SCP1 5 The World Copper Industry Its Changing Structure and Future Prospects Kenji Takeuchi, John E. Strongman, and Shunichi Maeda WORLD BANK STAFF COMMODITY WORKING PAPERS 1. The World Tin Economy: An Econometric Analysis (out of print) 2. International Cotton Market Prospects 3. An Econometric Model of the World Rubber Economy (out of print) 4. Industrial Processing of Natural Resources 5. The World Sugar Economy: An Econometric Analysis of Long-Term Developments 6. World Bank Commodity Models (2 volumes) 7. Analysis of the World Coffee Market 8. Analysis of the World Cocoa Market 9. The Outlook for Primary Commodities 10. World Rubber Market Structure and Stabilisation: An Econometric Study 11. The Outlook for Primary Commodities, 1984 to 1995 12. The Outlook for Thermal Coal 13. Jute Supply Response in Bangladesh 14. Prospects for the World Jute Industry WORLD BANK STAFF COMMODITY WORKING PAPERS Number 15 The World Copper Industy Its Changing Structure and Future Prospects Kenji Takeuchi, John E. Strongman, and Shunichi Maeda The World Bank Washington, D.C., U.S.A. Copyright (© 1986 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing November 1986 Commodity Working Papers are not formal publications of the World Bank, and are circulated to encourage discussion and comment and to communicate the results of the Bank's work quicldy to the development community; citation and the use of these papers should take account of their provisional character. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Any maps that accompany the text have been prepared solely for the convenience of readers; the designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. Because of the informality and to present the results of research with the least possible delay, the typescript has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The publication is supplied at a token charge to defray part of the cost of manufacture and distribution. The most recent World Bank publications are described in the catalog New Publications, a new edition of which is issued in the spring and fall of each year. The complete backlist of publications is shown in the annual Index of Publications, which contains an alphabetical title list and indexes of subjects, authors, and countries and regions; it is of value principally to libraries and institutional purchasers. The continuing research program is described in The World Bank Research Program: Abstracts of Current Studies, which is issued annually. The latest edition of each of these is available free of charge from the Publications Sales Unit, Department F, The World Bank, 1818 H Street, N.W, Washington, D.C. 20433, U.S.A., or from Publications, The World Bank 66, avenue d'lena, 75116 Paris, France. Kenji Takeuchi is an economist in the Economic Analysis and Projections Department, and John E. Strongman is a minerals economist and Shunichi Maeda a financial analyst in the Industry Department; all are at the World Bank. Library of Congress Cataloging-in-Publicafton Data Takeuchi, Kenji, 1932- The world copper industry. (World Bank staff conmodity working papers, ISSN 0253-3537 ; no. 15) Bibliography: p. 1. Copper industry and trade. 2. Market surveys. I. Strongman, John, 1947- . II. Maeda, Shurichi. III. Title. IV. Series: World Bank staff conmodity working paper ; no. 15. HD9539.C6T35 1986 338.2'743 86-26554 ISBN 0-8213-0852-1 ABSTRACT 1. The world copper market, depressed for more than a decade, has undergone a dramatic structural change. Depressed demand for copper was caused by the slowdown in world economic growth, a declining share of industry/manufacturing in GDP, substitution by other materials (aluminum, plastics, and, recently, optical fibers), and material saving technologies (downsizing/miniaturization). No clear evidence exists that increased participation of governments and oil companies has significantly changed the investment behavior and production policies of the industry. Since 1980, various cost cutting measures, closures of uneconomic operations by individual producers, and exchange devaluations have tended to lower production costs in real terms. 2. In the next decade, substitution by plastics and optical fibers will accelerate while substitution by aluminum, on balance, will slow down. Copper demand is expected to grow at 1.3Z per annum during 1984-95. Most additional demand is expected to emanate from industrializing developing countries while demand in industrial countries is expected to stagnate. While mining capacity in the United States, Zambia, and South Africa is expected to decline, capacity in several countries (notably Chile) is expected to increase. Apart from the new capacity already virtually completed (e.g., Sar Cheshmeh in Iran, Cananea in Mexico, Tintaya in Peru), major probable increases can be attributed to the expansion of the existing CODELCO and Disputada mines in Chile, additional leaching operations (mostly in North America), low-cost greenfield projects in Chile (Escondida) and Portugal (Neves-Corvo), and rich muiltiproduct projects in Australia (Olympic Dam) and Papua New Guinea (OK Tedi II). 3. The Base Case (most likely) outlook for the world copper market is that the price of copper will remain at the present low level in real terms through 1990 (63-64C/lb in 1984$) and then will rise moderately to 73-74c/lb (in 1984$) by the mid-1990s. This outlook is sensitive to certain key variables, such as economic growth assumptions and investment decisions regarding incremental copper capacity. Copper market prospects are less promising if economic growth is weak. In that case, prices would average only 58C/lb (in 1984 dollar terms). If, however, mine capacity grows more slowly than anticipated, resulting in a tighter supply/demand balance in the 1990s, copper prices could rise to 79¢/lb (in 1984 dollar terms) by 1995. - iv - ACKNOWLEDGMENTS The study was undertaken as a joint effort of the Economic Analysis and Projections Department and the Industry Department. The copper model used for this report was developed by C. Suan Tan (Economic Analysis and Projections Department) and will be reported in a separate World Bank publication. J.R. Bentjerodt (Industry Department) also contributed to the preparation of the report. G. Ukpong (Economic Analysis and Projections Department) provided research assistance. This report benefited from comments on earlier drafts by B.M. Chambers (UNCTAD), Phillip Crowson (RTZ Ltd.), Hans H. Landsberg (Resources for the Future), Raymond F. Mikesell (University of Oregon), S.N. Payton (International Wrought Copper Council), Marian Radetzki (Institute for International Economic Studies, University of Stockholm), G.F. Ray (National Institute of Economic and Social Research), Walter Sies (Metallgesellschaft AG), David B. Morris (Kennecott), R.R. Beebe (Newmont Mining Corp.), Douglas C. Yearley (Phelps Dodge Corp.), Alfredo Dammert (Empresa Minera del Centro del Peru S.A.) and Janice Jolly (United States Bureau of Mines). TABLE OF CONTENTS SUMMARY AND CONCLUSIONS ............................... xiii ABBREVIATIONS AND ACRONYMS .......................................... xxix I. INTRODUCTION .... ....................... ... .......1.. II. LONG-TERM HISTORICAL PERSPECTIVE ...... ................... 5 A. Consumption Trends, 1890-1983 ...........................5 B. Copper Production Trends, 1890-1983 ...........8...oo..oo... 8 C. Long-Term Price Trends, 1900-1984. .......8....... D. Highlights in the Period L900-1929 ................... 1 E. Highlights in the Period 1929-1950 ........................ 15 F. Highlights in the Period 1950-1963. ....................... 19 C. The High Price Period of 1964-1973 ..........................22 III. RECENT DEVELOPMENTS IN THE WORLD ECONOMY AND IN PRIMARY COMMODITY MARKETS. .......... o. .. . . ..... o....26 A. Changing World Economic Environment ......... ......26 B. Economic Performance in Developing Countries .........31 C. Primary Commodity Markets since World War II ........... 33 IV. STAGNANT CONSUMPTION AND OVERCAPACITY IN THE WORLD COP PER MARKET SINCE 1 974 ..............40 A. Slowdown in Demand Growth....... .........................40 B. Material Substitution and Material Saving Technology .......45 C. Chronic Overcapacity .........4.... .......... V. CHANGING COMPETITIVE POSITIONS AMONG PRODUCING COUNTRIES ........58 A. Changes in Production Cost Structure .........58 B. Impact of Co/By-Product Credits on Competitive Positions ....62 C. Impact of Cost Cutting Measures Undertaken by Producers since 1980.......o .... .......... . ....... .65 D. Currency Devaluations by Producing Countries .... 68 E. Trends in Production Costs in 1985........ ...............74 VI. IMPACT OF CHANGES IN INDUSTRY ORGANIZATION ON SUPPLY .......... 80 A. Changes in Ownership and Control ..........o.............. 80 B. Producers' Investment Decisions 1974-84 ......... o ............86 C. Production Policies 1975-84.o... .......................... .90 D. CIPECo ...... . o........... ..... o.o...o . o.o..o............ o ......... 98 - vi - VII. TRADE AND PRICES ..................... .......................... 99 A. Changing Patterns in Copper Trade ......................... 99 B. Impact of the Changing Exchange Value of the US Dollar on Copper Prices ...... ........................ 105 VIII. LONG-TERM MARKET PROSPECTS FOR COPPER ........................ 108 A. Likely Trends in the World Economy ................*........108 B. Projection Methodology and the "Base Case" Projections ......'.ll C. Demand Prospects ...... ................... .......................... 113 D. Supply Prospects ..... .. ........ ...... 123 E. Price Outlook ... ..#... .......... ............................132 ANNEXES ANNEX A. Copper Cartels, 1870-1940........ ........................ . 134 ANNEX B. Adjusted Net Costs of Copper Production: Method of Calculation.o ...... ..136 ANNEX C. Three Forms of Copper in Trade ....... .o.o..o.......138 ANNEX D. Oil Companies' Behavior in the Copper Industry ..... ........142 ANNEX E. Impact of Changing Value of US Dollar on Copper Prices ..... 146 ANNEX F. Summary of the Copper Model Structure.....................150 ANNEX G. China's Copper Demand Prospects-oo-oo ..................-154 ANNEX H. Impact of Fiber Optics on the Copper Market, 1986-95....... 160 ANNEX I. Long-Run Marginal Costs in Copper Production ... ............ 166 ANNEX J. The Survival Strategy of Japanese Copper Smelters .......... 171 REFERENCES ............................... ................ o..q ..o.9 ....... 175 - vii - LIST OF TABLES 2.1 Trend Growth Rates in World Copper Consumption and Industrial Production and Apparent Industrial Production Elasticities, 1890-1984...... ... ................ . .7 2.2 Trend Growth Rates in World Copper Mine Production (Copper Content) in Four Sub-Periods in 1890-1984 ............ 12 2.3 Average Refined Copper Price in London, Deflated by US Wholesale Price Index (1983 = 100), Covering Periods 1900-1984.. ................... 12 2.4 Productivity of Labor in Copper Mining in the U nited States, 1912-45 .Production.b CountryRegion,1600.. 4..........16 2.5 (A) Copper Mine Production by Country/Region, 1900-1984......... 17 2.5 (B) Copper Mine Production by Country/Region, 1900-1984 ......18 2.6 Prices in Defense Production Act Contracts.... ............ 21 2.7 Average Return on Investment, Major US Copper Companies, 1950-56 and 1957-63........ . ..... ... ............21 2.8 Prices of Copper, 1962-1969 ........... ...... .s* ............24 3.1 Real GDP Growth Rate, by Economic Region/Country 1960-1984 ......... . . . . . . .. . . . 27 3.2 Growth Rates of Consumer Price Index, Real GDP Per Labor Employed and World Trade Volume, Selected Region/Country, Selected Periods in 1960-1984 ............. e .. ....28 3.3 Exchange Rates of the US Dollar, 1970-1984...... ............30 3.4 Unemployment and Interest Rates in Industrial Countriesgt Index of3Cout..odi.y Prices.(Curren o .........32 3.5 Weighted Index of Commodity Prices (Current Dollars) ........34 3.6 Weighted Index of Commodity Prices (Constant Dollars) .....36 3.7 Manufacturing Unit Value (MUV) Index, 1948-95 ................38 4.1 World (Excluding CPEs) Consumption of Selected Major Metals, and Industrial Production and Real GDP in Industrial Countries: 1961-1984 ............ 41 4.2 Share of Value Added as a Percentage of CDP in OECD Countries, By Sectors, 1960-1983... ......... ......42 - viii - 4.3 Changes in Copper Intensity by Region, Selected Years ............................... . . 44 4.4 Refined Copper Consumption by Country/Region 1970-1984 ................................... 46 4.5 Apparent Copper Consumption by End-Use Industry, Western Europe, United States and Japan, 1970 and 1983 ....... 48 4.6 Consumption of Copper and Aluminum in Cable Manufacturing, 1964-1984 ....................... .. .......... 51 4.7 Consumption of Hot-Water Plastics Pipe In Western Europe as of 1980 ............................ . ........ ... 53 4.8 Refined Copper Stocks in Market Economies, 1960-84 ..........57 5.1 (A) Production Costs for Major Copper Producing Countries........................... . ............ . 60 5.1 (B) Production Costs for Major Copper Producing Countries .. . . . . . . . . . . . . . . . . . . . .* . . . .61 5.2 Average Prices of Selected Metals, 1978-1984 ............... 63 5.3 Change in Gross Costs Between 1980 and 1984 ...............65 5.4 Exchange Rate Changes and Inflation in Major Copper Producing Countries, 1980-1984............. .. .... . .69 5.5 Estimated Share of Local Content in Gross Production Costs in Major Copper Producing Countries, 1975, 1980 and 1984 .............................. 70 5.6 Ranking of Major Copper Producing Countries Based on Actual Net Costs, 1975, 1980 and 1984............ ... 71 5.7 Ranking of Major Copper Producing Countries by Net Costs as Adjusted Using the 1980 Exchange Rates as Equilibrium Rates, 1975, 1980 and 1984 ............. 72 5.8 Ranking of Major Copper Producing Countries Net Costs as Adjusted by Using the 1984 Exchange Rates as Equilibrium Rates, 1975, 1980 and 1984 .............. 72 5.9 Ranking of Major Copper Producing Countries Based on Actual Gross Costs, 1975, 1980 and 1984 ...... ............. 75 5.10 Estimated Production Costs in Major Copper Producing Countries in 1985 ................................77 5.11 Changes in Production Costs Between 1984 and 1985 ....1esee..78 - ix - 5.12 Ranking of Major Copper Producing Countries Gross Cost Basis in 1985 ................ * ................ . 79 6.1 Government Ownership Position in the Copper Industry, 1981 ...................................... ..82 6.2 Mine Production of Copper By State-Owned and Privately-Owned Companies, 1975, 1980 and 1983 ........ 85 6.3 Copper Mining Capacity at Beginning of Year, Selected Years, 1970-1983 ...... ...... so ............... 87 6.4 Copper Mining Capacity Utilization, Selected Countries/Regions, 1973-1983 ............................... 91 6.5 Changes in Capacity Utilization Rates in the Copper Industries of Selected Countries, in 1975, 1977, 1981 and 1982 ................. ..... ........... 94 7.1 Industrial and Developing Countries' Copper Exports, 1960-1984... . ......... * ........ ............... . 100 7.2 Imports of Copper in Ores and Concentrates, Selected Countries, 1960, 1973, and 1984 .................. 101 7.3 Imports and Consumption of Refined Copper in the United States, 1960-1984 .................. .......103 7.4 Copper Prices, 1950-1984 .............. 104 8.1 Industrial Countries: Real GDP Growth Rates, 1983-85 (Actual) and 1986-1995 (Projected) ... ............... 109 8.2 National Inflation Rates, SDR/US$ Exchange Rate and Interest Rates, 1973-85 (Actual) and 1986-95 (Projected) ......................................110 8.3 Past and Projected World GNP Growth Rates, 1973-95 ..........112 8.4 Summary of "Base Case" Projections for World (Excluding CPEs) Copper Market, 1983 and 1985 (Actual) 1990 and 1995 (Projected) .......................... 114 8.5 Consumption of Refined Copper - "Base Case" Projections for Selected Countries and the World (Excluding CPEs), 1990 and 1995 ............................ 116 8.6 Impact of Fiber Optics Technology on Copper Consumption in Major Industrial Countries, 1990 and 1995 .....................118 8.7 World Copper Reserves ... . .....oo9999...9.9 oo9.o..125 8.8 Copper Mining Capacity at Beginning-of Year, 1984-96 (Base Case) ..........*.*.*.*.*............................... ..... .126 8.9 Projected Mine Production of Copper, 1990 and 1995, "Base Case" (Case A) .......................... 129 8.10 Copper Mining Capacity at Beginning of Year, 1985-96 - Case C ..... ......................130 ANNEX TABLES C.1 Marketing Characteristics for Main Products of Copper .......141 D.1 Copper Mine Closures by Oil Company in the United States, 1981-1985..... ............................143 E.1 Copper Price Equations ................................... 147 E.2 Correlation Matrix for Inflation, Exchange-Rate and Interest-Rate Variables ................. ............ 149 G.A China, P.R.: Copper Production, Consumption and Trade Data, Industrial Production, Electricity Generation, 1967-1984 ............. o .................... .155 G.2 Per Capita Consumption of Copper, Selected Countries, 1982 ................................ 157 G.3 China: Alternative Projections for Industrial Production, Electricity Generation and Copper Consumption, 1990 ............................... 159 H.A Coefficients Used to Reduce Total Copper Consumption to Account for Fiber Optics Penetration ..................... 164 J.1 Copper Smelting and Refining Capacity, Production of Copper Ore, Blister and Refined Copper, Refined Copper Consumption in Japan, 1955-1984 ...................... 172 - xi - LIST OF FIGURES 2.1 World Primary Copper Consumption .............................. 6 2.2 World Copper Mine Production .................................. 9 2.3 London Copper Price ........... ..... ..... .. ....... . 10 2.4 Average Yield of Copper Mined in the United States . . 14 3.1 Weighted Index of Comodity Prices: 1948-1983 .......... 37 4.1 Market Economies' Copper Mine Capacity and Production ............. . . .... .............. 55 5.1 Classification of Copper Producers by Cost Characteristics as of 1 9 8 4 .............................. 64 6.1 Change in Net Cost by Co untry .................. 96 8.1 Refined Copper: Intensities of Use ................ 119 8.2 Refined Copper: Intensities of Use ........................ 120 F.1 Structure of the World Copper Market Model - Flowchart. ...... . . ......... ............ 151 --- - xiii - SUMMARY AND CONCLUSIONS Historical Perspective 1. World copper market conditions are influenced strongly by long-term world 1/ economic trends. During the past century four periods of alternating high growth and low growth have occurred--each lasting twenty to thirty years. Copper consumption and production grew strongly in the two periods of high world economic growth (from 1890-1918 and 1950-1973). Copper growth, however, was much lower during the economic depression of the period between the two World Wars (1918-1940) and in the period since 1973. 2. The first period of high economic growth, from 1890-1918, spanned three decades. Industrial production increased at about 5.8% per annum, and world primary copper consumption and mine production each grew at the same rate--of about 5.8% per year. In 1890, the average world ore grade for mined copper is estimated to have been about 6.0% and production cost for refined copper about US$1.25 per lb (in 1983 dollar terms). Over the 1900-18 period, the copper price averaged about US$1.45 per lb (in 1983 dollar terms). 2/ This was a period of major technological advances which included the introduction of large scale open-pit mining in copper and the development of froth flotation process of concentration. Both technologies helped in the rapid expansion of copper capacity as well as in reducing production costs. By 1920 the United States was established as the leading world producer of copper, responsible for over 60% of world copper production. 3. The second period from 1918-1949 was a period of relatively low copper market growth and prices. The end of World War I proved to be a turning point in world economic performance, as well as in copper market trends. The next three decades witnessed prolonged economic difficulties. From 1918 to 1948 industrial production growth averaged only about 3.3% per year and primary copper consumption and mine production averaged about 2.6% and 3.0% respectively. The dominance of the United States in the world copper industry declined as new mines were developed largely by American interests in Latin America (especially in Chile and Peru) and by European investments in Central Africa (especially Zaire and Zambia). In the face of recurring worldwide economic difficulties, demand did not keep pace with capacity development, a trend that resulted in oversupply for most of the period. The price in London averaged only 760/lb in 1983 dollar terms. 1/ Unless otherwise specified "World" excludes East European Nonmarket Economies--which are also known as Centrally Planned Economies (CPEs). 2/ Unless otherwise specified "production costs" refer to refined copper and the "price" refers to London Metal Exchange (LME) price for refined copper. - xiv - 4. The third period, from 1948 to 1973, witnessed a strong growth in the world economy that fuelled a major expansion of the world copper industry. During this period, world industrial production expanded at about 5.7% per year and primary copper consumption and mine production grew by about 4.8% and 4.7% per year, respectively. 5. During the late 1940s and early 1950s the copper market was subject to various forms of government intervention. To encourage the development of domestic copper capacity the US Government provided loans, tax amortization benefits and government purchase contracts under the Defense Production Act of 1950. Under this program some 55 projects were undertaken from 1950-56, and by 1960 a government stockpile of just over 1 million tons was built up. In an attempt to improve price stability, US producers maintained a producer price system. Although these producer prices tended to follow the LME price, they did not change as often or as much as the LME price. During much of the 1950s and early 1960s there was a tendency toward copper over-supply. From 1947 to 1963, the LME price averaged about US$1.00 per lb (in 1983 dollar terms), with US producer prices averaging at similar levels. 6. The period from mid-1960s to mid-1970s was a dynamic period for world economic growth and the world copper market, and LME copper price averaged US$1.67 per lb (in 1983 dollar terms). In this period, US producers maintained much lower prices than the LME price. Because of the low price US producers occasionally had to introduce quotas to customers who in turn had to purchase additional supplies at the much higher LME price. The strong copper demand growth and the high copper prices encouraged a rapid worldwide expansion of copper production facilities. In particular, large copper processing (smelting/refining) industries were established in Japan and Western Europe, largely based on supplies of primary materials (ores/concentrates) from new mine developments and expansions of existing mines in the Pacific Basin, Africa and Latin America. Many of these mine developments were supported by loans from official export financing institutions in Japan and Western Europe. As a result, exports of copper concentrates (including cement/matte) increased dramatically, rising from 280 to 1,130 thousand tons (copper content) in the 1960-73 period. The share of concentrates in total copper exports rose from 9.8% to 25.3% during the same period. 7. The fourth period--from 1974 to the present--has largely been a period of worldwide economic disarray, substantial copper over-supply and persistently low copper prices. From 1973 to 1984, world industrial production increased at 2.8% per year (vs. 5.7% p.a. in the 1948-73 period) and primary copper consumption and mine production increased at about 1.3% and 1.0% per year, respectively. The LME copper price averaged US$0.87 per lb (in 1983 dollar terms) in the 1975-85 period, well below the level in the period from 1948-63. The price has been especially low in the last five years, averaging 660/lb (in 1983 dollar terms) in the 1981-85 period or well below the 1920-47 average level. Since the late 1970s US producer prices have generally stayed closely tied to the LME price. - xv - 8. The difficulties of the world copper industry have arisen during a decade of deep-rooted macro-economic changes in the world economy from 1975 to 1985 following over two decades of sustained worldwide economic growth. By 1973, when the first oil shock occurred, most industral countries were facing a combination of high inflation and low economic growth. Industrialized countries then plunged into a period of continued low economic growth, and high inflation, with increased unemployment, persistent trade imbalances and shifting exchange rates. The US dollar steadily weakened against other major currencies during the 1970s. At the same time, large industrial restructuring efforts were called for to reduce requirements for energy, particularly for oil. 9. When the industrial countries seemed to be recovering from the first oil shock, the second round of oil price increases occurred in 1979 and 1980. Control of inflation became a priority policy objective for the industrial countries, especially the United States and Western Europe, and interest rates began to rise. In 1981, the US prime rate and the Eurodollar interest rates averaged over 18% and over 16% respectively--compared with an average of 8-9% for the period 1973. Although governmental efforts to contain inflation were generally successful, a severe recession occurred and real Gross Domestic Product (GDP) in industrial countries declined by 0.3% in 1982. Since 1983, the industrial economies have recovered with 2.2% GDP growth in 1983 and 4.7% in 1984. Recovery, however, has brought with it sharp trade imbalances--especially for the United States--and tendencies toward increased trade barriers. 10. Unlike the industrial countries, developing countries experienced a continued strong economic growth in the mid to late 1970s averaging 5.5% per year in the 1973-80 period. Most developing countries were able to mitigate the effects of the 1974 oil price increases by external borrowing--largely made possible by the ready availability of "petrodollars" at low real interest rates. This, of course, increased developing countries' debt as a percentage of Gross National Product (GNP) from 14.1% to 20.9% in the 1970-80 period. When the second round of oil price increases occurred in the 1978-81 period the developing countries had to deal with it at a time when their indebtedness was at an all time high and real interest rates were also rising sharply. Thus, economic growth for developing countries slowed down markedly to 2.5% p.a in the 1981-83 period and debt servicing for developing countries became a major international issue. 11. In the face of debt-service pressures, almost all oil importing developing countries have had to take drastic measures to reduce real domestic absorption and to increase exports to deal with their serious balance of payments problems. These measures include, among others, devaluing their currencies substantially (in nominal and real terms) and radically restructuring their economic management policies. 12. While the developing countries' macroeconomic stabilization efforts have generally been successful, two groups of countries still suffer from severe debt servicing difficulties due to past heavy borrowing, inadequate policy adjustment and over-dependence on primary commodity exports. One group - xvi - includes the low income African countries constrained by weak institutional structures, heavy dependence on a few primary export commodities, shortage of skills and the occurence of natural disasters; among these are two important copper exporters, namely, Zambia and Zaire. The other group includes countries, especially in Latin America, with a heavy international debt burden and relatively high inflation rates. This group includes important copper exporters such as Chile, Peru and Philippines. 13. Many countries in these two groups (including major copper-exporting countries) have initiated extensive efforts to increase exports (not only copper but all exports) and reduce imports. Exchange rates have been adjusted to improve export competitiveness. In major copper exporting developing countries, various measures are being undertaken by copper producing entities to improve efficiency and to reduce production costs, including the closing of non-economic operations. For many countries, copper exports represent the primary mechanism--at least, in the short term--for generating the foreign exchange earnings necessary to service their external debts. Competition in the world copper markets has thus increased substantially at a time when the growth of demand has been slow. This situation has tended to keep copper prices low by historical standards. 14. The poor economic performance of the industrial countries in the mid- 1970s and early 1980s tended to depress international primary commodity prices across the board. From 1973 to 1977, the price indexes for both metals/minerals and for agricultural commodities declined by about 30% in real terms. After falling further in 1978, both indexes reached cyclical peaks in 1980, but fell sharply in 1981-82. They recovered moderately in 1983 but recovery did not continue in 1984, in spite of the renewed economic growth in industrial countries. Measured in US dollars commodity prices were negatively affected, not only by low demand growth but also by the renewed strength of the US currency. The copper market trends have been no exception. Changes in Copper Demand 15. Refined copper consumption is presently concentrated in the industrial countries with about 26% of world (including CPEs) consumption in Western Europe, 23% in North America and 14% in Japan, compared with 39%, 29% and 6Z, respectively, in 1960. Developing countries, now account for 17% of world consumption, mainly concentrated in a few industrializing developing countries such as China, Brazil, Republic of Korea, Yugoslavia, India, Mexico, Turkey and Argentina. The balance (about 19%) is accounted for by CPEs. 16. In terms of end-use sectors, copper consumption takes place mostly in capital intensive sectors such as electrical (50%), construction (15%), general engineering or machinery (18%) and transportation (9%). The pattern of consumption has changed only slightly. Back in 1970, electrical uses accounted for 48% of the total, with construction, general engineering and transportation making up for 16%, 19% and 10% respectively of total use. 17. Copper consumption growth has been reduced both by weak economic growth in industrial countries and by sectoral shifts away from industry - xvii - (including manufacturing) towards services that are less copper intensive. Copper consumption growth has also been slowed by changes in product design, e.g., miniaturization in electronics products and downsizing and downgauging of final consumer products especially in the transportation sector. Overall, the intensity of copper use per million US dollars of GDP (measured in constant 1980 dollars) in industrial countries peaked at 1,125 kg in 1965 and declined to 835 kg in 1981. 18. The decline in copper intensity of use is also linked to competition from substitute materials, especially aluminum, plastics and recently optical fibers. Massive substitution for copper by aluminum took place in electricity conductors in the 1950s and 1960s when relative prices favored the use of aluminum. But aluminum's inroads into that important market for copper slowed down in the 1970s. Aluminum, however, is also a prime substitute for copper in heat exchange applications. Substitution in this area was relatively light in the 1950s and 1960s, but since the mid-1970s the use of aluminum for automobile radiators has developed rapidly, especially in Western Europe where 65% of all new cars were equipped with aluminum-based radiators in 1983. Aluminum's penetration in the United States and Japanese automobile radiator application was only 15% and 4% respectively in 1983 but is likely to increase in the future. 19. With regard to plastics, the major substitution for copper has been in hot and cold water pipes and fittings by polyvinyl chloride (PVC). This substitution has become significant in the last decade, and is likely to accelerate in the next decade. More recently, optical fibers have started to substitute for copper in the telecommunication industry especially in the United States, Japan and the United Kingdom. 20. Copper consumption in developing countries grew strongly in the 1970- 81 period increasing at 8.1% per year. Their consumption fell in 1982 and 1983 in the face of recessionary pressures but revived sharply in 1984 and continued to grow in 1985. Copper intensity of use is still increasing in developing countries where economic growth is expected to result in substantial increases in copper needs. However, the copper intensity in developing countries is unlikely to reach the same peak levels as in industrial countries because many current technological trends in industrial countries--miniaturization, use of substitutes, etc.--are already taking place in developing countries as well. For example, fiber optics technology, which has started to displace copper in telecommunication applications in industrial countries, is recently making an impact in developing country markets as well. Changes in the Producer Shares 21. Copper mine production today takes place predominantly in the developing countries, which account for 52% of world total. North America, accounts for 21%. Historically, the market share of North American mines has declined from over 60% in 1920 to less than 25% today and the market share of mines in developing regions has increased from less than 25% in 1920 to over - xviii - 50% today. Currently, developing countries in Latin America, Africa and Asia account for about 25%, 15% and 10% of world mine production, respectively. 22. Trade of copper products takes place in various forms. The trade includes both refined copper, which goes directly to copper fabricators and semifabricators, and copper concentrates and blister which require further processing (smelting and/or refining) There are large copper processing (smelting and refining) industries in Western Europe and Japan that depend on imported unrefined copper. In addition, in industrial countries, secondary refined copper is produced from old scrap which accounts for about 15% of refined copper consumption. Changes in Cost Structure 23. Copper mining costs are generally lower in the developing countries because they tend to have higher ore grades, lower labor costs and more valuable by- and co-products than the industrial countries. The highest-cost producers are those countries that have high gross production costs and low co/by-product credits. The United States and Zambia are in this category. These producers are most vulnerable to weakening copper prices. The majority of large copper producing countries have high gross costs and high co/by- product credits. A decline in their copper prices can be offset if the prices of co/by-products remain relatively high. Producers that fall into this broad category include Australia, Canada, Mexico, Peru, Papua New Guinea, Philippines and South Africa. Producing countries with low gross costs and low co/by-product credits include Indonesia and Chile. Only one country, Zaire, falls in the category of low gross production costs coupled with high co/by-product credits. The last two groups of producers, that is, Indonesia, Chile and Zaire, are the most resillient in the face of depressed copper and/or co/by-product prices. 24. Over the last decade, the copper industry has experienced dramatic changes in its production cost structure. External factors that triggered these changes include (a) drastic increases in energy costs caused by the two rounds of oil price hikes, (b) tighter pollution control standards and their strict enforcement in industrial countries, (c) high inflation followed by high interest charges, and (d) large and frequent changes in exchange rates between the US dollar and other major currencies and between the dollar and the currencies of copper producing countries. These external factors, combined with major efforts by key producers to ensure the survival of their copper sectors, have accelerated the pace of change in the copper industry. - xix - 25. From 1975 to 1980 average gross production costs (i.e., costs before adjustments for co/by-product credits 1/ in the world 2/ increased from about US$0.71 per lb to US$1.09 per lb (both in current dollar terms). However, during the same period, by-product credits increased from US$0.22 per lb to US$0.59 per lb, resulting in average net production costs 3/ remaining at about US$0.50 per lb. While production costs for many producers were relatively unchanged, the cost position of the US producers deteriorated considerably (from US$0.62/lb to US$0.73/lb from 1975 to 1980) because they did not benefit as much from the presence of co/by-products. 26. From 1980 to 1984, gross costs decreased to US$0.93 per lb, but by- product prices declined sharply (to US$0.37 per ib) with the result that net costs increased to US$0.57 per lb. Strenuous efforts were made by the producers in the higher end of the cost curve, especially in North America, to maintain or reduce their costs under the difficult market circumstances of a prolonged depression of copper prices and decreases in co/by-product prices. Most cost reductions resulted from (a) various direct cost cutting measures including large reduction in labor costs, especially in the United States, (b) cutbacks in and/or closing of uneconomic operations, and (c) changes in valuation, reflecting currency realignments vis-a-vis the US dollar. In early 1986, the sharp decline in oil prices aided cost-cutting efforts. 27. In the 1980-84 period, most major copper producing countries experienced a significant depreciation of their currencies against the US dollar. The depreciation applied not only to copper producing developing countries, but also to Japan, F.R. Germany and other European countries with smelting and refining industries. The depreciations, which exceeded the differentials between domestic and US inflation rates in many areas, reduced copper production costs in these countries in terms of US dollars. For countries whose economies are highly dependent on primary export commodities such as copper, depressed primary commodity prices tend to cause pressure for the countries to devalue their currencies because of the need to restore balance of payments positions. Improved cost competitiveness derived from the currency devaluation can lead to an increase in primary exports (volume) as long as unused production capacity, which can be operated to cover variable costs, is available. 1/ Unless otherwise noted, production costs in this report include all direct and indirect operating costs and interest payments but exclude depreciation/depletion charges. 2/ The mines covered here account for some 80% of the market economies' mine production. 3/ Net production costs = gross production costs minus by/co-product credits. - xx Types of Producers 28. Copper producers today can be classified into different categories according to the type of ownership and the extent of integration of production with processing operations. In terms of ownership, it is often considered useful to distinguish between state-owned companies and private companies. Private companies, in turn, can be divided into independent producers and those owned by oil companies. These four broad types of copper producers can be further classified by the stages of production in which they are involved, namely, those in mining only, those in smelting and refining only and those possessing fully integrated facilities (mining, smelting and refining). 29. Until the early 1960s, a large part of copper production was in the hands of a few multinational corporations which were mostly integrated domestically or internationally. Since then, the structure of the industry has changed, becoming more diversified both geographically and institutionally. The change occurred as a result of three major trends: (i) Increased production by non-integrated mining companies (i.e., engaged in mining only) in the mid-1960s to mid-1970s. (This trend became strong as demand for copper concentrates increased as a result of rapid capacity expansion of companies engaged in smelting and refining only, especially in Japan and Western Europe.) (ii) State participation in the ownership (or control) of copper producing activities (especially in Latin America and Central Africa in the late 1960s and early 1970s). (iii) Takeovers of copper producing companies by energy-related multinational companies (especially in North America in the mid-1970s to early 1980s). 30. In the early 1960s, with consumption rising rapidly, and with a significant lack of domestic copper resources, the governments in Japan and West European countries such as (F.R. Germany) became concerned with long-term security of supply of non-fuel minerals. They decided to provide financial assistance to their growing non-ferrous smelting/refining industries to secure overseas supply sources of ores and concentrates. Copper smelter production in Japan grew from 187,000 tons in 1960 to one million tons in 1973. In F.R. Germany it increased from 62,000 tons in 1960 to 233,000 tons in 1973. As a result of the Japanese and (to a lesser extent) German smelting companies' aggressive search for sources of copper concentrates, a number of new mines developed in this period, and additional mines developed in the 1974-77 period. Many of these new mines were non-integrated (i.e., strictly ore/concentrate producers) and were located in Canada, Latin America and the Pacific Basin. These changes in location and structure made the world copper industry more diverse. 31. For a variety of historical, political and economic reasons, state participation in the copper industry increased significantly up to the early 1970s. At the beginning of the 1960s copper production in which governments held any sort of interest amounted to less than 100,000 tons a year, or 2.5% of capacity in the market economies, but by 1970 the total had risen to some 2.25 million tons, or about 43% of market economies' capacity and has remained in the 40-50% range since. - xxi - 32. As steep increases in oil prices around 1973-74 led to increases in profits, oil companies searched intensely for investment opportunities and began to view the non-fuel minerals sector as a potential area for diversification. The copper industry became a prime candidate for diversification. The prospects for copper seemed good and the technology and marketing involved seemed not too difficult to acquire. However, after a flurry of copper company acquisitions and mine investments from 1977 to 1982, some oil companies found their involvement yielded only marginal returns. Many oil companies have already divested their copper interests, or are actively seeking opportunities to do so. Capacity Expansions since 1973 33. In the 1974-77 period, because of the high copper prices and optimistic market outlook prevailing in the 1964-73 period, a wide range of countries expanded their production capacities, primarily by the initiative of private mining companies (often involving multinationals). In contrast, in the 1977-83 period, the majority of capacity increases (notably in Chile and Mexico) were undertaken by state-owned companies. Their production costs ranged in the lower half of the worldwide cost curve. In this period, capacity was also increased significantly by private sector initiative. In the 1977-83 period, sizable reductions in capacity were made by both state-owned and private companies with production costs in the upper half of the worldwide cost curve. 34. The decisions to increase capacity by producing companies, whether state-owned or privately-owned, were often influenced by attractive offers of long-term purchase contracts and financing from potential buyers (often Japanese smelting/refining companies). Also, additional mine production was encouraged by smelters bidding up the prices for concentrates in the late 1970s/early 1980s when the supply/demand balance for concentrates was relatively tight. Another important reason for increasing capacity was to lower costs by spreading fixed charges across a greater output. 35. In terms of the economic viability of the added capacity during this period, some of the projects turned out to be very competitive, while others were not justified on cost competitiveness alone. Both state-owned and privately-owned companies appear to have made miscalculations in their investment decisions during this period. These miscalculations can be attributed partly to the lingering effects of the Club of Rome's Limits to Growth syndrome. 1/ 1/ In 1972, a report was made to the Club of Rome. Among other things, it concluded that the world's population would peak out in the mid-21st century and then slide into an abrupt decline, which would be caused by the virtual depletion of nonrenewable resources and the intolerable pollution. - xxii - 36. During the decade preceding 1975, mine capacity in market economies expanded rapidly (at 4.8% per annum) as new investment responded to high demand growth and high prices. Yet there was little excess capacity, and capacity utilization averaged about 93%. From 1975-83, worldwide capacity utilization averaged only 86.6% as demand growth declined. In spite of this decline, substantial overproduction occurred and large copper inventories were built up. Copper production is relatively inflexible and, because of inherent "exit barriers", does not respond readily to changing market conditions. Historically, copper supply has adjusted to demand cycles through the start up and close down of smaller high cost operating units. In the face of declining ore grades and rising labor costs, however, producers have increasingly emphasized large-scale, capital-intensive operations with low variable costs but high fixed costs and financial charges. Such operations are not readily amenable to cut-backs in production. Capacity Utilization Decisions in Surplus Market 37. A large portion of copper production is from multi-product ore bodies (especially in Canada, Mexico, Papua New Guinea, Peru, Philippines and Sweden) where production decisions are influenced by co/by-product market conditions as well as by copper market conditions. During the past decade, low cost producers, both state-owned companies (such as CODELCO in Chile and Gecamines in Zaire) and private companies (such as Southern Peru Copper Corp. (SPCC) in Peru, Palabora in South Africa and Bougainville in Papua New Guinea as well as all copper-leaching operations in the United States and elsewhere), have tended to produce at full capacity. So have many producers with major co- or by-products (such as nickel, lead, zinc, gold and silver) in Sweden, Australia, Mexico, Peru and Canada. Furthermore, many of these mines supply concentrates to the Japanese and European smelters under long-term contracts. With the opening of smelters in the Republic of Korea and Philippines in the last decade, concentrates were in short supply. Many mines began to produce at full capacity to meet the demands of the custom smelters. 38. To the extent that mine cutbacks and closures have occurred, they have been found among high cost producers with low copper grade and little or no by-product credits, located especially in the Philippines and in North America (although nickel-copper producers in Canada also cut back their operations because of depressed nickel market). In the decade 1974-83, capacity utilization in North America averaged 78% compared with 92% in the rest of the market economy countries. 39. The especially heavy concentration of capacity closures and production cutbacks in North America, particularly in the United States, can be attributed to the fact that the buyers in these areas are sensitive to the short-term changes in market conditions and rely on spot and short-term purchase contracts. Captive mines operated by companies fully integrated downstream also are sensitive to changing market conditions and react quickly in reducing mine production. In contrast, copper producers exporting concentrates in other major producing countries such as Chile, Peru, PNG and Indonesia who sell their output largely on the basis of long-term contracts covering at least one year (up to 15 years), tend to be more protected against - xxiii - sudden sharp declines in offtake. The smelting companies in Japan and Europe that purchase concentrates are much more sensitive to long-term security of supply and tend to accept the full contracted quantities in the face of depressed markets. Such long-term-minded buyers of copper concentrates have tended to get contract terms that ensure the continued operation of mines to supply their needs even during depressed price periods. Copper Pricing 40. Refined copper prices have historically followed the broad pattern of demand growth with periods of high prices (averaging in the range of US$1.35- 1.45 per lb in 1983 US dollar terms) associated with high demand growth. This was the case in the years prior to World War I and in the 1950-73 period. Between the two world wars and after the first oil crisis, prices averaged at little more than half the levels in the periods of high prices, averaging at between US$0.70 and US$0.80 per lb (in 1983 US dollar terms). Since 1981, prices have averaged only US$0.66/lb (in 1983 dollars). 41. Because copper is heavily used in the capital goods and consumer durables sectors, copper demand is sensitive to short-term changes in economic conditions, especially the investment cycle. Therefore, there have been sharp short-term fluctuations in copper usage within the long periods of high and low copper consumption growth noted previously. The cyclical nature of copper consumption is further accentuated by consumers' tendencies to build stocks of both refined copper and copper-based manufactured products downstream to ensure secure supplies during times of copper market upswings (which increase the requirements for copper) and to de-stock during downswings (which accentuate demand downturns). Copper prices are volatile and change rapidly in response to shifting supply/demand conditions. The volatility of the copper price reflects an inherent tendency of the market to supply/demand imbalances because supply is relatively inflexible in contrast to rapid changes in demand. 42. US copper producers (and Canadian producers selling in the US market) have been selling copper on the basis of producer prices since the 19th Century. Until 1978, the producer price had fluctuated considerably less than the LME price. This was possible because the United States was largely self sufficient in copper and partially isolated from outside influences because US copper prices were generally below world prices (until the mid-1970s). Other producers tried to establish producer prices unrelated to daily fluctuations of the LME price (e.g, RST in the 1950s, African and Chilean producers in the 1960s), but abandoned the attempt. In late 1978, in the face of strong import competition, US producers decided to let the producer price follow the LME price changes more closely. Since then they have been changing their producer prices more frequently (on a weekly basis). In recent years, the US producer prices have averaged about 6% higher than the LME price. 43. Refined copper prices, expressed in current US dollar terms, have been especially depressed since 1982 in part due to the strength of the US dollar against other currencies. In terms of constant value domestic currencies, copper prices have declined more sharply since 1980 for US - xxiv - producers than for non-US producers. Correspondingly, refined copper prices in terms of constant value local currencies have been relatively high by historical standards for consumers outside the United States until the exchange value of the US dollar vis-a-vis other major currencies declined sharply in the second half of 1985. 44. The strength of the US dollar against the currencies of Japan and major European countries in the period from mid-1981 to mid-1985 also had an effect of reducing the costs of smelting/refining in these countries in US dollar terms. This benefited the mines that export concentrates to Japan and Western Europe (including those in the United States), because the lowered (in US dollar terms) smelting/refining charges resulted in concentrate prices having been stronger than they would otherwise have been. This provided the mines with some relief from depressed refined metal prices. This tendency was further encouraged by a shortage of concentrate supply vis-a-vis smelter capacity in the early 1980s. As the Japanese and European currencies appreciated against the US dollar since the third quarter of 1985, however, smelting/refining charges for concentrate contracts have tended to rise sharply in US dollar terms. Projection Methodology 45. The outlook for the copper industry in industrial and developing countries over the next ten years has been examined with the help of an econometric model. A modelling approach is used since it provides the understanding of the dynamic structure of the industry and allows the derivation of alternative scenarios, consistent with a variety of assumptions concerning the future evolution of the key variables--e.g., income growth, exchange rate changes, substitution possibilities, etc. Different response parameters can also be evaluated with the help of a quantitative framework. 46. The model used has the following special characteristics. First, given the concern with stagnant consumption trends, emphasis was placed on the demand side. Consumption in six major consuming countries (United States, Japan and four major EEC countries) and in "rest of the world" was modelled for each of five major end-use sectors (i.e., electrical, construction, transport, general engineering and domestic use sectors). 47. Second, in view of the failure on the part of other econometric copper models to represent mine capacity investment behavior in a form that is useful for predictions (especially, on a country-by-country basis), it was decided to treat mine capacity exogenously. It is difficult to treat investment behavior by an econometric model in a satisfactory manner because of the wide range of investment criteria used by different investors, the lumpy nature of investment decisions, the complex nature of the effects of expectations on investment decilsions and the possible long (and varying) delays between investment decisiohs and full completion of new capacity. 48. Mine capacity is thus projected judgmentally for eleven main primary producing countries plus "rest of the world" on the basis of available information on planned and potential capacity increases/decreases andmine-by- - xxv - mine cost estimates. To the extent that the impact of projected price trends on capacity adjustments is taken into account (judgmentally), the process of projecting mine capacity is done "interactively." Mine production is then determined by projected mine capacity and the price of copper (projected endogenously) adjusted for inflation (which is assumed to reflect the trends in costs). 49. Inventories, scrap supply and the price of copper are all determined within the model. Net trade in refined copper with CPEs is treated exogenously and is assumed to remain constant at 120,000 tons per year for the projection period. The model represents an advance on other econometric models of the copper industry in that it allows for the effects of real interest rates and exchange rates on refined copper prices. 50. The projections derived from the model represent a combination of expert judgement as well as model behavior and are consistent with the "realities" of producers' cost structure, probable production and investment strategies. Considerable work has been undertaken on the cost structure of existing producers and for possible greenfield projects as well as for expansions of existing operations. 51. Copper market forecasting based on this model requires several other exogenous inputs to be specified explicitly for the projection period, e.g., the industrial production index for each major consuming country, prices for substitute materials, real interest rates and exchange rates and inflation indexes for major producing and consuming countries. The model results provide a useful basis for examining future market conditions and price movements, (a) because the model performs reasonably well in terms of simulations for the sample period (historical period), and (b) because it behaves robustly in providing future simulations, with broadly "sensible" results under a realistic range of assumptions. Demand Outlook 52. In view of the uncertainty regarding future economic growth rates, two economic scenarios have been examined. The base case (Case A) assumes that economic growth in industrial countries in the 1985-90 period returns to a moderately high growth path (3.1% p.a.) and improves further in the 1990-95 period (3.4% p.a.). Developing countries are assumed to grow at 4.3-4.7% per year in the 1985-95 period. A low alternative case (Case B) examines the consequences of 20% slower economic growth in industrial and developing countries for the entire period. 53. In the base case (Case A), copper consumption is projected to grow at 1.3% p.a. between 1984 and 1995 and mine production at 0.8% p.a. in the 1984- 95 period. For the next several years, supply is expected to remain plentiful vis-a-vis US demand and the LME price in 1984 dollar terms is projected to remain near present low levels (63-65c/lb) until 1990. After 1990, the supply-demand balance should be tighter and the price is expected to rise moderately to 74C/lb by 1995. These price estimates do not take into account - xxvi - short-term business cycles. Therefore, prices could be 10-20% higher or lower than these trend estimates in any single year. 54. Demand growth is expected to come mostly from developing countries, especially such industrializing countries as Brazil, China, India, Republic of Korea, Mexico and Turkey. Consumption in industrial countries as a whole is not expected to grow because any possible modest increase expected in some of these countries is likely to be offset by declines elsewhere. 55. The demand projection reflects likely effects of technological and substitution changes in end-use markets including the expanded use of fibre- optics technology in the communications sector. The loss of market for copper due to the effects of optical fiber technology is estimated to reach 400,000 tons a year world-wide by 1995. The projections imply continued declines in the copper intensity of use in industrial countries. 56. The base case projections assume a net increase of 300,000 tpy in world copper mine capacity by 1990 and a net addition of only 100,000 tpy in the 1990-1995 period. Most capacity increases are to be found in Latin America, especially Chile. In Chile, an incremental capacity of 150,000 tpy is expected to be developed by CODELCO, and Utah International, together with RTZ Ltd. and Mitsubishi Corp., are likely to develop a new 300,000 tpy capacity in Escondida. Other probable developments include the Olympic Dam project (150,000 tpy potential capacity by mid-1990s) in Australia, the OK Tedi project in Papua New Guinea (30,000-50,000 tpy by mid-1990s) and the Neves-Corvo project in Portugal (up to 120,000 tpy capacity by mid-1990s). These projects have very good copper ore grades (Olympic Dam and Neves-Corvo) and very good co-products (OK Tedi and Olympic Dam). RTZ (or its subsidiaries) is involved in both Neves-Corvo and OK Tedi, as well as Bougainville, giving it a significant stake in the likely copper expansion of the next decade. US mine capacity is expected to decline to 1.05 million tpy by 1995. Similarly, Zambia's capacity is expected to decline from 600,000 tpy as at the beginning of 1984 to 440,000 tpy by 1995. Weaker Economic Growth Scenario 57. Copper industry prospects are less promising if weaker economic growth occurs. A second set of alternative projections (Case B), which is based on lower economic growth rates (20% less economic growth than the base case, across the board) but with the same mine capacity assumption as the base case, indicates an annual average consumption growth rate of 0.8% in the 1984- 95 period and a copper price of US$0.58 per lb (1984$) in 1995. Case B reflects a tendency towards greater oversupply than in the base case. Promotion of Copper 58. If copper producers and fabricators are successful in promoting the use of copper, however, copper consumption growth and copper prices higher than indicated by the base case could materialize. Traditionally, the world copper industry has spent little on promoting copper consumption. Estimates show that, as of 1985, the total annual budgets of various copper promoting - xxvii - organizations amounted to only about US$10 million. Individual corporations have spent little for promotion purposes, in contrast to the producers of such substitutes as aluminum and plastics. Because the degree of vertical integration down to the final products is far weaker in the copper industry than in the aluminum and petrochemical industries, the individual copper producers have much less incentive to spend funds on product promotion, with the externality involved being so strong. Nevertheless, for a large copper producer such as CODELCO (with annual production of one million tons), it may make economic sense to take an aggressive initiative to spend funds for promotion of copper consumption on the basis of well identified opportunities. This is especially so if such efforts are undertaken jointly with other major producers and/or fabricators. No separate set of projections for such a scenario is presented in the report. Slower Supply Growth Scenario 59. A third set of projections (Case C) is based on the assumption that producers take a more cautious approach to new copper investments than expected at present so that mine capacity grows more slowly in the 1985-95 period than the base case, while other base case macro-economic assumptions remain the same. Under this scenario, capacity in the United States and Zambia decline more rapidly and Chilean capacity grows more slowly than in the base case. Consequently, world mine capacity would be lower by 3% in 1990 and by 4% in 1995 than the base case. This scenario indicates a tighter balance between supply and demand and a more rapid price increase. The price reaches 68C/lb (1984$) in 1990 (as compared with 63C/lb in the base case) and 79C/lb in 1995 (as compared with 74C/lb in the base case). Other Notable Trends 60. Copper prices over the next ten years are likely to be not only lower in real terms but also less volatile than they have been over the last two decades. Reopening retired mines and facilities takes less time and cost than developing new mines. Over the next ten years, many mines would remain closed if prices behave as forecasted in the base case. Some of these idle mines, however, could be brought back into production if prices start rising above the base case projections, thus moderating any price increase. Indeed, for prices to remain at a level well above the forecast for any extended period, it would be necessary for consumption to rise considerably faster than expected now and for output to experience significant shortfalls (a situation that could arise if a large producing country such as Chile experiences protracted production difficulties). 61. Despite the closure of a number of private mines in the United States, the share of state-owned copper companies in world copper supply is not likely to increase. Chile's CODELCO is likely to contribute significantly to expansion of world capacity, but Chilean private mines are also likely to expand by similar amounts. CODELCO has by far the largest potential for low cost expansion. However, few state-owned copper mines are expected to expand capacity outside Chile. Moreover, the expected reduction in Zambia's capacity is likely to contain state-owned companies' aggregate expansion worldwide. - xxviii - 62. Although oil companies have been divesting themselves of copper producing units in the last 2-3 years, they are likely to continue to play a noteworthy role in the copper industry. It is interesting to note that the closures and divestitures of copper producing units by oil companies have taken place mostly in the United States. Oil companies have retained their interest in promising copper properties located in such countries as Australia, Chile, Peru and Papua New Guinea. Oil companies' recent behavioral pattern suggests that: (a) they are quick to withdraw from economically marginal projects; (b) they maintain an international orientation; and (c) they are ready to inject their funds into promising projects wherever and whenever these are available. 63. Future industry production and investment decisions require great attention to cost competitiveness in order to generate earnings and avoid possible large losses. The various scenarios provided by the model underline the vulnerability of the industry's prospects to world economic trends. Over the past decade, the restructuring of the industry has occurred in the face of persistent over-supply. It has resulted in shifts in the geographical location of the industry, as well as changes in industry ownership structure, including divestment of copper (and other) mining operations by multi-national parent companies. 64. A possible turnaround in the industry's financial fortunes is indicated if a sustained economic recovery occurs or a large segment of high cost producers close permanently. However, it is possible that the industry will face several more years of intense competition and low prices if lower economic growth occurs and/or untimely investments take place. Under such circumstances, only low cost operations with cost conscious management will be able to remain profitable. State-owned and private producers alike will need to assess their competitive strategy carefully in the light of evolving demand trends not only for refined copper but also for trade in different forms of copper, i.e., copper concentrates, blister copper and refined copper. They will need contingency plans to minimize the costs of adjustments in the event that low prices persist. 65. Finally, if the evolving initiatives among the copper producers and fabricators for cooperative market development for copper uses should prove to be successful, copper consumption growth could be moderately stepped up, resulting in somewhat higher prices in the projection period. - xx ix - ABBREVIATIONS AND ACRONYMS bbl. Barrels CPEs Centrally Planned Economies CPI Consumer Price Index CODELCO Corporacion Nacional del Cobre de Chile COMEX New York Commodity Exchange E/MJ Engineering and Mining Journal CDP Gross Domestic Product IFC International Finance Corporation IMF International Monetary Fund IWCC International Wrought Copper Council LME London Metal Exchange LRMC Long-Run Marginal Cost MUV Manufactured Exports Unit Value MFN Most Favored Nation OECD Organization for Economic Cooperation and Development OPEC Organization of Petroleum Exporting Countries SPCC Southern Peru Copper Corporation SDR Special Drawing Rights tpy Tons per year USBM United States Bureau of Mines WBMS World Bureau of Metal Statistics ZCCM Zambia Consolidated Copper Mines i j t I. INTRODUCTION 1. The world copper market has been in a slump for the last ten years, except for a brief period around 1980. 1/ During the period 1975-84, the world market price of refined copper averaged $1,747/ton (or 79¢/lb) in constant 1984 dollars as compared to $3,593/ton (or 163¢/lb) in the period 1964-74. 2/ Indeed, the level of real prices over the last four years (1982- 85) has been perhaps the lowest in the last half century. 3/ Although world consumption of refined copper has recovered significantly since late 1983, copper prices have remained at low levels. Many copper-producing companies have accumulated considerable losses and substantial overcapacity persists. 2. Although exports of manufactures from developing countries have expanded rapidly over the last two decades, developing countries still rely heavily on primary commodities for their export revenues as well as for fiscal revenues. In 1982, primary exports accounted for 41% of non-fuel merchandise exports of developing countries. 3. Copper ranks among the top four non-fuel commodities in terms of contribution to the export earnings of developing countries. In 1983, export earnings from copper amounted to US$5.2 billion for all developing countries out of the total of US$7.7 billon from all industrial and developing countries. For several of the developing countries, copper is a vital export commodity, e.g., in 1980-82, the share of copper in total exports was as high as 92% in Zambia, 46% in Chile, 38% in Zaire and 30% in Papua New Guinea. 4. The recent sharp decline in copper prices, coinciding with the 1981- 82 recession, has been devastating for several major copper-exporting developing countries. From 1980 to 1982, for example, the total annual copper export earnings of the six major copper-exporting developing countries--i.e., Zambia, Zaire, Chile, Papua New Guinea, Peru and the Philippines--decreased from $5.8 billion to $4.3 billion, (or by 25.5%), despite the 6.6% increase in the annual export quantity, (from 2.92 to 3.11 million tons, copper content). 5. Broadly speaking, the ten-year-long stagnation in the world copper market has had both demand and supply dimensions. The growth of copper demand in the last decade has lagged well behind the historical trends established in the preceding two and a half decades (1.3% per annum in 1973-84 vs. 4.8% per annum in 1950-73). Two reasons for this stagnation are often cited. First, world economic growth has slowed markedly since 1975. Second, wide-spread 1/ The tonnage figures in this study refer to metric units unless otherwise noted. 2/ Prices deflated by the World Bank's manufacturing unit value (MUV) Index. 3/ The historical level of real prices varies depending on which deflator is used. - 2 - structural changes in copper-using industrial sectors have occurred, including significant inroads made by substitutes and material-saving technological changes. 6. Despite the very sluggish demand growth, world production capacity expanded during the 1974-83 decade resulting in an increasing imbalance between demand and supply potential. The oversupply conditions and low prices of the past decade have in recent years caused copper producers to direct management attention away from new projects and towards improvements in production efficiency for existing mines. As a result, average production costs have been declining. 7. At the same time a major shift has occurred in the regional distribution of production. Cutbacks in production have taken place mainly in North America where the higher-cost producers are located. In contrast, the majority of developing country producers (both in private and public sectors) have continued to produce copper at their maximum effective capacity and indeed some countries (notably Chile) have increased capacity and production. This uneven distribution of output reductions led to efforts by the copper industry in the United States to limit imports. While the United States has for a long time been the world's largest producer of copper, it has also been the largest consumer and, at least since the end of World War II, a net importer of copper. In 1978 and 1984, the US copper industry filed petitions with the International Trade Commission (ITC) seeking government actions to restrict copper imports. In both cases, ITC found substantial injury to the domestic industry from imports but the President decided not to take measures to restrict imports. Numerous bills have also been introduced in the US Congress to restrict copper imports; no legislative actions have materialized so far. 8. Still, many new copper deposits have been identified and explored, and some seem to be ready for development when the market outlook improves. In addition, plans exist to expand production at existing mines. A general view of the industry suggests that a plentiful supply of copper will be available in spite of little increase in copper prices for years to come. A majority of industry observers believe that, while growth in copper demand will continue to be sluggish, growth in new capacity is likely because of the availability of attractive, low-cost deposits and/or the expectation in some quarters of higher copper prices. 9. Because of the importance of the copper industry for developing countries, this study examines the nature of recent structural changes in the world copper industry and their implications for the market outlook in both the medium and the long term. More specifically, the study re-examines the demand relationships, focussing on end-uses and substitution possibilities as well as alternative scenarios for world economic development. The analysis of future supplies is based on available information on likely cost-of-production profiles of potential mines and metallurgical facilities as well as production costs of existing and temporarily-closed mines and plants. This paper examines the changing structure of primary supply, in particular the emergence of new suppliers, and the implications for changes in the degree of market -3- power over the past 10-15 years. Particular attention is given to the production strategy of different types of producers and the rationality of their decisions to reduce, maintain or expand production in the light of (a) changes in the industry's cost structure, and (b) changes in the ownership pattern of copper producers. 10. The insights developed about demand and supply behavior are incorporated in an econometric model of the world copper market. The model has been used in projecting consumption, production and prices under alternative scenarios with respect to global macro-economic developments, alternative investment scenarios and technological changes in the copper consuming industries. 11. The report is organized as follows. Chapter II reviews the long-term trends in the world copper industry to provide a historical perspective for understanding the so-called "structural" changes in the last decade and for assessing the future prospects for the industry. A short review of past attempts to "manage" supply by the industry is included in an annex. 12. Chapter III focuses on the world economic environment and primary commodity markets since the early 1970s. It provides a review of radical changes that have taken place in the global economic setting and that have had a profound impact on the copper industry in the last 15 years. The trends in other primary commodity markets are outlined to provide a comparative perspective for the trends in the copper market. 13. The next four chapters are devoted to analyses of some key elements of the structural changes in the copper market in the last decade. Chapter IV discusses the main factors behind the marked slowdown in copper demand growth over the last decade. An analysis of the persistent overcapacity conditions follows. 14. Chapter V discusses the recent changes in the competitive positions of major copper producing countries through an analysis of changes in the cost structure. 15. Chapter VI reviews the changes in owership/management control and organization of the industry in the last two decades, and analyzes the impact of these changes on the investment behavior of the industry and producers' responses to the overcapacity conditions. The impact of increased state participation, increased involvement of petroleum companies and the creation of CIPEC is assessed. 16. Chapter VII focuses on the changes in trade patterns and pricing mechanisms, in particular, the development of large smelting industries in Japan and Europe in the 1960s and 1970s and associated mine developments in the Pacific Basin, Latin America and Canada. The chapter also examines pricing mechanisms and price trends including the impact of fluctuations in the exchange value of the US dollar on the dollar price of copper. 17. Chapter VIII presents the analysis of the medium to long-term outlook for the copper market. A description of the methodology and assumptions used is followed by a presentation of the demand outlook, an assessment of the supply potential, and likely cost and price trends. The likely effects of optical fiber technology and lower-than-expected world economic growth on demand are discussed. Also, possible effects of capacity growth lower than expected on the market developments are discussed. The structure of the World Copper Market model used is summarized in an annex. -5- II. LONG-TERM HISTORICAL PERSPECTIVE A. Consumption Trends, 1890-1983 18. Annual world copper consumption over the period 1890-1983 is shown on a logarithmic scale in Figure 2.1. 1/ The period can be separated into the following four sub-periods, namely: (a) 1890-1918, (b) 1919-49, (c) 1950-73, and (d) 1974-84. Average annual growth rates for these sub-periods are summarized below in Table 2.1. Although this categorization of the sub- periods is somewhat arbitrary it helps in identifying certain broad changes in the trends in world copper consumption. 19. The first sub-period is a period when copper consumption expanded rapidly, basically sustained by rapid industrial growth in the United States, Japan and Germany, including, in particular, the expanding use of electric power and telephones/telegraphs. 2/ World industrial production expanded at an average rate of 5.8% per annum in this period. The second sub-period, from the end of World War I to the end of World War II was a period of low growth of copper consumption. In this sub-period, copper consumption experienced unusually sharp falls twice, around 1919-21 and 1930-32, when the world economy suffered severe economic depressions. World industrial production grew at 3.3% per annum in this period (vs. 5.8% in the preceding period). 3/ In the third sub-period there was sustained high growth in copper consumption, despite substantial substitution by aluminum. This high growth in consumption 1/ A consistent time series for world copper consumption going back a century is not readily available. However, Metallgesellschaft AG, Frankfurt, F.R. Germany has been publishing statistics on the copper industry since well before World War I. A time series for annual world copper consumption covering the period 1890-1983 has been constructed by using the Metallgesellschaft data, supplemented by estimates based on data available from the World Bureau of Metal Statistics (WBMS). In constructing the series, special care has been taken to maintain the consistency of the series over time. Nevertheless, it must be emphasized that the series is suitable only for the purpose of understanding the broad trends in consumption growth. 2/ Technological landmarks involving use of electricity (hence use of copper) were: the invention of the telegraph by Samuel F.B. Morse in 1840; the establishment of the first international cable between Dover, England and Calais, France in 1850; the demonstration of the principle of the electric speaking telephone by Alexander Graham Bell in 1875; and the registration of the patent on the first incandescent lamp by Thomas Alva Edison in 1879. The electrical industry began to receive considerable public attention around 1885. 3/ Copper consumption growth during World War II was slow because of slow growth in supply under war conditions. * SOIFSJIVS Te8aN Jo nvanlg PTTOM !'DV IJeDSTTasaBlTTaNH :ainoS 0361 OLSI 061t S0 0*" 61 Of OM6 0161 001s 0m - J ~ £ I f - J~ 4 I - I - 001 .. . . . . . . . . .. . . . . . . . . .. . . . . . . . . .. . .. . . .. .. . . .. . . . . . . ... . . . . . . . . . . . . . . . . . : ' . ' '. :. .. . . . . . .. . . ... .. . ,.. ,, , '. '. ................. ....... ............ ... . .. .. . . . .. *~ . . . // .... ............... ............... .............. .............. ............... ............ .... ................ . ......... ...........................7 .............. ............ .............. ..................... ................ ................ . ........................ .............................. .... , ,,,, . ,,,,,,,,,,, ..... jt|>.. .. . ...... ............... ............... ............... ......;....... .............. ............ .............. ............ .. oo .... ........ .......... ........... ........... ....... ...... . .... . e. * ' . ; ;? & e 8*. , . - ------- --.- - S 5 . ...............5..... .....5.......... I........................I..........5..........2.................. ............ *00001 Mat4 ONVSfOHL NOLLdvInSNO3 83dO3 AWY4I1Id G180M :JOzM z 'T - 7 - Table 2.1: TREND GROWTH RATES IN WORLD COPPER CONSUMPTION AND INDUSTRIAL PRODUCTION AND APPARENT INDUSTRIAL PRODUCTION ELASTICITIES, 1890-1984 Apparent Elasticity of Copper Consumption Trend Growth Rate with Respect to Copper Industrial Industrial Production Sub-Period Consumption Production Short-Term /a Long-Term /b ---Percent per annum /c-- (a) 1890 to 1918 5.8% 5.8 1.4 1.00 (b) 1918 to 1950 2.6% /d 3.3 1.0 0.79 (c) 1950 to 1973 4.8% 5.7 0.8 0.84 (d) 1973 to 1984 1.3% 2.8 0.7 0.46 /a The values of the apparent elasticity for each sub-period were calculated by OLS regressions using natural log formulae, disregarding possible effects of price fluctuations or serial correlation problems. Crudeness of the data does not warrant further analysis. /b The trend growth rate for copper consumption divided by the trend growth rate in industrial production. /c Based on least squares trend. /d Excluding the extreme "bottoms" of 1921 and 1932. Source: Metallgesellschaft AG.; World Bureau of Metal Statistics; World Bank, EPDCS - 8 - was driven by high growth in world industrial production (5.7X per annum), during two decades of unprecedented world economic growth in the 1950s and 1960s. Since 1974 (the fourth sub-period), copper consumption has been growing very slowly, partly because of low growth in world economy, and partly because of declining copper intensity with respect to industrial production. Simple regression analysis indicates that the apparent elasticity of world copper demand with respect to world industrial production has tended to decline over the whole of the period under review (Table 2.1). 1/ A more detailed analysis of the trends in copper consumption and the world economy in the post-World-War-II period will be presented in the chapters to follow. B. Copper Production Trends, 1890-1983 20. Figure 2.2 presents (in log scale) the broad trends in world copper mine production (that is, excluding copper recovered from scrap). 2/ The pattern of growth in mine production in the last 90 years or so (shown in Table 2.2) has been consistent with the pattern of growth in consumption largely because overproduction or underproduction cannot be sustained for an extended period. 3/ A more detailed discussion of production trends in the past 25 years is contained in Chapters VI and VII. C. Long-Term Price Trends, 1900-84 21. Figure 2.3 shows the London copper price during the period 1900-84, expressed in US dollars and deflated by the US Wholesale Price Index (base year = 1983). Five distinct sub-periods can be identified in terms of underlying average price levels; namely, (a) 1900-18; (b) 1919-46; (c) 1947- 63; (d) 1964-74; and (e) 1975-84. As might be expected, the sub-periods as defined in terms of average "real" price levels coincide broadly with the sub- periods demarcated by changes in consumption trends. 1/ Apparent "long-term" elasticity of demand, however, appears to have risen somewhat, or at least ceased to decline, in the 1950-73 period. One possible explanation for this "anomaly" could be that the average price of copper in the first half of this period remained relatively low (see Section C below). 2/ The production series relies on the Metallgesellschaft AG publications for the period up to 1950 and is linked at 1950 to the mine production series available from the World Bureau of Metal Statistics (WBMS). It is believed that the data so linked represent the trend throughout the period fairly well. 3/ Unfortunately, it is not possible to identify clearly sustained periods of mine capacity underutilization, because detailed data on capacity utilization on a world-wide basis are not available before the early 1960s. An analysis of capacity utilization changes is provided in chapter IV for the period since the early 1960s. -SDT3S9Tes Tulel Jo nvaing PTIOM !'-9V evDSTT9992TIIaW :sailnoS 0361 0t61 0961 OSII 0o"1 Of6 OZSI 0161 0061 0691 -...... - .- : - ........... *......... ....... . ........ ............ . ...... .......... ......... ........ ....... ....... ..... .... ....... ... .. .. ..... ....... .. ~* : * * * ............. ........... ............ ........... ......... ......... ................ .. .... ,,,,..........''' ........... ........ .......... ....... .... ........ Al-00001 'N. * N S , .,I ........... ................ .. :. . . ' .---- --------------- ..............-''''-'--- I .t ......................................................... ... ........ .. . .. ... . . .. .... . ... .. .. ... ..... .... .... .... .... ... .. .... ............ .. ...... . ................ .............-3. ..-. . . . . .. . . . . . . . . . . . ..... .. . . . . . . .... . . . . . .. .. .. . .... .. .. . .. .. .. . .. .. .. . .. .. .. . SNo1 aNvsnoHL NoLLonGOdd 3NIWI 83 O 0180M :z z aain8TA Figure 2.3: LONDON COPPER PRICE (1983$) 220 - 210 - 200 - 190 - 180- 170- 160- 150- 144.5,, 0 140- m 130- 120- L~110 - 100 - X oX Al 19 99.6¢e t 90 - 72 80- mV87.2 70 - 60- 60 - .. ... ..... .. ......... ..... ......... .................. 1900 1910 1920 1930 1940 1950 1960 1970 1980 YEAR Source: Metallgesellschaft AG.; World Bureau of Metal Statistics; American Bureau of Metal Statistics; United States Bureau of the Census. 22. Table 2.3 below shows the average "real" price levels for the sub- periods chosen. There were two sub-periods in which the "real" price averaged relatively high levels.: 1900-18 and 1964-74. In both of these sub-periods consumption growth was also generally high, i.e., 5.8% and 4.8% per annum, respectively. In contrast, in two of the other three subperiods, i.e., 1919- 46 and 1975-84, average prices were at relatively low levels. In these sub- periods, consumption growth was relatively slow. The period 1947-63 stands out as an exception. In this period, consumption growth was relatively rapid, but the price responded only moderately, perhaps because supply kept up extremely well with consumption growth. This is analyzed in greater detail in Section F below where developments on the supply side during this sub-period are examined. D. Highlights in the Period 1900-29 23. As seen earlier, in the period 1900-18, demand for copper increased rapidly and the price averaged high (145 USC per pound in 1983 dollars). In contrast, in the period 1919-29, consumption growth slowed (Figure 2.1) and the price in real terms averaged low (Figure 2.3). 24. The 1919-29 period is characterized as the "years of overcapacity" by Navin. 1/ Consumption of copper, which had grown considerably during World War I because of consumption for munitions, shrank suddenly after the war; indeed, primary consumption declined almost 60% from the peak of 1.48 million tons in 1918 to 0.62 million tons in 1921. Copper prices decreased sharply in 1919, and substantial inventories (of both raw and fabricated copper) piled up. After a 20-year experience of quick recoveries from unexpected slackened demand, companies were unwilling to cut back production immediately. Primary consumption did not return to the 1918 peak level until 1926. 2/ Modestly successful attempts were made to cartelize the market through restrictions on supply. 3/ 25. In addition to slow growth in demand and the chronic overcapacity conditions, an important factor that tended to help depress copper prices in the 1919-29 period was the declining trend in cost of production. Despite the steady decline in the average grade of ore in the United States (the largest and dominant producing country), world average production cost appears to have declined in the 1920s for two reasons; (a) the dramatic productivity improvements resulting from innovations in copper production technology, and (b) the emergence of new, low-cost producers, that resulted in a sharp reduction of the share of the dominant producer, the United States. 1/ Thomas R. Navin, Copper Mining and Management (University of Arizona Press, 1978), p. 125. 2/ The year 1926 was also a peak in the world economy before the crash of 1929-30. 3/ For a brief review of copper cartels before World War II, see Annex A. - 12 - Table 2.2: TREND GROWTH RATES IN WORLD COPPER MINE PRODUCTION (COPPER CONTENT) IN FOUR SUB-PERIODS IN 1890-1984 Average Growth Sub-Period Rates (Z per annum) Characterization (a) 1890-92 to 1916-18 6.3 high growth (b) 1916-18 to 1947-49 1.6 low growth (c) 1947-49 to 1972-74 4.8 high growth (d) 1972-74 to 1982-84 1.1 low growth Source: Metallgesellschaft AG.; World Bureau of Metal Statistics; World Bank, EPDCS. Table 2.3: AVERAGE REFINED COPPER PRICE IN LONDON, DEFLATED BY US WHOLESALE PRICE INDEX (1983 = 100), COVERING PERIODS 1900-1984 (US $ per pounds, in 1983 dollars) Sub-Period Average Price (f/lb) (a) 1900-1918 144.5 (b) 1919-1946 75.7 (c) 1947-1963 99.6 (d) 1964-1974 167.0 (e) 1975-1984 89.6 Source: Metallgesellschaft AG.; Metal Bulletin. - 13 - 26. The average grade of copper ore mined has declined through the history of copper mining, but because of technological progress in mining and metallurgical processing, the cost of producing copper has tended to decrease over the long term. Lowell cites the 15% Cu ore mined by Indian pits in the Lake Superior district (of what now is the United States), before 3000 B.C. 1/ The copper mined in northern Europe in the year 1540 averaged about 8% Cu, and the cost is estimated to have been about US$30 per lb (in 1983 terms). By 1890, the average ore grade had fallen to about 6% Cu while the cost of copper had decreased to US$1.25 per lb. (in 1983 terms). 27. Reliable data on the long-term trend in the grade of copper ore mined for the world as a whole are not available. However, for the United States, data on the average "yield" of copper ore mined are available from the US Bureau of Mines (USBM) for the period back to 1906 (Figure 2.4). 2/ Over the 77-year period the yield declined steadily--from 2.23% in 1906-08 to 0.52% in 1981-83. 3/ Because of the narrowing gap between the yield and the ore grade over time resulting from technological improvements 4/, it is safe to infer that the ore grade in the United States also declined over the same period. 28. The decline in the yield of copper ore mined in the United States was offset by improvements in productivity. While many innovations contributing toward productivity increases were adopted by the copper mining industry during the period, the most outstanding was the advent of open-pit mining. Daniel C. Jackling of Salt Lake City first applied the method (which had already been well-developed in iron ore mining) to copper mining at Kennecott's Bingham Canyon mine in 1906. The technique was also applied at the great Chuquicamata mine in Chile in 1915, and at other new mines in the l/ J. David Lowell, "Copper Resources in 1970," the 1970 Jackling Award Lecture, The Transactions of the Society of Mining Engineers, AIME, June 1970, pp. 97-103. 2/ Unpublished data made available to the World Bank by Janice L.W. Jolly, USBM. The "yield" percent of recoverable copper from ore mined, as reported by USBM, is calculated to the smelter level, and is not the same as the average ore grade, reserve grade or "mill-head" grade. 3/ The only periods in which the declining trend of the yield was temporarily reversed were 1919-22 and 1932-35; these were the periods in which copper markets were extremely depressed. Many high-cost mines (which tended to be the relatively low-yield mines) were shut down while the mines which remained in operation resorted to "high-grading". 4/ Before the flotation process was introduced in the milling stage, the recovery rate was only 70-75%. Flotation helped increase the recovery rate. Today, the recovery rate is in the order of 90%. Thus, because of the improvements in recovery rates, the gap between the grade of ore mined, and the "yield" has been narrowing over time. Figure 2.4: AVERAGE YIELD OF COPPER MINED IN THE UNITED STATES 2.5- 2.4- 2.3 - 2.2 - 2.1 - - --\ - . . - 2- 1.9- 1.8- 1.7- 1.6- 1.5 - 1.4 - 1.3 - 1.2 - 1.1 - _ _ _ _ _ 1- 0.9- 0.8- 0.7- 0.6- 0.5- 1906 1916 1926 1936 1946 1956 1966 1976 YEAR Source: U.S. Bureau of Mines. - 15 - United States. 1/ Another new technology that had a considerable impact on production costs was the froth flotation process of concentration, which also had the effect of facilitating open-pit mining. 2/ This method was introduced to the US copper industry in 1915 at an Anaconda affiliate in Arizona. "Flotation was capable of recovering up to 50 percent more ore than had been recovered by previous methods; hence, it was the equivalent of increasing the sulfide ore-mining capacity of the country, in a very short period, by half again what it had been before--all of which contributed significantly to the overcapacity that developed after World War I." 3/ 29. Data on productivity changes in copper production for the early decades of the 20th Century are scarce. However, data given in Table 2.4 indicate that labor productivity (as measured by the volume of copper produced per man-hour worked) increased by 5.2% per annum in the United States in the period 1912-25. 30. The adverse effects of the declining grade of ore in the United States on the world average production cost were also mitigated by the increasing exploitation of newly discovered rich deposits outside the United States. While the United States has been the historically dominant producers, the share of the United States in world mine production declined from 57% in 1920 to 47% in 1929, mainly as a result of rapid expansions in production in Chile, Zaire and Canada (Table 2.5A). The combined shares of Chile, Zaire and Canada increased from 16% in 1920 to 29% in 1929 (Table 2.5B) E. Highlights in the Period 1929-50 31. The world copper industry was hard hit during the depression of the 1930s. Sharp price declines failed to stimulate demand. Production declined from close to two million tons in 1929 to less than 0.9 million tons in 1932. World production did not surpass the 1929 level until 1937. Despite the extreme stagnation and slow recovery in the copper market, production in some areas expanded significantly. By 1938, Northern Rhodesia (Zambia of 1/ Thomas R. Navin, Copper Mining and Management (University of Arizona Press, 1978), pp. 31-34. To quote Sir Ronald Prain, "Jackling's important contribution was simply the realization that in a large-tonnage operation fixed unit costs could be reduced to a minimum which would more than compensate for the bigger tonnages of material that had to be mined and for the larger plants necessary to treat the ore and produce the finished product." Sir Ronald Prain, Copper: The Anatomy of an Industry, (Mining Journal Books Ltd., London, 1975), p. 14. 2/ To quote Sir Ronald Prain again: "The story of the discovery of this process is always the same--of a lead miner's wife washing her husband's overalls and noticing that particles of galena, a lead sulphide mineral, always stuck to the soap bubbles..." Ibid., p. 16. 3/ Thomas R. Navin, Op. cit. pp. 46-47. - 16 - Table 2.4: PRODUCTIVITY OF LABOR IN COPPER MINING IN THE UNITED STATES, 1912-45 Pounds of Copper Year/Period Per Man-Hour 1912 9.86 1919-21 13.18 1928-30 21.68 1934-36 33.87 1939-41 28.48 1944 31.95 Source: U.S. Federal Trade Commission, Report on the Copper Industry, Part I, "The Copper Industry of the United States and International Copper Cartels," U.S. Government Printing Office, 1947, p. 115, Table 23. Table 2.5 (A): COPPER MINE PRODUCTION BY COUNTRY/REGION, 1900-1984 ('000 tons (cu content) Country/Region 1900 1920 1929 1938 1950 1961 1964-66 1969-71 1974-76 1979-81 1982-84 United States 275 555 905 506 825 1,057 1,218 1,447 1,395 1,388 1,088 Canada 9 37 113 259 240 398 454 595 774 681 648 North America 284 592 1,018 765 1,065 1,455 1,672 2,042 2,169 2,069 1,736 Australia 23 27 13 20 15 97 103 155 230 238 252 Other Industrial 60 110 196 181 107 191 192 238 253 224 234 INDUSTRIAL TOTAL 367 729 1,227 966 1,187 1,744 1,967 2,435 2,641 2,530 2,222 Zambia 0 3 7 254 298 575 651 685 695 591 557 Zaire 0 19 137 124 176 295 294 386 480 455 502 Other Africa 7 5 22 23 47 115 158 210 299 302 329 AFRICA TOTAL 7 27 166 400 520 985 1,103 1,281 1,474 1,348 1,388 Chile 26 99 321 352 363 547 615 694 912 1,071 1,263 Peru 8 35 56 38 30 198 179 204 204 364 348 Mexico 22 51 87 42 62 49 55 64 83 171 210 Other LAC 2 20 22 16 26 19 26 22 16 13 37 LAC TOTAL 59 204 485 447 481 814 874 984 1,215 1,619 1,858 Philippines 0 0 0 4 10 52 66 163 230 302 266 Papua New Guinea 0 0 0 0 0 0 0 0 178 161 173 Other Developing Asia 0 0 7 9 7 83 103 135 261 340 511 ASIA & OCEANIA TOTAL 0 0 7 13 17 135 169 298 669 803 950 Southern Europe 54 3 27 81 81 109 130 150 164 148 165 DEVELOPING TOTAL 120 234 684 941 1,100 2,043 2,275 2,712 3,521 3,916 4,361 INDUSTRIAL & DEVELOPING 487 963 1,911 1,907 2,286 3,787 4,242 5,148 6,161 6,447 6,583 East European 7 4 37 103 238 610 830 1,095 1,457 1,583 1,705 Non Market Economies WORLD TOTAL 494 967 1,948 2,010 2,524 4,397 5,072 6,243 7,618 8,030 8,288 Source: Metallgesellschaft AG.; World Btureau of Metal Statistics. Table 2.5 (B): COPPER MINE PRODUCTION BY COUNTRY/REGION, 1900-1984 (percentage share in world total) Country/Region 1900 1920 1929 1938 1950 1961 1964-66 1969-71 1974-76 1979-81 1982-84 United States 55.7 57.4 46.5 25.2 32.7 24.0 24.0 32.2 18.3 17.3 13.1 Canada 1.7 3.8 5.8 12.9 9.5 9.1 9.0 9.5 10.2 8.5 7.8 North America 57.4 61.2 52.2 38.1 42.2 33.1 33.0 32.7 28.5 25.8 20.9 Australia 4.7 2.8 0.7 1.0 0.6 2.2 2.0 2.5 3.0 3.0 3.0 Other Industrial 12.2 11.4 10.1 9.0 4.2 4.3 3.8 3.8 3.3 2.8 2.8 INDUSTRIAL TOTAL 74.3 75.4 63.0 48.1 47.0 39.7 38.8 39.0 34.7 31.5 26.8 Zambia 0.0 0.3 0.3 12.6 11.8 13.1 12.8 11.0 9.1 7.4 6.7 Zaire 0.0 2.0 7.0 6.2 7.0 6.7 5.8 6.2 6.3 5.7 6.1 Other Africa 1.4 0.6 1.1 1.1 1.9 2.6 3.1 3.4 3.9 3.8 4.0 AFRICA TOTAL 1.4 2.8 8.5 19.9 20.6 22.4 21.7 20.5 19.3 16.8 16.7 Chile 5.3 10.2 16.5 17.5 14.4 12.4 12.1 11.1 12.0 13.3 15.2 Peru 1.7 3.6 2.9 1.9 1.2 4.5 3.5 3.3 2.7 4.5 4.2 x Mexico 4.5 5.2 4.4 2.1 2.4 1.1 1.1 1.0 1.1 2.1 2.5 1 Other LAC 0.4 2.1 1.1 0.8 1.0 0.4 0.5 0.4 0.2 0.2 0.4 LAC TOTAL 11.9 21.1 24.9 22.2 19.0 18.5 17.2 15.8 15.9 20.2 22.4 Philippines 0.0 0.0 0.0 0.2 0.4 1.2 1.3 2.6 3.0 3.8 3.2 Papua New Guinea 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.3 2.0 2.1 Other Developing Asia 0.0 0.0 0.4 0.4 0.3 1.9 2.0 2.2 3.4 4.2 6.2 ASIA & OCEANIA TOTAL 0.0 0.0 0.4 0.6 0.7 3.1 3.3 4.8 8.8 10.0 11.5 Southern Europe 10.9 0.3 1.4 4.0 3.2 2.5 2.6 2.4 2.2 1.8 2.0 DEVELOPING TOTAL 24.2 24.2 35.1 46.8 43.6 46.5 44.9 43.4 46.2 48.8 52.6 INDUSTRIAL & DEVELOPING 98.6 99.6 98.1 94.9 90.6 86.1 83.6 82.5 80.9 80.3 79.4 East European 1.4 0.4 1.9 5.1 9.4 13.9 16.4 17.5 19.1 19.7 20.6 Nonmarket Economies WORLD TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Source: Metallgesellschaft AG.; World Bureau of Metal Statistics. - 19 - today) had become a major producer (producing over 250,000 tons), and major production expansions occurred in Canada, Southern Europe and the USSR. US production, in contrast, remained at only 56% of its 1929 level, with its share in world production having fallen to 25%. 32. During World War II governments of the major producing countries froze copper prices at low levels and imposed stiff excess profit taxes. The industry also suffered from a manpower shortage. Even after World War II, governmental control of the copper market in major consuming countries continued for several years. In the United States, although price controls were lifted as early as November 1946, exports remained subject to loose control. With the advent of the Korean War in the fall of 1950, the U.S. Government re-established various controls to ensure the adequacy of copper supply for defense requirements; these remained in effect until February 1953. 1/ The trend of the United States losing its share in world production was reversed in the 1938-50 period primarily because of the encouragement and support provided by the US Government 2/ (Table 2.5B). The British Covernment sanctioned resumption of trading on the London Metal Exchange in August 1953, re-establishing the world's free market for copper. F. Highlights of the Period 1950-63 33. The price of copper rose significantly in the first half of the 1950s because of the rapid rise in consumption stimulated by the Korean War and the stockpiling of copper (along with other strategic raw materials) by the US Government. Consumption continued to grow at a fast pace throughout the 1950s and 1960s. Nevertheless, the real price remained at relatively low levels largely because of the rapid expansion of production (partly due to the incentives provided by US Government programs). 34. Thus, from the end of World War II through the 1950s, the US Government stimulated domestic copper production through various measures. The declared objective was to ensure an adequate level of national defense preparedness in terms of industrial mobilization and raw material supply. The Critical Materials Stockpiling Act of 1946 set the target stock level for copper at 1.13 million tons. The target was increased to 1.9 million tons in 1950, then reduced to 1.0 million tons in 1952. In 1954, two types of objectives (targets) were established. Basic objectives were set at 1.45 million tons and maximum objectives at 3.17 million tons. In 1959, the basic and maximum objectives were reduced to zero and 0.91 million tons, respectively. The actual level of US Government stockpile of copper reached 543,000 tons in 1951, and increased to 1.04 million tons (including 1/ Orris C. Herfindahl, Copper Costs and Prices: 1890-1957 (The Johns Hopkins Press, Baltimore, 1959), pp. 127-136; and A.D. McMahon, Copper: A Materials Survey (USBM, IC 8225, 1965), pp. 330-336. 2/ For details, see the next section. - 20 - supplementary stockpiles) in 1960. Accumulated stocks in the government stockpiles substantially exceeded the maximum target as of 1961. 35. The Defense Production Act of 1950 "provided for expansion of copper output at mines then operating or idle or for maintenance of production that might be lost without such aid." 1/ The properties were to receive government loans, government purchase contracts, or tax amortization benefits, or combinations of the three types of assistance. The contracts offered by the government under this Act, in effect, provided floor prices for the projects involved through the guaranteed prices. Under the program some 55 projects were undertaken in the period 1950-56. In the period 1951-53, the guaranteed prices negotiated under these contracts were close to the quoted market prices as shown in Table 2.6, and "were adequate, together with the other features of the contract, to induce the agreed amount of investment." 2/ "The properties in question were generally viewed, however, as 'marginal r in the sense that it was doubtful whether in the normal course of events over the next decade or so they could have earned a satisfactory return on the additional investment required for their development and production." 3/ 36. During the period 1950-63, copper companies generally enjoyed favorable returns on investment. 4/ However, after the boom years of 1955-56, the copper market experienced a period of oversupply until 1963. This was mainly because of an excessively rapid expansion in capacity; market economies' mine capacity rose from 2.88 million tons per year (tpy) at the beginning of 1956 to 3.98 million tpy at the beginning of 1961, or at 6.7% per annum, compared with consumption growth of 3.4% per annum in the 1955-62 period. 5/ Consequently, copper prices remained relatively low and the return on investment was reduced. For example, the return for US copper producers in the period 1957-63 was reduced, compared with the period 1950-56 (Table 2.7). In the oversupply period of 1957-63, production cuts were implemented by major producers around the world. 37. The price of copper began to fall in the spring of 1956 and continued falling through 1957-58. In 1958, mine production was cut. The market balance was temporarily recovered in 1959 when consumption rose substantially 1/ McMahon, op. cit., p. 332. 2/ Herfindahl, op. cit., p. 143. 3/ Herfindahl, Op. cit., p. 144. 4/ For example, Table 2.7 shows the average return on investment during this period for the six major US copper companies. 5/ "1967 Study by Andean Region of Economic Affairs Department of Pan American Union", Engineering and Mining Journal, February 1967. - 21 - Table 2.6: PRICES IN DEFENSE PRODUCTION ACT CONTRACTS Average US Domestic Contract Coming Tonnage Commitment Price, f.o.b. into Effect in: (m.t.) Price (¢/lb) Refinery (¢/lb) 1951 378,200 23.8 24.4 1952 727,400 24.5 24.4 1953 136,100 28.0 28.9 Total 1951-1953 1,241,700 24.7 25.9 Source: Orris C. Herfindahl, op. cit., p. 143. Table 2.7: AVERAGE RETURN ON INVESTMENT, MAJOR US COPPER COMPANIES 1950-56 AND 1957-63 Company 1950-56 1957-63 Phelps Dodge 19.0 9.6 Newmont 17.4 12.1 AMAX 16.7 15.0 /a Kennecott 16.6 8.4 ASARCO 13.6 6.9 Anaconda 8.0 5.0 /a Newmont and AMAX were much more diversified companies and contributions of copper toward their revenues were only a part of their total revenues. Source: Company annual reports; cited from Thomas R. Navin, Op. cit., pp. 155-159. - 22 - and a long strike in the US copper industry 1/ contained production increases. During 1959-60, however, almost half a million tons of new mining capacity came onstream. Thus, in 1960, even though consumption rose by almost 9%, world production rose by 17% and the LME price declined sharply. 38. The sharp fall in the price led major producers (see above) to take concerted action to reduce production. In October 1960, a 10% cutback of planned production was announced by the major producers operating in Zambia, Zaire, Canada and the United States and was continued in 1961 despite a further 7% increase in consumption. In January 1961, Southern Peru Copper Corporation and the Chilean mines of the US producers joined in. In July 1963, following a year in which there had been no increase in consumption, the producers announced to increase their cutback to 15%. The actual reduction in production, however, was not as great as the announced cutbacks would suggest; the mines indeed operated at 92-93 percent of existing capacity in the 1961-63 period. 2/ By the close of 1963, market balance was restored. 39. During the period 1960-63, in addition to concerted efforts to cut back production, major non-US copper producing companies supported the price of copper in London by a series of purchases. Consequently, the price of copper remained stable; the LME settlement price was confirmed in the narrow band of 220-235 pounds sterling per long ton from September 1960 to November 1963. The companies that cooperated officially in the support operation were RST, Anglo-American and Union Miniere. In addition to purchases at LME, they bought copper from smaller producers, who were not participating in the scheme. The amount of funds used for these purchases and the amount of copper stockpiled have never been disclosed. The company officials involved in this support operation reportedly consider the operation a great success. G. The High Price Period of 1964-73 40. The decade 1964-73 became a dynamic period for the copper industry, with consumption growing at 4.8% per annum and prices soaring. The high pace of consumption growth was influenced by high economic growth, and the impact of the escalation of the Vietnam conflict. On the supply side, although capacity expansion was rapid, it still tended to lag behind demand growth; moreover, a series of events (strikes, civil disturbances, etc.) prevented full utilization of installed capacity. The boom lasted until 1975, when it collapsed. 41. The year 1964 began with the LME price of copper at 30 cents a pound, and US producer price at 30.6C/lb. Consumption rose by 11% in 1964, as the world-wide communications explosion gathered momentum and the Vietnam conflict 1/ Production loss estimated at 300,000 tons. 2/ The cutbacks were measured on the basis of anticipated maximum production over the next 12 months. The cutbacks are estimated to have amounted to some 250,000 tons per year. - 23 - escalated. The industry was already operating at over 92% capacity, and with little additional mine capacity nearing completion, was in no position to meet the increase in demand. Stocks that had accumulated over the preceding three years were depleted quickly. By November 1964, the LME price reached 63.5C/lb. US producer price, however, remained low throughout 1964, rising only slightly to 33.7¢/lb. by November-December. 42. The rising trend in the LME price continued until it reached a peak of 86.4c/lb. in April 1966. Although it declined in the next few months, the price remained mostly in the 50-60¢/lb. range from August 1966 through the spring of 1969, when the price began to break away above 60C/lb. US producer price remained substantially below the LME price in the 1964-69 period, rising only gradually to an average of 47.5C/lb. in 1969 (as compared with 66.5¢/lb. at LME). 43. In the meantime, beginning in January 1964, the Zambian copper producers stopped selling on the basis of the LME price and began deliveries to their contract customers at non-US destinations at pre-announced producer prices disassociated from the LME. The other major producers soon followed. 1/ These prices were similar in level to US producer prices which were much lower than the LME price (Table 2.8). However, in the spring of 1966, Chilean and Zambian producers decided to give up their producer pricing and to sell their copper on the basis of the LME price. The other producers soon followed (except for US domestic production sold inside the United States). 2/ 44. In addition to the effects of rapid demand growth, the generally tight market conditions in the 1963-73 period were also influenced by a series of supply-related events. First, the market economies' mine capacity, which had grown at 6.7% per annum in the 1956-60 period, grew at only 2.4% per annum in the period 1961-65, a consequence of the slowdown in investment during the surplus market period. Second, political difficulties and strikes reduced production in one major producing country after another. Civil disturbances in Zaire reduced production by 25-30,000 tons in each of the three years 1963- 65 compared with the earlier annual peak. Zambian production fell 75,000 tons in 1966 and was 40,000 tons below the 1965 peak in 1967 because of transportation difficulties and fuel shortages caused by the unilateral declaration of independence in Zimbabwe (Rhodesia). Moreover, the political difficulties in Zaire and Zambia discouraged the expansion of capacity. Strikes cut Chilean production by 35,000 tons in 1965 and Australian production by 15-20,000 tons in 1964 and 1965. A 9-month industry-wide strike in the United States reduced US production by 450,000 tons in 1967 and by 300,000 tons in 1968. 1/ The only earlier attempt by non-US producers to institute a producer price unrelated to the LME price was the RST's attempt from 1956 to 1958 (the so-called RST price). 2/ Traditionally, major US producers have sold their copper at the so-called US producers price since well before World War II. - 24 - Table 2.8: PRICES OF COPPER, 1962-1969 (US ¢/lb.) Producer Price London Metal Year United States la Chile /b Zambia /c Exchange /d 1962 31.0 29.3 1963 31.0 29.3 1964 32.3 31.8 31.0 43.9 1965 35.4 36.0 35.2 58.5 1966 36.0 54.1 55.2 69.3 1967 38.0 51.0 1968 41.0 56.2 1969 47.5 66.5 /a United States domestic electrolytic, delivered Connecticut Valley. 7W Chile, electrolytic, delivered London. /c Zambia, electrolytic, delivered London. 7dT Spot, electrolytic, ex-warehouse London. Source: American Metal Market, Metals Week, Metal Bulletin. - 25 - 45. The shortage of primary copper supplies during this period was offset partly by increases in the supplies of secondary copper, but more significantly, by releases from the US Government stockpiles. Copper stocks in the US Strategic Stockpile declined from 1,035,600 tons at the end of 1963 to 249,600 tons in 1967. 46. In summary, the LME price of copper remained at extremely high levels throughout the 1964-73 period, because of the following factors: (a) Rapid growth in demand persisted, reflecting the sustained world economic boom and the escalation of the Vietnam conflict. (b) World mine capacity tended to lag behind the rapidly growing demand--a lagged effect of the overcapacity conditions in the 1957-63 period. (c) Significant production losses resulted from a series of unanticipated supply-side related events including political disturbances in Africa and labor strikes elsewhere. (d) Attempts by US and other producers to sell at producer prices which were substantially below the LME price had the effect of raising the LME price. - 26 - III. RECENT DEVELOPMENTS IN THE WORLD ECONOMY AND IN PRIMARY COMMODITY MARKETS 47. To provide a perspective on the economic environment from which the difficulties of the world copper industry arose over the last decade, this chapter reviews changes in the world economy and in the primary commodity markets since the early 1970s. The dramatic decline in world economic growth in the decade 1973-83 was accompanied by other significant marcro-economic changes. World trade growth slowed markedly. Growth in productivity in industrial countries also slowed significantly. Unemployment levels increased in industrial countries. Inflation, which began to creep up in the late 1960s, escalated during the 1970s. Following the quadrupling of oil prices in 1973-74, energy prices increased sharply, fuelling the already galloping inflation. The relatively stable international monetary system established in the 1950s and 1960s was converted to a system of fluctuating exchange rates among the major currencies (effectively "block floating", with many developing countries' currencies tied to the US dollar and the French franc). 48. Since the late 1970s, attempts by major industrial countries to control inflation resulted in very high interest rates, aggravating the debt- servicing problems of developing countries. These and other macro-economic developments since the early 1970s have deeply affected the world copper industry and primary commodity markets in general. A. Changing World Economic Environment 49. The most significant change in the world economy has been the sharp decline in economic growth (as traditionally measured) from that experienced in the preceding two decades (Table 3.1). The industrial market economies' real GDP growth rate declined from 4.9% per annum in the period 1960-73 to 2.8% per annum in the period 1973-80 and to 1.1% in the period 1980-83, although it rebounded to 4.5% in 1984 and 2.8% in 1985. The developing countries' real GDP growth rate experienced only a modest decline after 1973; from 6.1% per annum in the period 1960-73 to 5.5% per annum in the period 1973-80. This rate declined sharply to 2.5% in the period 1980-83, because the prolonged recession in the industrial market economies, historically high interest rates and more expensive oil exacerbated the external debt problems of many developing countries. 50. In retrospect it can be seen that the health of the world economy was already deteriorating before the end of the "golden" 1960s. The rising inflation and slowdown in productivity gains in the United States and other industrial countries are evidence of that deterioration. The annual growth rate of consumer prices in the United States, which had averaged 2.0% in 1960- 68, rose to 5.0% in 1968-73 (Table 3.2). For the OECD group the rate of price increase rose from 2.7% in 1960-68 to 5.6% in 1968-73. Real GDP per person employed (an indicator for labor productivity) in the United States grew at 2.6% a year in the period 1960-68, but slowed to an average growth of 1.3% per annum in 1968-73. For OECD as a whole, however, the growth rate in the same indicator declined only moderately, from 4.0% per annum in 1960-68 to - 27 - Table 3.1: REAL GDP GROWTH RATE, BY ECONOMIC REGION/COUNTRY, 1960-1984 (annual average percentage) Region/Country 1960-73 1973-80 1981 1982 1983 1984 All Developing Countries 6.1 5.5 2.6 2.2 2.6 3.9 Low Income 5.7 4.9 4.0 5.0 7.2 6.6 China 8.5 5.8 3.0 7.4 9.0 9.0 India 3.6 4.1 5.8 2.6 6.5 4.2 Africa 3.7 2.7 1.7 0.7 0.7 1.6 Zaire 5.6 -2.0 2.9 -3.0 1.3 2.7 Middle Income 6.3 5.7 3.0 0.8 0.0 3.1 Oil Exporters 6.1 5.8 4.6 0.9 -1.0 2.7 Indonesia 4.5 7.1 7.9 2.2 4.2 6.6 Mexico 6.7 6.0 7.9 -0.6 -5.3 3.5 Peru 5.1 1.9 3.0 0.9 -13.4 4.8 Venezuela 5.3 4.5 -0.3 0.7 -5.6 1.4 Oil Importers 6.3 5.6 2.0 0.8 0.7 3.3 Chile 3.7 3.7 5.5 -14.1 -0.7 6.3 Zambia 7.5 -0.2 4.7 -2.0 1.7 -1.3 Major Exporters of 6.8 5.9 1.6 1.3 0.7 3.5 Manufacturers Brazil 8.6 a/ 6.6 -1.6 0.9 -3.2 4.5 Philippines 5.1 6.2 3.8 2.9 1.1 -5.2 South Africa 5.3 2.9 4.9 -1.2 -3.2 4.7 Other 5.2 4.6 3.3 -0.7 0.5 2.7 High Income Exporters 8.6 8.3 0.1 -1.7 -7.0 0.6 Industrial Market Economies 4.9 2.8 1.4 -0.3 2.2 4.7 W. Europe 4.6 2.4 -0.3 0.5 1.0 2.5 United States 4.0 2.2 3.0 -2.3 3.6 6.7 Japan 10.3 3.8 4.2 3.0 3.0 5.8 Australia 5.3 2.5 3.9 0.7 0.4 6.3 Canada 5.4 3.0 3.3 -4.4 3.3 4.7 East European Nonmarket NA NA 1.5 2.1 3.0 2.8 Economies /a For the period 1963-73. Source: The World Bank, EPD. Table 3.2: GROWTH RATES OF CONSUMER PRICE INDEX, REAL GDP PER LABOR EMPLOYED AND WORLD TRADE VOLUME, SELECTED REGION/COUNTRY, SELECTED PERIODS IN 1960-1984 (percent per annum) 1960-68 1968-73 1973-79 1980 1981 1982 1983 1984 Consumer Price Index OECD - Total 2.9 5.6 10.0 12.8 10.5 7.8 5.2 5.3 Major 7 Industrial Countries 2.7 5.5 9.4 12.2 10.0 7.0 4.4 4.5 United States 2.0 5.0 8.5 13.5 10.4 6.1 3.2 4.3 Real GDP Per Person Employed OECD - Total 4.1 3.6 1.6 0.6 1.3 0.2 2.2 3.3 Major 7 Industrial Countries 4.0 3.4 1.5 0.3 1.4 0.1 2.1 3.3 United States 2.6 1.3 0.2 -0.8 1.2 -1.3 2.4 2.5 Trade Volume OECD Imports 8.6 10.2 3.7 -0.4 -2.5 -0.5 4.7 11.5 Market Economies Exports 7.7 /a 9.3 /b 4.6 0.0 0.0 -3.0 3.1 8.0 /a 1960-69. 7b 1969-73. Source: OECD, Historical Statistics, 1960-1983; Economics Outlook, December 1985; United Nations, Monthly Bulletin of Statistics. - 29 - 3.6% in 1968-73, because of an increase in the growth rate in Oceania and continued high growth in Western Europe and Japan. At the same time, direct and indirect economic consequences of the escalating Vietnam conflict began to show up in the balance of payments of the United States. The deficit in the basic balance of the United States rose from $366 million in 1964 to $3.8 billion in 1969, and to over $10 billion in 1971 and 1972. 51. In 1971, in the face of rising inflation, slowing productivity gains, growing balance of payments deficits and continued drain on its gold reserves, the United States decided to abandon the gold convertibility of the US dollar and in effect to devalue the dollar vis-a-vis other major currencies. Soon the international monetary system moved from the fixed exchange rate system operating within the IMF framework to a system of floating exchange rates, spelling the end of the Bretton Woods system. As US inflation and US balance of payments deficits continued to escalate, the value of the US dollar declined further: the US dollar to SDR rate, which was at 1.0 in 1970, rose to 1.214 by 1975 (Table 3.3). Though it increased somewhat in 1976-77, the value of the dollar vis-a-vis the SDR continued to decline to reach the ratio of over 1.3 by 1979. 52. After the recession of 1970-71, the world economy experienced an extraordinary boom with real GDP in OECD countries growing at 5.4 and 6.0 percent in 1972 and 1973 respectively. The boom, which was remarkably simultaneous in all the industrial market economies, was the culmination of the twenty-year period of sustained growth. However, the quadrupling of the petroleum price by the OPEC oil exporters in 1973 provided a sufficiently strong blow to push the already shaky world economic structure into a period of stagflation. The price of oil, which had remained remarkably stable in current dollars (i.e., declining in real terms) for two decades, began to rise in February 1971 when major exporters in the Middle East decided to index their posted price of oil to the inflation in industrial countries. The price thus rose from about $1.80/bbl. in January 1971 to $3.00 by October 1973, when war broke out between Israel and Egypt and the price was sharply raised to over $5.00 in the midst of the so-called Arab oil embargo. In December 1973, a historic decision by the OPEC led to a further increase of the posted price to over $11.00/bbl. 53. In 1974, the world economic boom collapsed and a deep recession set in through 1975. For the OECD as a whole, the growth of real CDP declined to 0.8% in 1974 and to a negative 0.2% in 1975. Despite the slowdown of world economic activity, inflationary pressures continued to surge; on the average, OECD consumer prices rose by 13.4% in 1974 and 11.3% in 1975. Although inflation in the OECD area was somewhat moderated in the period 1976-78 (CPI growth averaged 8.5% per annum), the pace was high enough to sustain the inflationary psychology around the world. The price of gold rose from $41 per oz. in 1971 to $193 by 1978. The value of the US dollar continued to depreciate. - 30 - Table 3.3: EXCHANGE RATES OF THE US DOLLAR, 1970-1984 Major US$ Currencies US$ per Year per SDR per US$ /a British Pound 1968 1.000 1.398 2.394 1969 1.000 1.401 2.390 1970 1.000 1.374 2.396 1971 1.003 1.336 2.444 1972 1.086 1.222 2.502 1973 1.192 1.103 2.452 1974 1.203 1.140 2.339 1975 1.214 1.116 2.222 1976 1.155 1.209 1.806 1977 1.168 1.183 1.746 1978 1.252 1.055 1.220 1979 1.292 1.009 2.122 1980 1.301 1.000 2.326 1981 1.179 1.177 2.028 1982 1.104 1.335 1.751 1983 1.069 1.445 1.517 1984 1.025 1.607 1.336 1985 1.015 1.671 1.296 /a An index of the value of major currencies relative to the US dollar (1980 - 100). The "basket" includes currencies of Canada, U.K., Germany, France, Italy and Japan. The average merchandise imports of these countries during 1979-1981 were used as weights for constructing the composite index. Source: IMF, International Financial Statistics. - 31 - 54. Most industrial market economies recovered sharply in 1976, with OECD countries' real GDP growing by 4.8%, and continued to grow at moderately high rates in the period 1977-78 (3.8 - 3.9% p.a.). Nonetheless, unemployment levels in these economies were not reduced significantly. The unemployment level in the seven major industrial market economies averaged 4.9% in the 1973-79 period, compared with 2.9 and 3.2 percent in the periods 1960-67 and 1968-73, respectively (Table 3.4). 55. In 1979 and 1980, petroleum prices increased sharply again (the so- called second oil shock). After a period of falling real prices, conditions in the petroleum market changed rapidly in the fourth quarter of 1978, with the drastic decline in exports from Iran. Production increases effected by other exporting countries were not sufficient to compensate fully for the loss of supply from Iran. The weighted average OPEC crude oil price rose from $12.90/bbl. in 1978 to $18.60 in 1979, $30.50 in 1980 and finally to $34.30 in 1981 (in current $). 56. As the governments of major industrial countries began to focus on the objective of controlling inflation in 1979, interest rates began to rise. The Eurodollar rate rose from 6.0% in 1977, and to 14.4% in 1980. 1/ In the United States, prime rates rose from 6.8% in 1977 to 18.87% in 1981. Real GDP growth in the industrial market economies slowed down to rates less than 2% in 1980 and 1981 and to a negative 0.3% in 1982. However, the rampant inflation of the 1970s was brought under control in the period 1981-83. Since 1983, the world economy has been on a recovery path, with industrial market economies' real GDP growing at 2.2% in 1983 and 4.7% in 1984. B. Economic Performance of Developing Countries 57. During the decade of the 1970s, the developing countries sharply increased their external debts. Availability of petrodollars at low real interest rates made this possible. Also, the prospects for continued growth by the developing countries made lenders comfortable with the increased borrowing. Since 1979, however, developing countries suffered from a combination of external shocks: (a) more expensive oil; (b) historically high real interest rates; (c) prolonged recession in industrial economies; and (d) increased trade barriers. Consequently, economic growth in developing countries slowed down markedly in the 1981-83 period--to 2.5% per annum from the 5.5% per annum average in the 1973-80 period. 58. Debt as a percentage of GNP in the developing countries, which had increased from 14.1% to 20.9% over the 1970-80 period, rose sharply to 31.3% and 33.8% in 1983 and 1984, respectively. The debt service percentage for the developing countries as a whole rose from 16.0% to 20.5% in the 1980-82 period, although it has since declined somewhat. And yet, as many as 100 developing countries have continued to service their foreign debt without 1/ Broad movements of short-term and long-term interest rates over the 1960- 1984 period are shown in Table 3.4. Table 3.4: UNEMPLOYMENT AND INTEREST RATES IN INDUSTRIAL COUNTRIES (percent) 1960-67 1968-73 1973-79 1980 1981 1982 1983 1984 1985 Unemployment (% of total labor force) OECD Total 3.1 3.4 5.2 6.1 6.9 8.3 8.8 8.4 8.3 Major 7 Industrial Countries 2.9 3.2 4.9 5.5 6.5 7.9 8.2 7.6 7.5 United States 5.0 4.6 6.7 7.0 7.5 9.7 9.6 7.5 7.3 Short-term Interest Rates (% per annum) United States Prime Rate 4.83 6.86 8.87 15.27 18.87 14.86 10.79 12.04 9.94 3-Month Treasury Bill Yield 3.49 5.66 6.90 11.51 14.03 10.69 8.61 9.52 7.60 Real Yield of 3-Month Treasury Bill 1.53 /a 0.51 -0.38 2.61 4.93 4.69 4.33 4.68 4.00 Eurodollar Rate, 3-Month Deposits 4.52 7a 7.65 8.50 14.36 16.51 13.11 9.60 10.78 8.35 Long-Term Interest Rates (Z per annum} United States, 10-Year Treasury Bond Yield 4.30 /a 6.48 7.90 11.46 13.91 13.00 11.10 12.44 10.70 Real Yield of 10-Year Treasury Bond 2.30 7 1.33 0.61 2.56 4.61 7.00 6.80 7.54 7.10 /a The period is 1961-1967. Source: OECD, Economic Outlook, December 1985; Economic Outlook: Historical Statistics, 1960-1983. IMF, International Financial Statistics. US Council of Economic Advisers, The Annual Report, February 1985. - 33 - interruption. Some have experienced only minor shocks, e.g., oil exporters and countries with large remittances from their overseas workers. Some had borrowed only a little or mainly on concessional terms in the 1970s (e.g., China, Colombia and India). Others undertook economic policy reforms quickly to facilitate debt servicing (e.g., Indonesia and Korea). Nonetheless, the number of formal reschedulings of payment for World Bank members increased from an average of less than four a year in the period 1975-80 to 13 in 1981 and 31 (involving 21 countries) in 1983. 1/ 59. Countries that experienced debt-servicing difficulties, however, were not necessarily the ones that had suffered the greatest economic shocks. Those that had borrowed heavily and failed to adjust policies quickly were the ones that ran into the most serious debt-servicing problems. One such group of countries includes the low-income African countries who are constrained by weak institutional structures and a shortage of skills, as well as by natural disasters. In recent years, these countries have depended primarily on concessional credit from abroad. Loans from commercial sources have been relatively small. Among the copper-producers, Zambia and Zaire belong to this group. 60. Another group of countries that has experienced debt difficulties includes many countries in Latin America that were able to borrow heavily from commercial sources because they were considered eminently creditworthy (most of them are middle-income countries). Common features of these countries were: (a) fiscal and monetary policies that were too expansionary to achieve a sustainable external balance; (b) overvalued exchange rates; and (c) savings efforts that lagged behind their investment expansions. Some of these countries, e.g., Chile and Uruguay, attempted comprehensive economic reforms, but these were often defective in their coverage and timing. Brazil, the Philippines and Ivory Coast undertook some policy changes but underestimated the length and depth of the recession and the rise of interest rates. Among the copper exporting countries, Chile, Mexico, Peru and the Philippines belong to this group. C. Primary Commodity Markets since World War II 61. That the acceleration in world inflation during the 1970s affected the price levels of all commodities is clear. 2/ Table 3.5 shows price indices for petroleum and the 33 non-oil primary commodities that are important for developing countries, in current dollars, for the period 1948- 1/ The World Bank, World Development Report 1985 (Oxford University Press, 1985). 2/ The question of whether and to what extent commodity price increases in the 1970s "caused" the inflation of the decade has never been settled. For an example of an investigation of this question, see Barry P. Bothworth and Robert Z. Lawrence, Commodity Prices and the New Inflation (Brookings Institution, 1982). - 34 - Table 3.5 IGT'MMD InDX OF CDITT PUQS /A (CURRENS DOLLARS) (1977-1979-100) - ~ANNUAL AVESRUGES 33 CUIOlDITIhS AGRICULTURE TINBER METALS (EXCLUDING TOTAL FOO NON-FOD & (WEIGHTS- PETROLEU EEG) TOTAL smHAGES CEREALS FATS A OILS OTHER MINERALS Z SARE) (100.0) (71.8) (58.3) (26.5) (7.5) (8.6) (15.6) (13.6) (4.8) (23.3) 1948 16 36 39 37 19 54 54 48 46 18 33 1949 13 35 36 35 22 54 40 46 41 19 34 1950 12 41 44 40 28 51 43 53 64 20 35 1951 12 49 53 45 31 56 51 60 87 30 44 1952 12 44 43 39 29 56 42 46 61 22 51 1953 12 40 40 38 30 56 43 41 47 20 46 1954 13 43 43 43 42 52 40 40 46 28 44 1955 13 43 41 38 34 46 37 41 57 22 51 1956 13 43 41 39 35 46 38 43 52 22 54 1957 13 43 43 41 32 44 38 58 50 21 48 1958 12 39 38 36 29 45 36 43 46 20 45 1959 11 38 37 33 26 42 39 38 53 23 44 1960 10 38 37 33 25 40 37 39 55 26 44 1961 10 36 34 31 24 43 39 35 47 26 44 1962 10 36 34 31 23 48 36 36 47 29 42 1963 10 42 44 44 22 48 38 81 44 28 42 1964 9 43 41 40 26 48 39 62 44 24 51 1965 9 40 34 32 25 47 44 31 43 28 59 1966 9 41 34 32 24 53 42 30 44 29 63 1967 9 38 34 32 23 56 40 31 42 31 53 1968 9 39 34 32 23 54 39 31 43 32 54 1969 9 42 37 35 24 54 38 43 46 31 59 1970 9 44 39 38 28 49 45 47 44 33 62 1971 12 42 39 38 24 47 45 53 44 33 53 1972 13 47 46 46 27 50 44 77 45 33 53 1973 18 70 69 68 34 102 91 98 71 58 77 1974 76 110 115 123 39 149 105 262 80 69 105 1975 74 87 88 92 37 116 70 188 69 52 90 1976 79 86 87 87 71 94 74 117 87 70 88 1977 87 97 101 104 120 87 92 91 89 79 90 1978 87 93 94 93 89 104 97 93 96 80 92 1979 126 110 106 103 92 109 111 116 115 141 118 1980 207 140 140 142 82 130 102 272 132 170 132 1981 232 114 111 111 68 139 98 176 115 127 119 1982 225 95 88 85 70 101 79 105 101 128 109 1983 197 98 93 89 71 109 94 107 113 119 109 1984 192 96 91 88 82 106 112 75 106 134 104 /A COMPUTD FM UNROUNDED DATA. /S WEIGTED BY 1977-1979 DEMLOPIIG COUNTRIES' EXPORT VALUES. NOTE: TIIE 'SDITIES INCLUDED IN ZACK GROJP ARU: BEVERAGES-COFFEE, CDCOA. TEA; C2RALS-MAZE5 RICE, MEAT, OAN SORGHUM; FATS AND OILS-PALX OIL, COCONUT OIL, ROUNDUT OIL, soYANS, COPLA. GROUNDNIrr MEAL, SOYBEAN MEAL; OT FooDS-SGAR. EF, NAKAN , ORANGS3; NDE-FODS-COTTON. JT. 0IlDR, ToBAoo; TINER-LOGS; METALS AND MINERALS-COPPER, TIN, NICKEL, BAUXITE. ALUMINM, INO OU, MANI SE OR, LEAD, ZINC, PHOSPRATE ROCK. SOURCE: S WOUD M . EOU NIC MALYtlS & POLJECTIONS DEPARTMT , CQIMDDIl STUDIES a ProJEcTzoS DIVISIOm - 35 - 84. From 1948 to 1970, the price of petroleum declined in current dollars; or more precisely, it declined until 1964 and remained at the 1964 level until 1971. The aggregate index for 33 non-fuel primary commodities shows that, after reaching a peak during the Korean War (1951), prices declined gradually until 1961-62 and then were stable until the end of the 1960s. During the 1970s, the price of petroleum increased dramatically; the level of oil prices in 1980 represented a 2200% increase over the 1970 level. The average price level of non-oil commodities rose by a factor of more than 3 over the same 10- year period. 62. Table 3.6 and Figure 3.1 show how primary commodity prices have moved relative to prices of manufactures exported by industrial countries during this period. The latter price series is represented by the World Bank's Manufacturing Unit Value Index (MUV) as presented in Table 3.7. 63. After increasing sharply during the 1970s, the price of petroleum has been receding steadily in "real" terms since 1981. The market power of the OPEC cartel has been reduced because of (a) the recession in the world economy; (b) steadily increasing supplies of oil from non-OPEC sources; (c) increased supplies of alternative energy sources; and (d) increased application of energy-saving innovations and conservation techniques. 64. The trend in real prices of non-fuel primary commodities has been somewhat different from oil. After peaking in the early 1950s (the Korean War), the aggregate constant dollar index of 33 nonfuel commodities declined until 1958. It "stabilized" at that level until 1972, except for a peak in 1963-64 1/ and a small peak in 1969. 2/ 65. In 1973-1974, as a result of inflation and the sharp increase in demand for non-fuel commodities triggered by the boom in industrial market economies, non-fuel commodity prices rose sharply in real terms, although not as dramatically as the price of oil. Prior to this, the Limits to Growth report of the Club of Rome, published in 1972, foretold a doomsday for the world due to the constraints imposed on the material progress of mankind by 1/ This increase was associated with sharp rises in sugar, beef, some cereals and some nonferrous metals. 2/ Mainly due to increases in sugar, beef, rubber, and, to a lesser extent, lead and nickel. - 36 - Table 3.6 WZIGCTED INDEX Of COODITy P ICES /A (CONSTANT DOLLARS) (1977-1979.100) ANNUAL AVERAGES 33 COUOD0ITSIS -AGRICULTURE TINIZR METALS (EXCLUDING TOTAL FOOD NON-FOOD & (WEIGHTS- PETROLEUM ENERGY) TOTAL JRVERAWES CEREALS FATS & OILS OTHER MINERALS X SHuR) /S (100.0) (71.8) (58.3) (26.5) (7.5) (8.6) (15.6) (13.6) (4.8) (23.3) 1948 50 117 124 118 62 174 174 155 149 58 107 1949 43 117 121 117 72 180 135 155 137 64 114 1950 44 156 169 152 106 196 163 203 243 77 134 1951 37 159 169 143 98 179 163 192 280 95 141 1952 36 137 135 122 91 174 133 145 189 69 159 1953 40 131 130 124 97 182 141 133 153 66 150 1954 43 142 144 142 141 172 133 135 154 93 147 1955 42 139 134 122 109 149 121 132 185 73 167 1956 41 138 132 124 112 147 121 136 164 68 172 1957 39 132 131 126 98 134 116 176 153 63 147 1958 35 112 109 104 85 129 104 123 134 57 130 1959 33 115 112 101 79 129 118 115 160 71 132 1960 30 113 110 97 74 120 110 117 164 76 132 1961 30 107 102 93 70 128 113 103 139 78 129 1962 28 108 103 94 68 143 107 107 141 87 127 1963 28 127 130 130 67 143 113 241 130 84 124 1964 26 124 119 117 75 138 114 179 129 70 149 1965 25 114 99 92 71 135 126 90 126 81 170 1966 24 111 93 87 66 144 116 81 120 80 173 1967 24 103 91 86 63 150 108 83 113 84 143 1968 25 111 97 91 67 156 111 89 125 92 156 1969 25 120 106 100 68 153 109 122 132 88 169 '1970 23 114 101 9 72 126 116 122 112 85 160 1971 27 100 93 91 58 112 108 126 104 80 127 1972 28 102 99 99 58 109 96 167 98 72 115 1973 33 128 126 124 62 187 165 179 130 105 140 1974 110 160 166 178 57 216 152 381 116 101 152 1975 94 111 113 118 48 148 89 240 89 67 115 1976 100 109 109 109 89 119 93 147 110 88 110 1977 101 113 117 120 139 101 107 105 103 92 105 1978 86 92 93 92 88 103 96 92 95 79 91 1979 112 98 94 92 82 97 99 103 102 125 105 1980 169 115 115 117 67 107 84 223 108 139 108 1981 200 98 96 95 59 120 85 152 99 110 103 1982 198 83 77 74 62 89 70 92 89 112 96 1983 181 90 86 82 65 100 87 98 104 109 100 1984 182 91 86 83 77 100 106 71 100 127 99 /A COMPUTED FROM UNROUNDED DATA AND DEFLATED BY MANUFACTURING UNIT VALUE (W6V) INDEX. /2 WEIGHTED BY 1977-1979 DEVELOPING COUWTRIES' CPORT VALDIS. OTE: THi COI60DITIS8 INCLUDED IN EACH GROU ARE: XVIRAA U-comS, COCOA, TEA; CERRALS--MAIZE, RICE, IMEAT, GRAIN SO RUN; FATS AND OILS-PALI OIL, COCOWUT OIL, GROUNDIUJ OIL, SO13NS, COFRA, GCOUNDONUT MAL, SOT51M MMAL; OTHER FOODS-USDGA, 3 -, BANANS , ORANGS; NON-FOODS--COTTON, JUE IusQ. TORACCO; TINDER-LOGS; I TALS AND MINERALS-COPPER, TIN, NICFEL, BAUXITE, ALMNUM, 110 ORE, MNUGANESE OU, LUAD, ZINC, PUOSFWE ROCK. SOUCE: WVDLD MAM, ECONOMIC ANALYSIS 4 PROJECTIONS DNFAETNUT., COmOIT! STUDIES G POJECTIONS DIVISION. Figure 3.1: WEIGHTED INDEX OF COMMODITY PRICES: 1948-1983 CONSTANT US DOLLARS (1977/79=100) 240 180- 200 160 33 COMMODITIES 160 140 120 |120 80 100 40 80 0 ... ,,,,,,,,6v- ...,,,,,, 48 52 5660646872 7680 48 525660646872 7680 *AVERAGE OPEC PRICE 200 180 290 160 250 NON-FOODS 140 / ATOTAL FOOD | 210 ,. 100 2130 so 90 METALS & MINERALS 48 52 56 60 64 68 72 76 80 48 52 56 60 64 68 72 76 80 Source: Table 3.6. - 38 - Table 3.7: Manufacturing Unit Value (Muv) Index, 1948-95 /a Year (1980-100) (1984-100) (Z Change) 1948 26.8 28.3 1949 25.7 27.1 -4.0 1950 22.6 23.8 -12.2 1951 26.1 27.5 15.5 1952 27.3 28.8 4.7 1953 26.6 28.1 -2.7 1954 26.0 27.4 -2.1 1955 26.5 28.0 2.0 1956 27.4 28.9 3.2 1957 28.0 29.5 2.3 1958 28.5 30.1 1.7 1959 28.1 29.6 -1.3 1960 28.7 30.3 2.2 1961 29.2 30.8 1.5 1962 29.7 31.3 1.6 1963 29.2 30.8 -1.6 1964 29.8 31.4 1.9 1965 30.0 31.6 0.8 1966 31.1 32.8 3.4 1967 31.4 33.1 1.2 1968 31.2 32.9 -0.7 1969 32.8 34.6 5.0 1970 34.9 36.8 6.1 1971 36.8 38.8 5.3 1972 40.0 42.2 8.6 1973 46.4 48.9 14.7 1974 56.5 59.6 19.8 1975 62.8 66.2 10.6 1976 63.7 67.2 1.4 1977 70.0 73.8 9.4 1978 80.5 84.9 14.0 1979 91.2 96.2 12.4 1980 100.0 105.5 9.2 1981 100.5 106.0 0.5 1982 99.1 104.5 -1.4 1983 96.5 101.8 -2.6 1984 94.8 100.0 -1.8 1985 96.0 101.3 1.3 1986 102.9 108.5 7.2 1987 109.9 115.9 6.8 1988 117.4 123.8 6.8 1989 125.6 132.5 7.0 1990 134.5 141.9 7.1 1995 163.7 172.7 4.0 /a Unit Value Index of Manufactured Export. (SITC 5-8) from five industrial market economies to developing countries on a CIF basis. For history, the source is the UN, Monthly Bulletin of Statistics. Computed from unrounded data. Source: World Bank, Economic Analysis and Projections Department, Commodity Studies and Projections Division and Global Analysis and Projections Division. - 39 - the limitations of nonrenewable natural resources. 1/ Concerns about availability of raw materials, food and energy, and security of supply of "critical" raw materials (because of sudden interruptions of supply, intentional or otherwise), occupied high priority inside and outside the governments of leading economies. 66. This climate of opinion gave birth to several attempts at controlling supplies of nonfuel commodity markets by producers. However, these attempts did not play a prominent role in raising prices of non-fuel primary commodities, except for a very few commodities and then only on a limited scale and for a short duration. Collaborative supply management by producers may be said to have played a role in price increases in bauxite, phosphate rock and potash. However, the high prices achieved in 1974-75 for phosphate rock and potash did not last more than two. Only in the case of bauxite were high prices continued into subsequent years. 67. The deep recession of 1974-75 severely reduced demand for most nonfuel commodities. A close similarity exists between the real-price movements of the metals/minerals and agricultural groups after the peak in 1973. From 1973 to 1977, price indexes in constant dollars of both groups declined by 30%. After dropping further in 1978, both reached moderate peaks in 1980. Both indexes fell sharply in 1981-1982, recovered moderately in 1983 but remained flat in 1984. 68. The current dollar prices of non-fuel commodities have been declining since 1980, partly as a result of the strengthening of the value of the US dollar against other major currencies. The 1981-82 world-wide recession was a major reason for the fall in commodity prices. However, in the 1983-84 recovery, the current-dollar commodity prices failed to respond as in past upswings. The aggregate non-fuel commodity price index in current dollars increased only slightly in response to the world economic recovery in 1983- 1984 (and appears to have declined in 1985). This puzzle can be explained largely by the strengthening of the value of the US dollar (in which the commodity price index is measured) since 1980. 1/ If the present growth trends in world population, industrialization, pollution, food production, and resource depletion continue unchanged, the limits to growth on this planet will be reached sometime within the next 100 years." See Meadows, D.H., Dennis L. Meadows, Jorgen Randers and William W. Behrens III, The Limits to Growth, A Report for the Club of Rome's Project on the Predicament of Mankind, Universe Books Publishers 1972, p. 29. - 40 - IV. STAGNANT CONSUMPTION AND OVERCAPACITY IN THE WORLD COPPER MARKET SINCE 1974 69. As described earlier, by historical standards the world copper market has been depressed since 1974, except for a brief period around 1980. Growth in world demand for copper has been sluggish and the average constant dollar price has been not only substantially lower than the average level in the "high-price" period of 1964-74, but even lower than in the "medium-low" price period of 1950-63. This prolonged stagnation in the world copper market has been caused by several factors, some general and some copper-specific. These factors are discussed in this and the following three chapters. This chapter analyses the factors that may have contributed to the depressed market conditions from the demand side and discusses the consequent overcapacity conditions. A. Slowdown in Demand Growth 70. World (excluding CPEs) consumption of refined copper, which peaked at 7.2 million tons in 1973, surpassed this peak level again only in 1977-78; after that it stagnated until 1984 when it sharply rose to surpass the 1978 peak. The trend growth over the period 1973-83 was 1.1%, as compared with the average rate of 4.7% per annum during the period 1950-73. 71. The marked slowdown in demand growth since 1973 has not been specific to copper. Demand growth in all other major metals and minerals has also been sluggish. Table 4.1 compares the demand growth trends in the periods 1961-73 and 1973-83 for selected major metals and minerals. 72. One obv-ious, and perhaps most important, factor responsible for the slowdown in demand growth for copper (and for other metals as well) is the slowdown in world economic growth since 1973. This decline in GDP growth was reflected in the consumption of all metals and minerals (indeed, in consumption of all industrial raw materials including such agricultural products as rubber and cotton.) 73. The share of value added in industry, particularly in manufacturing, as a percentage of GDP in industrial countries has been steadily declining (Table 4.2), while the share of services has increased. Reflecting this, the rate of growth in industrial production has slowed down even more dramatically than the slowdown in GDP growth in industrial countries (Table 4.1). Minerals Table 4.1: WORLD (EXCLUDING CPES) CONSUMPTION OF SELECTED MAJOR METALS, AND INDUSTRIAL PRODUCTION AND REAL GDP IN INDUSTRIAL COUNTRIES: 1961 TO 1984 Consumption /a Growth Rates /b Unit 1961 1970 1973 1982 1983 1984 1961-73 1973-84 Aluminum, Primary '000 tons 3,471 8,149 11,182 10,906 12,074 12,652 9.1 1.4 Copper, Refined '000 tons 4,219 6,023 7,215 7,065 6,820 7,551 3.9 1.2 Lead, Refined '000 tons 2,507 3,486 3,797 4,033 3,810 3,959 3.1 -0.2 Zinc, Refined '000 tons 2,688 4,368 4,829 4,491 4,582 4,713 4.5 -0.3 Tin '000 tons 169 185 200 153 141 154 1.4 -2.4 Nickel '000 tons 223 444 502 422 498 581 7.0 1.3 Steel million tons 262 441 488 460 473 647 /c 5.3 -2.6 Industrial 1975 - 100 56.7 90.9 98.6 118.2 125.9 135.1 4.7 2.9 Production Index /d Real GDP /d Billion US$ at 3,499 5,422 5,954 7,525 7,885 8,248 4.5 3.0 1980 prices /e /a Includes secondary metal, unless otherwise noted. 7W_ Least squares trend growth rates for aluminum, copper, lead and zinc; end-point growth rates for tin, nickel and steel. 7T Estimate. 7T Industrial countries only. T At 1980 exchange rates. Sources: Metallgesellschaft AG; World Bureau of Metal Statistics; International Lead and Zinc Study Group; International Tin Council; International Iron & Steel Institute; OECD; the World Bank. Table 4.2: SHARE OF VALUE ADDED AS A PERCENTAGE OF GDP IN OECD COUNTRIES, BY SECTORS, 1960-1983 Average during Period of 1960- 1968- 1974- 1980 Sector 1960 1970 1973 1975 1980 1983 1967 1973 1979 1983 Agriculture 6.4 4.3 5.0 4.5 3.6 3.0 5.7 4.5 4.2 3.4 Services 53.6 57.4 56.8 58.7 59.6 61.9 54.6 57.3 58.6 60.7 Industry 40.0 38.3 38.2 36.8 36.8 35.0 39.7 38.2 37.1 35.8 of which, Manufacturing 29.7 28.2 28.0 26.1 25.3 23.7 29.4 28.1 26.5 24.3 Source: OECD, Economic Outlook Historical Statistics, 1960-1983 (Paris, 1985). - 43 - consumption is more directly affected by changes in the activity level of the industrial sector than by changes in any other sector. 1/ 74. The increased service-sector orientation of industrial economies has accelerated--if not caused--the declining trend in the "intensity of use" of metals/minerals in industrial countries in the last 10-15 years. The concept of "intensity of use" is simply a ratio of the volume of consumption of a particular raw material to a unit of GNP; it is typically expressed in tons (or kilos) per one million constant US dollars of GNP. It is believed that intensity of use increases as per capita GNP rises, and that, at high levels of per capita GNP, it begins to decline. This observation is based on the fact that, as per capita GNP rises, the importance of the manufacturing sector typically increases, entailing an increase in consumption of raw materials; in contrast, at high per capita incomes the service sector's share in GNP typically increases and the relative importance of the manufacturing sector decreases. 2/ 75. The intensity of use of most major metals has declined in industrial countries over the last 10-15 years. For instance, steel consumption per million US dollars of GDP (measured in constant 1980 dollars) in industrial countries, which had increased from 67 tons to 70 tons in the period 1960-70, declined to 54 tons by 1975 and to 47 tons in 1980. The intensity of copper use per million US dollars of GDP (measured in constant 1980 dollars) in industrial countries, which had increased from 1,107 kg in 1960 to 1,125 kg in 1965, decreased to 745 kg in 1983 (Table 4.3). The declining pattern in the intensity of use of copper holds for every major industrial country/region shown in Table 4.3. In contrast, the intensity of copper use in developing countries has generally risen over the last two decades, except that for Southern Europe and China it has declined somewhat since the mid-1970s. 3/ 76. In industrial countries, while the total fixed investments have stagnated, those in "high-technology" and communications equipment have grown rapidly. This disparity in growth rates between the "high-tech" and traditional industries appears to have had differential impacts on consumption of raw materials. Namely, in contrast to the stagnant demand for traditional, 1/ Richard M. Auty, however, cautions: "Vogely, noting that government expansion accounts for a large fraction of the growth in US services, points out that expenditure on schools, highways, defense and energy R and D may be relatively resource consuming rather than resource conserving". See "Materials Intensity of GDP," Resources Policy, December 1985, p. 281. 2/ For an application of the concept for raw materials demand forecasting, see Wilfred Malenbaum, World Demand for Raw Materials in 1985 and 2000, New York, McGraw-Hill, 1978. 3/ In the case of China, statistics on both copper consumption and GNP data are subject to large margins of errors. As for Southern Europe, the reason for the recent decline needs to be investigated. Table 4.3: CHANGES IN COPPER INTENSITY BY REGION, SELECTED YEARS 1960 1965 1970 1974 1975 1980 1981 1982 1983 --------------------(KG per million constant 1980 US dollars----------------- Industrial Countries 1,107 1,125 1,017 936 779 827 835 752 745 United States 940 1,110 963 908 643 722 767 640 664 Canada 1,074 1,598 1,384 1,295 930 863 971 571 654 EEC 1,332 1,201 1,079 997 917 932 875 870 836 Other W. Europe 758 625 501 437 370 370 354 NA NA Japan 1,365 1,193 1,282 1,094 1,013 1,081 1,138 1,097 1,035 Oceania 870 1,071 846 841 688 774 744 720 666 Developing Countries 540 547 533 620 574 640 720 NA NA Southern Europe 1,550 1,444 1,587 1,680 1,459 1,438 1,379 1,357 1,312 Africa 236 241 202 317 295 352 344 NA NA Latin American & Caribbean 343 526 399 563 500 633 556 NA NA Asia Excl. China 432 352 299 318 309 579 679 NA NA China 1,481 1,138 1,389 1,549 1,584 1,233 1,270 1,323 1,239 Centrally Planned Economies 1,116 1,148 1,105 1,111 1,117 1,035 1,010 NA NA Selected Developing Countries Brazil 473 393 607 982 834 960 730 999 627 Korea, Republic of 26 131 285 623 692 1,470 2,353 2,046 2,065 India 846 738 459 403 347 512 472 513 562 Source: World Bureau of Metal Statistics for Refined Copper Consumption; and the World Bank for GNP Data. - 45 - major metals/minerals such as copper, iron ore, lead, zinc and tin, consumption of "new" materials such as beryllium, tantalum, silicon, zirconium, titanium, lithium and platinum has been growing rapidly. Thus, the changing composition of industry seems to have had an impact on the intensity of use of "older" metals/minerals including copper. 1/ 77. Despite the generally rising intensity of copper use, copper consumption in developing countries has stagnated since 1980 as a result of the sharp decrease in their economic growth. However, as economic growth resumes in these countries, copper consumption is expected to resume rapid growth. 78. In the 1981-82 recession, there was a clear disparity in the extent of decline in copper consumption among various regions, especially between North America and the rest of the world (excluding CPEs). Table 4.4 illustrates the recent trends in consumption of refined copper. 2/ In 1982, with the recession underway, cc-s-ption in North America fell sharply, by 20%. In contrast, in the rest of the industrial countries, the effect of the recession on copper consumption was much milder and more gradual; i.e., consumption fell by only 9% and the decline was spread out over a four-year period from 1979 to 1983. 79. In developing countries, there was a sharp fall in consumption in Latin America (including the Caribbean)--from 494,000 tons in 1980 to 303,000 tons in 1983. However, in the rest of the developing regions, the recession- caused fall in consumption was less dramatic. This contrast between Latin America and other developing countries resulted from extraordinarily severe belt-tightening of economic management undertaken by major Latin American countries in the face of the debt crisis in that region. B. Material Substitution and Material Saving Technology 80. Substitution for copper by other materials has adversely affected copper consumption. In addition, technological improvements leading to materials-saving in the use of copper (e.g., miniaturization in electronics products) has reduced copper consumption. Considered in the historical perspective, however, it appears that the loss of copper markets directly 1/ Fernando Gonzalez-Vigil, "New Technologies, Industrial Restructuring and Changing Patterns of Metal Consumption", Raw Materials, vol. 3, no. 3. 2/ Before the recession, different countries/regions experienced peaks in different years; industrial countries generally in 1979, developing America and Africa in 1980, Southern Europe and developing Asia in 1981. In the United States and Canada, there were significant production losses due to an industry-wide labor strike in 1980. Consequently, consumption dipped in 1980, but recovered in 1981. It may be that without the significant production loss in North America, consumption in industrial countries in 1980 would have been as high as in 1979. Table 4.4: REFINED COPPER CONSUMPTION BY COUNTRY/REGION, 1970-1984 ('000 tons) Country/Region 1970 1973 1975 1979 1980 1981 1982 1983 1984 Industrial Countries 5,387 6,331 4,801 6,535 6,024 6,147 5,671 5,746 6,333 North America 2,005 2,386 1,574 2,468 2,076 2,271 1,823 1,970 2,272 United States 1,860 2,221 1,397 2,165 1,868 2,030 1,664 1,775 2,041 Canada 145 164 177 303 209 242 159 195 231 EEC-9 2045 2,180 1,967 2,337 2,318 2166 2,145 2.096 2,227 Japan 821 1,202 828 1,330 1,158 1,254 1,243 1,216 1,368 Developing Countries 636 905 969 1,378 1,476 1,494 1,410 1,465 1,663 Asia 302 418 425 610 663 735 651 844 927 China 208 287 300 360 330 330 288 383 393 America 179 268 306 469 495 432 445 319 396 Africa 48 87 86 97 119 116 103 96 109 S. Europe 108 131 151 182 199 212 212 206 232 Industrial & Developing 6,023 7,236 5,770 7,904 7,501 7,641 7,174 7,209 7,981 North America 4,018 4,850 4,196 5,436 5,425 5,378 5,351 5,239 5,709 Source: World Bureau of Metal Statistics. - 47 - attributable to substitution during the last decade was less dramatic than in the preceding two decades. 1/ 81. Before examining substitution of other materials for copper, it is useful to review the special properties of copper and its use for industrial and other purposes. Copper is one of the most versatile metals; it is used both in pure form and in alloys such as brass (with zinc), bronze (with tin), and nickel-silver (with zinc and nickel). The basis for copper's extensive industrial applications includes the following properties of copper: superior electrical conductivity, high thermal conductivity, strong resistance against corrosion, ductility, malleability, high strength, lack of magnetism, bacteria resistance, and a pleasing reddish color. Because of these characteristics, copper consumption is spread over a variety of industrial sectors. 82. End-use data on copper consumption are relatively scarce, and, when available from published sources, are usually spotty in coverage. However, Brook Hunt & Associates, Ltd. (a consulting firm based in London) has continued to collect such statistics after the CIDEC (based in Switzerland) was disbanded in the early 1970s. Table 4.5 summarizes Brook Hunt data on end-use consumption for "Western Europe", 2/ Japan and the United States in 1970 (the earliest year for which such data are available for all three regions) and 1983. 3/ These three regions together account for about 83% of refined copper consumption in the market economies, or 61% of world consumption. 83. The electrical industry has been by far the most important copper consuming industry, accounting for 50.2% of total apparent consumption in the three regions in 1983. General engineering and construction industries accounted for 18.4% and 14.7%, respectively, of the total. The other two major copper consuming end-uses are the transport industry and domestic uses, which together account for 16.7% in 1983. The use of copper in the production of wires and cables account for over 90% of copper consumption in the electrical industry. 84. The main substitutes for copper have been aluminum and plastics; aluminum has been used primarily for electrical and thermal conductor purposes and plastics for anti-corrosion purposes. Comprehensive data showing trends in substitution of aluminum for copper are not available. However, as 1/ Brook Hunt & Associates, Ltd., Quarterly Service, Issue XV, Copper, July 1984, p. 26. 2/ Brook Hunt's definition of Western Europe includes the Bank's Western Europe and Southern Europe. 3/ In dealing with copper consumption by end-use, Brook Hunt refers to refined copper consumption plus the direct use of scrap (scrap used without refining). - 48 - Table 4.5: APPARENT COPPER CONSUMPTION BY END-USE INDUSTRY, WESTERN EUROPE /a, UNITED STATES AND JAPAN, 1970 AND 1983 1970 1983 000 tons Z Share 000 tons x Share Western Europe /a Electrical /9- 1,487 48.0 1,658 49.1 Construction 489 15.8 582 17.2 Transport 308 9.9 261 7.7 General Engineering 574 18.5 615 18.2 Domestic Uses 240 7.8 262 7.8 Sub-Total 3,098 100.0 3,378 100.0 United States Electrical /b 1,018 44.4 1,220 50.6 Construction 445 19.4 368 15.2 Transport 196 8.6 178 7.8 General Engineering 460 20.1 478 19.7 Domestic Uses 173 7.5 164 6.7 Sub-total 2,292 100.0 2,408 100.0 Japan Electrical /b 624 55.3 769 52.1 Construction 90 8.0 122 8.3 Transport 169 15.0 228 15.4 General Engineering 178 15.7 239 16.2 Domestic Uses 68 6.0 118 8.0 Sub-total 1,129 100.0 1,476 100.0 Total of Three Regions Electrical /b 3,129 48.0 3,647 50.2 Construction 1,024 15.7 1,073 14.7 Transport 673 10.3 667 9.2 General Engineering 1,212 18.6 1,333 18.4 Domestic Uses 481 7.4 544 7.5 Total 6,519 100.0 7,264 100.0 /a Inclues Southern Europe. 71W Includes Telecommunications. Source: Brook Hunt & Associates, Ltd., Quarterly Service, Issue XV, Cogper, July 1984, p. 78. Cited with permission of Brook Hunt & Associates, Ltd. - 49 - fragmentary evidence presented below indicates, it appears that the substitution by aluminum for copper has slowed down since the early 1970s. 85. The most important area of aluminum substitution for copper has been the area of electrical conductivity. Aluminum is the only metal with suitable properties available in sufficient quantities for general use in place of copper as electrical conductors. 1/ The weight of an aluminum conductor with equivalent electrical conductivity is only 50% of the weight of the copper conductor. However, in practice, because more insulation is needed for the bulkier aluminum conductor and because aluminum has poorer mechanical and ductile properties, copper needs to be only 35% (rather than the theoretical 50%) cheaper than aluminum (per unit of weight) for copper to be competitive. 2/ 86. Substitution of aluminum for copper in electrical cable and wire markets has been a long standing practice. The first recorded instance of such substitution was in 1897 when copper telephone wiring was replaced by aluminum wiring in Chicago. 3/ However, until the 1950s, substitution by aluminum in electrical cables and wires progressed only gradually. 87. In the late 1940s, aluminum became less costly than copper in terms of price per unit of weight. 4/ Substitution for copper by aluminum in electrical conductors accelerated in the 1950s and 1960s, with the copper- aluminum price ratio distinctly favoring the use of aluminum as an electrical conductor. For example, an earlier study indicates that in the decade up to 1971, conductor uses (in terms of electrical equivalency) increased on the order of 50% for copper and over 100% for aluminum. 5/ Also, data available from the International Wrought Copper Council (IWCC) indicate that the share of aluminum in consumption of copper and aluminum (in electrical equivalency) 1/ National Research Council, National Materials Advisory Board, Committee on Technical Aspects of Critical and Strategic Materials, Mutual Substitutability of Aluminum and Copper, April 1972. 2/ Brook Hunt Associates Ltd., Quarterly Service, Issue XII, Copper, December 1982, pp. 44-45. Strictly speaking, this is too simplistic. There is clearly a sliding scale of trade-offs for price-related substitution, depending on the specific end-uses. 3/ Rhea Berk, Howard Lax, William Prast, and Jack Scott of Atlantis, Inc., Aluminum: Profile of the Industry, McCraw-Hill, Inc., 1982, p. 157. 4/ Aluminum has always been less costly than copper on a per unit of volume basis since the late 1800s (aluminum being the much lighter metal with its specific gravity at 2.7 compared with copper's 8.94). 5/ National Research Council, National Materials Advisory Board, Mutual Substitutability of Aluminum and Copper, p. 37. - 50 - in cable manufacturing increased from 7.5% in 1964 to 25.8% in 1973 in a wide range of countries (Table 4.6). 1/ 88. However, substitution by aluminum in the electrical cable market has slowed down since the early 1970s. The IWCC data indicate that total consumption of copper and aluminum (in electrical equivalency) in cable manufacturing has stagnated since 1973-74 and that the share of aluminum has been fairly constant at around 25-27% since 1973 (Table 4.6). The Aluminum Association (of the United States) data also indicate that in the United States, shipments of ACSR 2/ and bare cable made of aluminum, which increased from 83,000 tons in 1960 to a peak of 231,000 tons in 1966, and shipments of aluminum bare wire, which increased from 25,000 to 48,000 tons in the same period, have not surpassed these 1966 levels since (except for the sharp rise in shipments of the the latter category during the period 1972-74). 3/ Similarly, shipments of insulated or covered wire and cables made of aluminum, which grew steadily to a peak in 1973-74, have not grown at all since. 4/ 89. Aluminum has also been the prime substitute for copper in heat exchange applications, though not in all such applications. Substitution in this area, in contrast to electrical areas, seems to have accelerated in the 1970s. For example, in Western Europe the use of aluminum radiators in automobiles has made significant inroads in the last decade. As of 1983, 65% of all new cars produced in Western Europe were equipped with aluminum radiators. 5/ In contrast, in the United States and Japan penetration by aluminum in the new car radiator market was only 17% and 4%, respectively, as of 1983, but automobile producers' interest in switching from copper to aluminum in this end-use has since increased as have their research and development expenditures in this area. In the United States, an increased penetration in this area is imminent. 90. A model-based simulation of a counterfactually higher aluminum price in the 1970-83 period indicates that an assumed copper-to-aluminum price ratio of 0.65 for the entire period would not change the total volume of copper consumption significantly, although the price of copper would have averaged 1/ For the list of countries covered, see Table 4.6, footnote (a). 2/ Aluminum stranded conductor reinforced by a core of steel. 3/ The Aluminum Association, Aluminum Statistical Review: Historical Supplement, 1982; and Aluminum Statistical Review, 1981, 1982 and 1983. 4/ Ibid. 5/ Brook Hunt Associates, Ltd., Quarterly Service Issue XV, Copper, July 1984, p. 60. - 51 - Table 4.6: CONSUMPTION /a OF COPPER AND ALUMINUM IN CABLE MANUFACTURING /b 1964-1984 Combined Copper Aluminum /c Total Share Share (-100%) Consumption in Total Consumption in Total Year 000 Tons 000 Tons % 000 Tons % 1964 816 755 92.5 61 7.5 1965 997 907 91.0 90 9.0 1966 1,067 915 85.7 152 14.3 1967 1,407 1,184 84.2 223 15.8 1968 1,329 1,116 84.0 213 16.0 1969 1,630 1,378 84.5 252 15.5 1970 1,725 1,419 82.2 307 17.8 1971 1,617 1,288 79.6 330 20.4 1972 1,706 1,362 79.8 345 20.2 1973 2,122 1,574 74.2 548 25.8 1974 2,309 1,788 77.4 521 22.6 1975 1,993 1,500 75.3 492 24.7 1976 2,179 1,640 75.3 539 24.7 1977 2,382 1,797 75.4 585 24.6 1978 2,431 1,821 74.9 610 25.1 1979 2,184 1,597 73.1 588 26.9 1980 2,265 1,581 69.8 684 30.2 1981 2,150 1,561 72.6 589 27.4 1982 2,095 1,560 74.5 535 25.5 1983 2,012 1,479 73.5 534 26.5 1984 2,045 1,533 75.0 512 25.0 /a Reported consumption by the member companies of the International Cable Development Corp. (ICDC) and the International Telephone Cable Development Association (ITCDA) and in Japan. Members of ICDC/ITCDA are; Austria, Belgium, Denmark, Finland, France, Germany, F.R., Italy, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom, India, Israel, and South Africa. /b Figures exclude winding wires from 1979 onwards. 7- The aluminum figures are expressed in the approximate electrical equivalent of copper tonnages, i.e., actual weight of aluminum multiplied by two. Source: International Wrought Copper Council. - 52 - up to 17% higher than the actual level. 1/ This is because the effect of raising copper consumption in Western Europe would be offset by the effect of higher copper price reducing copper consumption in the United States. In Japan, a hypothetically higher aluminum price has only negligible effect on the volume of copper consumed because copper consumption in Japan is essentially insensitive to the price of copper. 91. Substitution for copper by plastics is a relatively new phenomenon; it has become a significant factor only in the last decade. It has affected mainly the construction industry, with respect to cold and hot water distribution systems (tubing, plumbing, toilet fittings) and other miscellaneous copper and brass products such as door knobs and locks. Several different kinds of plastics are involved; PVC, PB, PE and PP. 2/) As prices of plastics came down rapidly (in real terms) in the second half of the 1970s and as new varieties of plastics were put into commercial use, plastics rapidly substituted for copper and copper-based alloys as well as for a host of other materials. A rapid further penetration by plastics is widely predicted. 3/ 92. The most important recent trend in substitution against copper is the use of optical fiber cables in telecommunications but its impact on copper consumption so far has been relatively small. 93. There have been instances in which copper has substituted for other materials, especially in the construction industry. For example, copper tubes have substituted for black iron, galvanized and lead pipes, particularly in Europe. On balance, for a variety of reasons (including building codes and industry practice), copper consumption in the construction industry has increased, except in the United States (Table 4.5). l/ See C. Suan Tan, "A Study of the World Copper Market, 1964-1995: An Econometric Analysis," The World Bank, 1986. This price effect is an overestimate because the model used includes mining capacity exogenously. Namely, it assumes that capacity does not change in response to changes in copper prices. If the model assumed that copper mining capacity would increase as the copper price is raised for the period covered as a result of a higher aluminum price, then the results of the above simulation would have indicated an increase in copper consumption (and production) and an increase less than 17% in the copper price. 2/ PVC is polyvinyl chloride; chlorinated polyvinyl chloride (CPVC) is especially important in the present context. PB is polybutylene. Crosslinked PE and PP stand for crosslinked polyethylene and polypropylene respectively. Table 4.7 gives indications of what types of plastics are used in what types of applications in Europe. 3/ See, for example, "Growing Challenge to Metal Pipe: CPVC & PB Pipe," Plastics World, September 1981. - 53 - Table 4.7: CONSUMPTION OF HOT-WATER PLASTICS PIPE IN WESTERN EUROPE /a AS OF 1980 Type of Plastics Crosslinked Application PVC PB PE PP TOTAL -----------------------tons--------------------- 1980 District Heating 0 30 150 0 180 In-Floor Heating 0 2,085 3,300 5,915 11,300 Central Heating 115 150 365 50 680 Plumbing 215 150 100 15 480 Other 110 50 320 70 550 TOTAL 440 2,465 4,325 6,050 13,190 /a Countries included are Belgium, France, Germany (F.R.), Netherlands, Italy, Finland, Norway, Sweden, Demark, Switzerland and the United Kingdom. Source: A study by Consultex SA (Bogis-Bossey, Geneva, Switzerland); cited in "Big Growth Seen for Crosslinked PE and PB in Hot-Water Pipe," Modern Plastics International, December 1981, p. 6. - 54 - 94. Technological progress in the last decade appears to have had an effect of "saving" copper use as a material. Examples abound in the downsizing and miniatuarization of electronic and electrical equipment (printed circuits, for example). Multiplexing in telecommunications has multiplied the transmission capacity of copper telecom cables. New technologies have made in possible to achieve the same degree of heat exchange with less volume of copper use. However, it is difficult to estimate the aggregate copper-saving effects that have resulted from such technological changes. 95. In sum, in the last decade or so, penetration by aluminum has slowed down while new penetration by plastics has begun to gather momentum, especially since the late 1970s. On the whole, the rate of substitution against copper since the early 1970s seems to have been no greater than in the preceding two decades. In the decade ahead, substitution by plastics is likely to accelerate and the impact of optical fiber technology on copper use in telecommunications will become significant. C. Chronic Overcapacity 96. The world copper industry has suffered through a period of sustained overcapacity since 1975 largely as a result of the slowdown in consumption growth. Market economies' mine capacity utilization rates averaged 86.6% in the 1975-83 period as compared to 93.5% in the 1960-74 period (excluding 1967- 68 when there was a 9-month-long, industry-wide strike in the United States, the world's largest producer). 97. During the decade prior to 1975, mine capacity in the market economies expanded rapidly (at 4.8% per annum) as new investment responded to the high growth of demand and the high prices. And yet, there was no prolonged period of overcapacity in that decade (Figure 4.1). 98. During the recessionary period of 1974-76, mine capacity in the market economies continued to expand significantly (2.3% p.a.) because of the inertia inherent in the investment process of the mining industry. This capacity increase in the midst of a deep recession set the stage for the prolonged period of overcapacity. Although the rate of expansion in mine capacity in the market economies further slowed to 0.8% p.a. in the 1977-83 period, capacity utilization rates have remained well below 90% for most of this period because of the sustained stagnation in consumption (Figure 4.1). Figure 4.1: MARKET ECONOMIES' COPPER MINE CAPACITY AND PRODUCTION (MILLION TONS) 7.5- 7- 6.5 6- 0 YEA o 5.5- 5- 4.5- 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 YEAR Source: Mine capacity- Phelps Dodge Corporation; Mine production - World Bureau of Metal Statistics. - 56 - 99. Relative to the early 1970s, commercial copper stocks in the market economies have been at very high levels for most of the time since 1974, 1/ with the exception of the 1979-81 period (Table 4.6). 2/ From the end of 1983, stocks declined steadily. However, having reached a bottom of just over 1.0 million tons at the end of March 1985, the level of commercial stocks edged up again, reaching 1.1 million tons by the end of September 1985. This figure is perhaps only a shade higher than the level traders consider to be "normal" relative to annual consumption. 100. The prolonged period of excess capacity conditions has led to reductions in average capacity utilization rates in most of the market economies, but the impact has been uneven across countries. Broadly speaking, capacity utilization in North America has been consistently lower than in the rest of the market economies since 1974. In the decade 1974-83, capacity utilization in North America averaged 78%, compared to 92% in the rest of the market economies. 3/ Much of this difference in capacity utilization resulted from the production cutbacks implemented by North American producers in the face of depressed market conditions and large financial losses, although some allowance must be made for periodic production losses resulting from labor strikes in North America. 101. Some producers outside North America also cut back production. Significant production cutbacks occurred in Australia and the Philippines from time to time. But these, on the whole, were on a smaller scale than the cuts by their North American counterparts. The differences in producers' behavior in this respect will be analyzed below. It suffices here to point out that, because the "burden" of output adjustments to the depressed demand has fallen mainly on North American producers, it has become a political problem assuming international proportions. 1/ Perhaps the fact that during 1975, world reported commercial stocks increased by 85% or 700,000 tons, warrants some comment. Sixty percent of this increase was accounted for by the sharp increase in stocks at LME and COMEX (mainly LME). A large part of this stock buildup at LME was attributable to sales of refined copper at LME by Japanese smelting companies which were forced to take delivery of concentrates under long- term concentrate contracts even though their market could not absorb the copper processed by them. See Robert H. Lesemann (Commodities Research Unit Ltd.), "Copper--Huge Stock Buildup Will Slow Market Recovery," E/MJ, March 1976, pp. 88-91. 2/ As explained in a footnote to Table 4.8, available statistics on stocks are not consistent over time, thus caution must be exercised in interpreting trends in these data. 3/ See Chapter VI, Section C below. Table 4.8: REFINED COPPER STOCKS IN MARKET ECONOMIES, 1960-84 ('000 Tons) Commercial Stocks (1) Total Stocks Reported Commercial Stocks /a Japanese as Percent as Percent Level at End Change in (1) during US Govt. Industry Total of Refined of Refined of Period Period LME COMEX Stockpile Stockpile /b -(1)+(5)-(6 Consumption /c Consumption /c Year (1) (2) (3) (4) (5) (6) (7) (8) (9) 1970 582 263 72 16 230 - 812 10.0 14.0 1971 612 30 140 18 229 - 841 10.7 14.7 1972 659 47 183 52 229 - 888 10.5 14.2 1973 394 -265 35 5 229 - 623 5.7 9.0 1974 832 438 133 39 32 - 864 12.8 13.3 1975 1,536 704 504 91 24 - 1,560 28.1 28.6 1976 1,828 292 603 182 38 50 1,916 28.4 29.8 1977 1,964 136 641 167 21 65 2,050 28.5 29.7 1978 1,535 -429 374 163 21 72 1,628 21.1 22.4 1979 1,090 -445 127 98 26 37 1,153 14.5 15.3 1980 1,029 -61 123 163 20 8 1,057 14.5 14.9 1981 1,087 58 127 170 20 5 1,112 15.0 15.4 1982 1,503 416 253 249 20 0 1,523 22.2 22.5 1983 1,562 59 436 371 20 0 1,582 23.2 23.5 1984 1,193 -369 126 251 20 0 1,213 16.0 16.2 /a Includes LME stocks (Column 3) and COMEX stocks (Column 4) but excludes US Government stockpile (Column 5) and Japanese industry stockpile (Column 6). Data shown are based on the series reported as "total commercial stocks of refined copper" by the World Bureau of Metal Statistics (WBMS). The coverage is incomplete in that the series exludes stocks in countries such as Finland, Norway, Spain, Sweden, Yugoslavia, and some other small producing countries (for years before 1976) as well as US merchant stocks. WBMS started to report stocks of "other countries as producers," beginning 1976, which are included in (1). Apart from this problem, the data for earlier years suffer from further errors of omission in various ways. US merchant stocks (not included here) are estimated to have amounted to 200,000 tons at mid-1978 (CRU Metal Monitor on Copper, February 1982). /b In 1976, the Japanese Government decided to provide financial support for an industry-held stockpile of nonferrous metals. The stockpile was financed by government-guaranteed, interest-subsidized funds. /c Refined copper consumption in market economies. Source: World Bureau of Metal Statistics. - 58 - V. CHANGING COMPETITIVE POSITIONS AMONG PRODUCING COUNTRIES A. Changes in Production Cost Structure 102. The copper industry has experienced dramatic changes in its production cost structure over the last decade. A variety of factors, beyond the industry's control, triggered these changes. These factors include; (a) drastic increases in energy costs during two rounds of oil price hikes led by the OPEC (1973 and 1978), (b) tighter pollution control standards and their stricter enforcement, especially in industrial countries, (c) high inflation followed by high interest charges, and (d) considerable and frequent changes in exchange rates between the US dollar and other major currencies and also between the dollar and the currencies of major copper producing countries. The dynamics between these external factors and the vigorous efforts by the producers to ensure their own survival accelerated the changes in the production costs. 1/ 103. This section reviews the changes in the copper industry's cost structure, and highlights the changes in the relative positions of major copper producing countries. The analysis focuses on cash production costs rather than full costs (which include depreciation charges) because in times of oversupply and production cutbacks, the cash costs are more relevant in a comparison of the cost competitiveness of existing producers. 104. Production cost data provided by Brook Hunt & Associates Limited are used in the analysis. 2/ The data cover 14 major copper producing countries and represent about 80% of the market economies production. 105. Production costs are categorized into four parts, i.e., direct costs, indirect costs, interest and co/by-product credits. These four items have been estimated on the assumption that production schedules are uninterrupted. The specific definitions of the categories of costs are as follows: (a) Direct Costs: Direct cash operating costs include (i) mining and milling costs, (ii) smelting and refining costs, (iii) all freight costs to market and (iv) marketing costs. 1/ Country specific factors such as transportation problems, strikes, and debt servicing problems also have helped accelerate the structural changes in production costs in some countries. 2/ Brook Hunt & Associates Limited, Western World Copper Costs and Production Strategy 1975-85, 5 vols., London, December 1984 (a multiclient study for subscribers only). - 59 - (b) Indirect costs: include (i) administration and corporate overheads, (ii) research and exploration and (iii) royalties and "front-end taxes" (excluding all corporate income and profit taxes). (c) Interest: All interest expenses (net of any interest receivable) on short-term loans, long-term loans, overdrafts, commercial paper, etc. (d) Gross costs: The sum of the costs of (a), (b) and (c) above. (e) By/co-product credits: Revenues from by/co-products. (f) Net costs: Gross costs less by/co-product credits. Tables 5.1 (A) and (B) show the production cost data for the major copper producing countries for the years 1975, 1980 and 1984. 106. These tables show that in the 1975-80 period, gross production costs increased by an average of 53%. 1/ But in the 1980-84 period these trends were reversed and world average gross production costs declined by 14%. In that period, average direct and indirect costs both declined by 14% and 32%, respectively, while average interest costs continued to increase by 50% over the period. A detailed analysis by each category of costs is given below. 107. Direct Costs: The share of direct costs in total cash costs was constant at around 80% over the 1975-84 period. Throughout the period, that producers with the highest direct costs were Canada and Sweden, both of whom had high co/by-product credits. The lowest direct cost producers were Chile and Indonesia where by-product credits were of minor importance. 108. Indirect Costs: Both absolute amounts of indirect costs (including front-end taxes) and their average share in gross production costs increased during the 1975-80 period. However, the average level of indirect costs in 1984 decreased again to the 1975 level. There were wide differences in indirect costs among the countries studied. Producers with relatively low indirect costs were Papua New Guinea, Indonesia, the United States, South Africa, the Philippines and Australia. Producers with high indirect costs were Mexico, Peru, Chile and Canada. 109. Interest Payments: The importance of interest charges increased substantially during the 1975-84 period both in absolute dollar terms and in terms of the percentage share in cash costs. Average interest costs rose from C2.9/lb in 1975 to c7.2/lb in 1984. The producers with relatively lower 1/ What counts, of course, is net production costs (i.e., net of by/co- product credits) rather than gross production costs. An analysis of net production costs will be done below, in the section on the impact of by/coproducts. - 60 - Table 5.1 (A): PRODUCTION COSTS FOR MAJOR COPPER PRODUCING COUNTRIES (C/lb.) 1975 Co/By- Direct Indirect Gross Product Net Costs Costs Interest Costs Credits Costs (A) (B) (C) (D) (E) (F) Zambia 52.6 7.9 2.7 63.2 1.6 61.6 Zaire 42.8 23.3 1.7 67.8 12.7 55.1 S. Africa/Namibia 45.7 3.0 0.1 48.8 7.5 41.3 Australia 49.3 2.4 - 51.7 13.4 38.3 P.N.G. 46.4 0.8 1.1 48.3 24.5 23.8 Philippines 43.7 4.6 1.5 49.8 11.7 38.1 Indonesia 36.2 2.1 7.9 46.2 10.7 35.5 Canada 98.5 15.8 4.4 118.7 90.3 28.4 U.S.A. 59.9 4.6 4.1 68.6 7.0 61.6 Mexico 61.3 22.6 3.1 87.0 59.7 27.3 Chile 36.5 14.1 2.1 52.7 5.5 47.2 Peru 87.3 22.1 2.2 111.6 60.5 51.1 Sweden - - - - - - Average 57.6 10.4 2.9 70.9 22.0 48.9 1980 Co/By- Direct Indirect Gross Product Net Costs Costs Interest Costs Credits Costs (A) (B) (C) (D) (E) (F) Zambia 72.6 16.8 4.0 93.4 9.1 84.3 Zaire 90.2 22.8 7.2 120.2 69.1 51.1 S. Africa/Naimibia 87.1 7.5 4.1 98.7 56.0 42.7 kustralia 70.5 7.3 0.7 78.6 50.9 27.7 P.N.G. 107.1 2.0 - 109.1 91.2 17.9 Philippines 84.4 7.1 9.9 101.4 44.1 57.3 Indonesia 61.2 2.7 - 63.9 30.6 33.3 Canada 130.6 22.3 5.6 158.5 168.1 -9.6 U.S.A. 95.7 6.2 4.0 105.9 32.5 73.4 Mexico 82.9 35.4 20.8 139.1 97.0 42.1 Chile 50.8 25.4 0.6 76.9 20.1 56.8 Peru 90.2 30.6 9.2 129.9 88.8 41.1 Sweden 210.4 22.7 4.6 237.7 222.0 15.7 Average 87.0 17.0 4.8 108.7 58.8 49.9 1984 /a Co/By- Direct Indirect Gross Product Net Costs Costs Interest Costs Credits Costs (A) (B) (C) (D) (E) (F) Zambia 53.3 13.0 7.5 74.8 7.8 67.0 Zaire 56.7 6.9 6.2 69.8 24.6 45.2 S. Africa/Namibia 69.3 5.0 4.4 78.7 33.1 45.6 Australia 91.0 7.3 3.1 103.9 37.6 66.3 P.N.G. 86.2 1.5 1.2 88.9 56.5 32.4 Philippines 78.0 5.4 9.4 92.8 37.3 55.5 Indonesia 63.1 1.5 1.0 65.6 19.6 46.0 Canada 150.5 17.2 14.6 182.2 126.2 56.0 U.S.A. 78.6 4.6 5.5 88.7 10.6 78.1 Mexico 63.3 25.0 21.2 109.5 71.6 37.9 Chile 39.5 15.4 5.0 59.8 11.1 48.7 Peru 84.9 19.8 7.5 112.2 55.4 56.8 Sweden 123.9 13.1 2.9 139.9 124.6 15.3 Average 75.0 11.6 7.2 93.4 36.9 56.7 /a Estimated. Source: Brook Hunt & Associates, Ltd. - 61 - Table 5.1 (B): PRODUCTION COSTS FOR MAJOR COPPER PRODUCING COUNTRIES (X) 1975 Co/By- Direct Indirect Gross Product Net Costs Costs Interest Costs Credits Costs (A) (B) (C) (D) (E) (F) Zambia 83 13 4 100 3 97 Zaire 63 34 3 100 19 81 S. Africa/Namibia 94 6 0 100 15 85 Australia 95 5 0 100 26 74 P.N.G. 96 2 2 100 51 49 Philippines 88 9 3 100 23 77 Indonesia 78 5 17 100 23 77 Canada 83 13 4 100 76 24 U.S.A. 87 7 6 100 10 90 Mexico 70 26 4 100 69 31 Chile 69 27 4 100 10 90 Peru 78 20 2 100 54 46 Sweden - - - - - - Average 81 15 4 100 31 69 1980 Co/By- Direct Indirect Gross Product Net Costs Costs Interest Costs Credits Costs (A) (B) (C) (D) (E) (F) Zambia 78 18 4 100 10 90 Zaire 75 19 6 100 57 43 S. Africa/Namibia 88 8 4 100 57 43 Australia 90 9 1 100 65 35 P.N.G. 98 2 0 100 84 16 Philippines 83 7 10 100 43 57 Indonesia 96 4 0 100 48 52 Canada 82 14 4 100 106 -6 U.S.A. 90 6 4 100 31 69 Mexico 60 25 15 100 70 30 Chile 66 33 1 100 26 74 Peru 69 24 7 100 68 32 Sweden 89 10 2 100 93 7 Average 80 16 4 100 54 46 1984 /a Co/By- Direct Indirect Gross Product Net Costs Costs Interest Casts Credits Costs (A) (B) (C) (D) (E) (F) Zambia 73 17 10 100 10 90 Zaire 81 10 9 100 35 65 S. Africa/Namibia 88 6 6 100 42 58 Australia 88 7 3 100 36 64 P.N.G. 97 2 1 100 64 36 Philippines 84 6 10 100 40 60 Indonesia 96 2 2 100 30 70 Canada 83 9 8 100 69 31 U.S.A. 89 5 6 100 12 88 Mexico 58 23 19 100 65 35 Chile 66 26 8 100 19 81 Peru 76 18 7 100 49 51 Sweden 89 9 2 100 89 11 Average 80 12 8 100 40 61 /a Estimated. Source: Brook Hunt i Associates, Ltd. - 62 - interest costs were Indonesia and Papua New Guinea and the countries with higher interest costs were Mexico, Canada and the Philippines. B. Impact of Co/by-Product Credits on Competitive Positions 110. For the majority of producers, co/by-product credits play an important role in reducing the production costs for copper. However, the co/by-products credits fluctuate widely due to the volatility of co/by-product prices. The important co/by-products for copper are gold, silver, lead, zinc, nickel, cobalt and molybdenum. Increases in co/by-product prices aided the financial performance of many producers during the late 1970s when co/by- product revenues increased from US$0.22 per lb (1975) to US$0.59 per lb (1980). Thus, in 1980, when the prices of these products were generally high, co/by-product credits accounted for about US¢59/lb, more than 50% of gross production costs on the average. By 1984, the prices of these co/by-products had declined substantially and their credits were worth only C36.9/lb or equivalent of 40% of gross production costs. 111. Table 5.2 shows the average market prices of major co/by-products along with the copper price in the 1978-84 period. Between 1980 and 1984, prices of most of these metals declined more sharply than the copper price, except for zinc and nickel. Nevertheless, co/by-product credits are still an important factor in reducing the average production cost of copper especially in Sweden, Canada, Mexico, Papua New Guinea, Peru and Zaire. Co/by-product credits are especially important for copper producers in Sweden, Canada and Zaire. In Canada, nickel, lead and zinc are the main co-products and in Sweden, lead, zinc and gold are the main coproducts. In Zaire, cobalt is the important byproduct. In contrast, producers in Zambia, Chile, Indonesia and the United States have relatively low co/byproduct credits. 112. The matrix presented in Figure 5.1 classifies the major copper producing countries, based on the gross cost and co/by-product credit dependency in 1984. The dividing lines chosen for gross costs and co/by- product credits are arbitrary. However, the matrix broadly characterizes each major copper producing country and its vulnerability to the changes in copper and co/by-product metal prices. 113. Zambia and the United States are Type I countries, where producers' gross costs are high and co/by-product credits are low. Therefore these producers' net incomes are most vulnerable to weakening of copper prices. From 1975 to 1980, when co/by-product prices increased strongly, net costs increased by 37% in Zambia and 19% in the United States--compared with 2% for the average of all countries. The majority of large copper producing countries are Type II countries where production is characterized by high gross costs and high co/by-product credits. The impact of a decline in their copper revenues could be eased by co/by-product revenues if the prices of co/by-products remain relatively high. Type III countries would be the most competitive producers with low average gross costs and high co/by-product credits. However, Zaire is the only country that falls into this category even marginally. Type IV countries are those with low gross costs and low Table 5.2: AVERAGE PRICES OF SELECTED METALS, 1978-1984 Ratio of Product Unit 1978 1979 1980 1981 1982 1983 1984 1984 to 1980 Copper (¢/lb) 61.8 90.1 99.3 79.0 67.2 72.2 62.7 0.63 Lead (t/lb) 29.8 54.5 41.2 33.0 24.7 19.3 20.2 0.49 Zinc (t/lb) 26.9 33.6 34.5 38.4 33.8 34.8 41.1 1.19 Gold ($/oz) 193.2 306.7 611.8 459.7 377.4 423.9 363.9 0.59 Silver ($/oz) 5.4 11.1 21.0 10.5 8.0 11.6 8.3 0.40 Nickel (¢/lb) 188.3 253.5 295.9 270.0 218.1 212.5 215.2 0.73 Cobalt (Wlb) 11.5 24.6 25.0 20.0 8.2 5.68 10.25 0.41 Molybdenum (t/lb) 4.5 7.8 9.8 8.5 4.1 3.64 3.62 0.37 Source: Metal Bulletin, various issues. Figure 5.1: CLASSIFICATION OF COPPER PRODUCERS BY COST CHARACTERISTICS AS OF 1984 190 - 180 - Cal 0 Canada 170 - . 160 - _ _ _ _ _ _ _ _ _ _ _ 1 50 - _ _ _ __ _ _ __ _ _ _ _ _ 140 140-(I) -_ = - (II) -=-=Sweden 130- 0, 9120 - U%Peu - 110 Indon s iaiaexi 6 0 100- 0 _____0 Philippines_ _ ___ _ _ USA PNG 80ma - ()frica 0 Zabia Za~ re 70 - - m Indon sia 60 - _ _ _ _ _ _ Chile 50 - _ _ _ _ _ _ _ _ (IV) (III)__ _ _ _ _ 40 - _ _ _ _ _ _ _ _ 0 20 40 60 80 100 120 By/Co-Product Credits (¢/lb) Source: Brook Hunt & Associates. - 65 - co/by-product credits. Indonesia and Chile fall into this category. These producers are the most resilient in the face of depressed copper and/or co/by- product prices. C. Impact of Cost Cutting Measures Undertaken by Producers since 1980 114. As mentioned earlier, a notable feature of the 1980-84 period was the substantial reductions in the average gross production costs for most of the major copper producing countries. Table 5.3 shows the changes in gross costs by country in this period, with countries shown in descending order according to the decrease in costs. Most of these cost reductions resulted from; (a) cost cutting measures; (b) cutbacks in and/or closures of uneconomic operations, and (c) changes in exchange rates especially vis-a-vis the U.S. dollar. Table 5.3: CHANGE IN GROSS COSTS BETWEEN 1980 AND 1984 Country Change in c/lb. Change in % 1. Sweden -97.8 -41.1 2. Zaire -50.4 -41.9 3. Mexico -29.6 -21.3 4. Papua New Guinea -20.2 -18.5 5. S. Africa -20.0 -20.3 6. Zambia -18.6 -19.9 7. Peru -17.7 -13.6 8. United States -17.2 -16.2 9. Chile -17.1 -22.2 10. Philippines -8.6 -8.5 11. Indonesia 1.7 2.7 12. Canada 23.7 14.9 13. Australia 25.3 32.2 Source: Table 5.1 (A) and (B). 115. During this 1980-84 period, strenuous efforts were made by all copper producers, especially those in North America, to contain or reduce their costs because of the depressed copper prices. Decreases in co/by-product prices exerted added pressure. Furthermore, a change in management outlook accelerated the producers efforts to rationalize operations and increase productivity. Traditionally, the copper industry considered itself a cyclical industry in which losses in the low copper price periods would be offset by profits in high copper price periods. The prolonged period of low prices, however, appears to have forced the producers to think of the industry as one in which the product prices are expected to be flat at low levels for an - 66 - extended period. Because copper is a homogeneous product, cost competitiveness is the only way to survive. Because product prices are determined by terminal free markets, a mere relative cost advantage over competitors in the copper industry does not guarantee that producers can cover their production costs. Therefore, in addition to striving for cost competitiveness, producers place increasing emphasis on reducing the absolute cost level. External factors such as declines in interest rates, energy prices and inflation rates, have also contributed to the decline in copper production costs. 116. Apart from cutbacks or shutdowns of uneconommic operations, the cost-reduction measures are of two kinds, i.e., those that have lasting effects on operating cost levels and those that would have only short-term effects and cannot be sustained on a longer term basis: (a) Measures with long-term effects include: o work force reductions; o reductions in overhead costs; o modernization and improvement of mining plans; o increased use of computer process control (such as computer dispatch systems for mine trains); o expansion of mine capacity (to achieve size economies); o utilization of futures markets to lock in prices; o opening of solvent extraction and/or electrowinning plants; and o abandoning smelters and refining stages of integrated operations and selling copper concentrates instead of smelted or refined copper. (b) Measures with only short-term effects include: o containing and reducing wage rates; o deferment of development expenditures; o reductions in waste stripping ratios; o high-grading. 117. The measures with only short-term, or temporary cost-reducing effects tend to have adverse long-term consequences both for cost levels and production capability. These kinds of measures have been taken in countries where foreign exchange shortages have constrained the availability of critical - 67 - imports (e.g., spare parts/supplies in Zambia), and also by North American producers that are economically marginal and/or in a weak financial position, in order to "buy time". Other independent copper producers in North America have relied more on measures that can be sustained. Kennecott Corp., before its acquisition by SOHIO, embarked on a vigorous modernization program at its Chino mine. St. Joe Minerals has modernized their copper and zinc operations and Phelps Dodge and INCO have drastically reduced work force levels. 118. Some major changes have occurred in the labor market, especially in the United States, where the copper industry has been a strike-prone industry. The wage negotiations in 1983 between the United Steelworkers of America and the major US copper producers were settled quickly with the workers accepting a wage freeze. The labor union, noting that attempts to obtain a significant wage hike or resorting to strike actions could threaten job security in the face of the growing copper imports and the deteriorated competitiveness of US producers, cooperated with management. Some producers, e.g., Phelps Dodge, obtained large concessions from their work force in the wage negotiations. Phelps Dodge's management successfully froze wages in its labor contract in 1983. 119. Several US producers decided to sell concentrates instead of smelted or refined copper to avoid relatively high costs of smelting and refining these concentrates themselves. They took advantage of the favorable terms being offered for custom smelting by Japanese smelters who were short of concentrate material. Thus, in 1982 US exports of copper concentrates amounted to nearly 200,000 tons compared with 107,000 tons and 151,000 tons in 1980 and 1981, respectively. Anaconda and Anamax contracts with the Japanese smelters were the main contributors to this increase in concentrate exports from the United States. Amoco (Cyprus Minerals) also shipped concentrates to Japan. 120. Another measure taken to reduce long-term costs was expansion of mining capacity, where it could benefit from economies of scale. Producers in Chile and Sweden resorted to this strategy. CODELCO's recent mine capacity expansions along with high capacity utilization seem to have strengthened its competitive position in the market. There are signs that some North American producers may also follow the capacity expansion strategy. For example, Kidd Creek in Canada hopes to reduce unit cost by increasing its copper smelter/refinery capacity to 90,000 tons a year from the current design capacity of 59,000 tons per year. 121. Treatment of tailings and dumped ore by leaching and electrowinning techniques is another way often used to expand production at low cost. Projects for processing of concentrator tailings accumulated from past operations have been implemented in Zambia and the United States to reduce the overall production costs and to offset the decline in output from the existing mines. The process produces copper without any mining cost, using less manpower. Thus, it enables producers to reduce their average cash production costs. - 68 - D. Currency Devaluations by Producing Countries 122. In the 1980-84 period, most copper producing countries experienced a significant depreciation of their currencies against the US dollar. The depreciation was a result of the appreciation of the U.S. dollar against other major currencies and the devaluations undertaken by the producing countries (or in Australia's case, the depreciation after floating the currency). The currency depreciations had the effect of reducing copper production costs in terms of US dollars in these countries. Table 5.4 shows the currency changes against the US dollar and local Consumer Price Index (CPI) inflation relative to US CPI inflation for major copper producing countries during the period. All the countries listed experienced depreciations more than sufficient to offset their differential inflation, as can be seen from Column (G). 123. The effects of exchange-rate adjustments on copper production costs depend on (i) the relative share of the local portion of production costs, (ii) the extent of local inflation relative to US inflation and (iii) the size of the currency devaluation. The higher the share of the local portion in the gross costs, the greater the cost-reducing leverage of a depreciation. 1/ Another factor is the rate of domestic inflation (relative to that in the United States), i.e., the extent to which the benefit derived from a currency devaluation is offset by domestic inflation. 2/ Even if the differential inflation exactly offsets the extent of depreciation, non-US producers may be able to take advantage of their currency devaluations to "buy time" to improve their competitive position and to maintain or expand their market shares. 124. For countries whose economies are highly dependent on primary commodity exports such as copper, depressed primary commodity prices tend to necessitate devaluation of their currencies. Improved cost competitiveness derived from the devaluation can lead to an increase in the volume of primary exports to the extent that unused production capacity is available. However, a devaluation can cause living standards to fall and can lead to politically negative consequences. 125. How have changes in the exchange rates vis-a-vis the US dollar affected the relative cost competitiveness position of each producing country? The effects of the changes in the exchange rates for the competitive positions of major copper producing countries can be seen from Tables 5.6, 5.7 and 5.8 which show the average net production cost (i.e., the cash cost after 1/ Table 5.5 shows the percentage share of the local cost portion in total gross costs for major copper producing countries. 2/ A'devaluation would have negative effects on the cost of production in local currency terms; a devaluation increases the local currency cost of imported goods and services and the local currency cost of repaying loans denominated in US dollars. These negative effects on the costs, of course, are more than offset by the increases in the price of the product in local currency terms. Table 5.4: EXCHANGE RATE CHANGES AND INFLATION IN MAJOR COPPER PRODUCING COUNTRIES, 1980-1984 Purchasing Ratio of Power Local CPI Parity Exchange Rate Inflation to Exchange Local Currency Per US$ US Inflation Rate Country 1980 1984 % Change 1980-1984 in 1984 /a (C) + (F) (A) (B) (C) (D) (E) (F) (G) Zambia 0.789 1.870 137.0 1.47 1.160 1.61 Zaire 2.800 37.100 1,225.0 4.02 11.256 3.30 S. Africa 0.779 1.570 101.5 1.33 1.036 1.52 Australia 0.878 1.190 35.5 1.10 0.966 1.23 PNG 0.671 0.929 38.5 1.04 0.698 1.33 Philippines 7.511 18.000 139.6 1.75 13.144 1.37 Indonesia 626.994 1,038.400 65.6 1.21 758.663 1.37 Canada 1.169 1.310 12.1 1.09 1.274 1.03 United States 1.000 1.000 - 1.00 1.000 1.00 Mexico 22.951 173.730 657.0 6.61 128.755 1.35 Chile 39.000 95.630 145.2 1.58 61.620 1.55 Peru 288.653 3,707.100 1,184.3 10.78 3,111.680 1.19 Sweden 4.230 8.410 98.8 1.13 4.780 1.76 /a The rate that would exactly offset the differential inflation rates between the country concerned and the United States in 1980-1984. Source: IMF, International Financial Statistics, various issues. - 70 - deducting by/co-product credits) for 1975, 1980 and 1985 for each country as estimated on the basis of three different exchange rates. Table 5.6 ranks countries on the basis of the net production cost at actual exchange rates. 126. Table 5.7 shows the country ranking, based on the net production cost calculated on the assumption that the exchange rates prevailing in 1980 could be considered to represent "equilibrium" rates; for 1975 and 1984, for each country the exchange rate is adjusted to "exactly offset" the differential inflation between the country concerned, on the one hand, and the United States, on the other, on the basis of the naive purchasing power parity formula. Similarly, Table 5.8 shows the ranking of countries by net production cost as calculated on the basis of the exchange rates derived from the assumption that the exchange rates prevailing in 1984 could be considered to represent "equilibrium" rates; the hypothetical exchange rates for 1975 and 1980 are derived on the basis of the naive purchasing power parity principle Table 5.5: ESTIMATED SHARE OF LOCAL CONTENT IN GROSS PRODUCTION COSTS IN MAJOR COPPER PRODUCING COUNTRIES, /a 1975, 1980 AND 1984 Country 1975 1980 1984 ----------- percents--------- United States 100 100 100 Sweden - 90 90 Canada 80 80 80 Australia 80 80 80 Chile 70 70 65 South Africa/Namibia 70 70 65 Zambia 65 65 60 Zaire 65 65 60 Philippines 50 50 60 Peru 65 65 60 Mexico 60 60 55 Papua New Guniea 50 50 50 Indonesia 50 50 50 /a Smelting and refining charges for producers which ship concentrates are considered as foreign components. Source: The World Bank, Industry Department to "exactly offset" the differential inflation rates in the 1975-84 and 1980- 84 periods between the United States and the country concerned. - 71 - Table 5.6: RANKING OF MAJOR COPPER PRODUCING COUNTRIES BASED ON ACTUAL NET COSTS, 1975, 1980 AND 1984 1975 1980 1984 Country ¢/lb Country ¢lb Country ¢/lb PNG 23.8 Canada -9.6 Sweden 15.3 Mexico 27.3 Sweden 15.7 PNG 32.4 Canada 28.4 PNG 17.9 Mexico 37.9 Indonesia 35.5 Australia 27.7 Zaire 45.2 Philippines 38.1 Indonesia 33.3 S.Africa /a 45.6 Australia 38.3 Peru 41.1 Indonesia 46.0 S.Africa /a 41.3 Mexico 42.1 Chile 48.7 Chile 47.2 S.Africa /a 42.7 Philippines 55.5 /b 48.9 /b 49.9 Canada 56.0 Peru 51.1 Zaire 51.1 /b 56.7 Zaire 55.1 Chile 56.8 Peru 56.8 Zambia 61.6 Philippines 57.3 Australia 66.3 United States 61.6 United States 73.4 Zambia 67.0 Sweden ---- Zambia 84.3 United States 78.1 /a South Africa plus Namibia. /b Weighted averages. Source: The World Bank, Industry Department, based on data provided by Brook Hunt Associates. - 72 - Table 5.7: RANKING OF MAJOR COPPER PRODUCING COUNTRIES BY NET COSTS AS ADJUSTED USING THE 1980 EXCHANGE RATES AS EQUILIBRIUM RATES, 1975, 1980 AND 1984 1975 1980 1984 Country {TIlb Country f/lb Country j/lb Canada 15.2 Canada -9.6 PNG 46.4 Mexico 23.9 Sweden 15.7 Mexico 55.0 PNG 24.7 PNG 17.9 Indonesia 58.1 Peru 30.0 Australia 27.7 Canada 60.2 Indonesia 33.2 Indonesia 33.3 Chile 68.6 Australia 35.9 Peru 41.1 Peru 68.9 Philippines 40.0 Mexico 42.1 South Africa /a 70.9 South Africa /a 44.1 South Africa /a 42.7 Philippines 74.1 /b 50.6 /b 49.9 /b 77.4 Chile 53.1 Zaire 51.1 United States 78.1 United States 61.6 Chile 56.8 Australia 84.2 Zambia 64.9 Philippines 57.3 Zambia 92.0 Zaire 83.4 United States 73.4 Sweden 108.2 Sweden ---- Zambia 84.3 Zaire 132.9 /a South Africa plus Namibia. /b Weighted average. Source: The World Bank, Industry Department, based on the data provided by Brook Hunt and Associates. Table 5.8: RANKING OF MAJOR COPPER PRODUCING COUNTRIES NET COSTS AS ADJUSTED BY USING THE 1984 EXCHANGE RATES AS EQUILIBRIUM RATES, 1975, 1980 and 1984 1975 1980 1984 Country :/1b Country jr/lb Country 7/lb Mexico 11.7 Sweden -74.5 Sweden 15.3 Canada 13.7 Canada -13.2 PNG 32.4 PNG 18.9 Zaire -0.1 Mexico 37.9 Peru 22.2 PNG 4.8 Zaire 45.2 Australia 28.6 Australia 16.0 South Africa /a 45.6 Indonesia 28.7 South Africa /a 20.0 Indonesia 46.0 South Africa /a 31.6 Mexico 23.7 Chile 48.7 Zaire 32.6 Indonesia 24.6 Philippines 55.5 Philippines 33.0 Peru 28.6 Canada 56.0 /b 39.4 /b 34.4 /b 56.7 Chile 40.0 Chile 37.7 Peru 56.8 Zambia 48.6 Philippines 44.9 Australia 66.3 United States 61.6 Zambia 62.3 Zambia 67.0 Sweden ---- United States 73.4 United States 78.1 /a South Africa plus Namibia. /b Weighted average. Source: The World Bank, Industry Department, based on the data provided by Brook Hunt and Associates. - 73 - 127. The choice of the years 1980 and 1984 as two alternative "base" years is convenient because during the 1975-1984 period 1980 was the year when the US dollar was at its weakest point vis-a-vis other major currencies and 1984 was the year when the US dollar was at its strongest point against other major currencies. It should be noted that, given the debt crisis, the exchange rate relationships of 1984 (greatly influenced for many countries by the need to generate foreign exchange surplus) are probably more indicative of future relationships than the years of 1975 and 1980 for the countries in Tables 5.6-5.8. The precise method used to calculate the adjusted net production costs presented in Tables 5.6, 5.7 and 5.8 is explained in ANNEX B. 128. Comparison of Tables 5.6, 5.7 and 5.8 highlights the strong influence that exchange rate adjustments have had on the level of production cost in each country over the last decade. For example, the actual average net production cost of Papua New Guniea in 1980 was 3.7 times the level of its hypothetical net production costs in 1980 as calculated on the basis of the adjusted exchange rate using the 1984 exchange rate as the equilibrium base rate. Similarly, the actual net production cost for Zaire in 1984 was only 34% of the level of its hypothetical net production costs as calculated for 1984 using the 1980 exchange rate as the equilibrium base rate. 129. A comparison of the three tables shows the following: (a) For the majority of countries selected for comparison, the relative competitive position is rather stable. (b) The United States and Zambia both fall into the above-average- costs category in all three cases. They are Type-I countries as defined in Figure 5.1, i.e., those most vulnerable to declines in copper prices. (c) For 1984, Chile's competitiveness ranking is even higher in the adjusted 1980 exchange rate case (the "weakest dollar" case) than in the adjusted 1984 exchange rate case (the "strongest dollar" case) and also higher than in the actual costs case. Chile's competitive position has been strong whichever year is chosen as the base year for the "equilibrium" exchange rate. (d) Chile's competitive position improved considerably between 1980 and 1984 in all cases. Chile's net cost position, which was near the world average in 1975 came down to well below average by 1980. By 1984, Chile's net cost position improved further, moving into the lowest-cost producer category. The fact that by/co-product prices were exceptionally high in 1980 - 74 - worked against Chile in terms of relative net cost ranking in that year. 1/ (e) Papua New Guinea, Indonesia and Mexico maintained strong competitive positions throughout the 1975-84 period, regardless of the exchange rate adjustments used. These countries have had sufficiently large co/by-product credits to remain competitive in the face of depressed copper prices. (f) Exchange rate adjustments made sharp differences for Sweden and Zaire. With its high local content (90%), Sweden was able to reap almost full benefits (or very large cost reductions) from the sharp devaluation of its currency between 1980 and 1984. For Zaire, although the local portion of total net costs is in the range of only 60-65%, the drastic devaluation of its currency (i.e., 1,225%) during the 1980-84 period dramatically improved its competitive position in US dollar terms. E. Trends in Production Costs in 1985 2/ 130. Preliminary cost estimates for 1985 indicate an impressive reduction in copper production costs. Table 5.10 shows the production cost estimates for 1985 and Table 5.11 shows the changes in production costs between 1984 and 1985. In 1985 average gross cost appears to have declined by 15¢/lb to 78¢/lb from 93¢/lb in 1984. This is a 16% reduction in production costs. Of the 15¢/lb reduction in average cost, 70% is attributable to the reduction in direct costs. Average direct costs fell llC/lb to 64¢/lb in 1985 from 75¢/lb in 1984 or 15% reduction. Average indirect costs declined by 40% (!) contributing about 30% to the gross cost reduction. The countries where gross costs declined by more than 15¢/lb (average) are Canada, Australia, Peru, South Africa/Nambia, Sweden and the United States. Of these countries, Peru, South Africa and Australia are the countries where the production cost improvements resulted mainly from the devaluations of their currencies against the US dollar. The improvement in gross costs in Canada, Sweden and the United States, came mainly from the producers' aggressive efforts to rationalize their operations including steps for streamlining the labor force, modernizing production facilities and new investments in solvent extraction processes to produce low cost copper, as mentioned in an earlier section. 1/ Chile's ranking in terms of gross cost measured at actual exchange rates remained strong over the 1975-84 period (Table 5.9). Chile was the second lowest cost producer after Indonesia in terms of gross cost. 2/ The preceding analysis in Chapter V relates to estimated data for 1984 because actual cost data for 1984 and preliminary estimates for 1985 became available only at the final stages of the study. Table 5.9: RANKING OF MAJOR COPPER PRODUCING COUNTRIES BASED ON ACTUAL GROSS COSTS, /a 1975, 1980 AND 1984 1975 1980 1984 Country ¢/lb Country ¢/lb Country ¢Ilb Indonesia 46.2 Indonesia 63.9 Chile 59.8 PNG 48.3 Chile 76.9 Indonesia 65.6 South Africa /b 48.8 Australia 78.6 Zaire 69.8 Philippines 49.8 Zambia 93.4 Zambia 74.8 Australia 51.7 South Africa /b 98.7 South Africa /b 78.7 Chile 52.7 Philippines 101.4 United States 88.7 Zambia 63.2 United States 105.9 PNG 88.9 Zaire 67.8 /c 108.7 Philippines 92.8 United States 68.6 PNG 109.1 /c 93.4 /c 70.9 Zaire 120.2 Australia 103.9 Mexico 87.0 Peru 129.9 Mexico 104.5 Peru 111.6 Mexico 139.1 Peru 112.2 Canada 118.7 Canada 158.5 Sweden 139.9 Sweden ----- Sweden 237.7 Canada 182.2 /a Gross operating costs, i.e., before deducting co/by-product credits. lb South Africa plus Namibia. 7 Weighted average. Source: Brook Hunt Associates. - 76 - 131. In contrast, in the Philippines and Mexico gross production costs rose in 1985. The Philippines showed a 30% deterioration in gross costs in 1985, while Mexico showed a 10% deterioration. In both countries, increases in interest payments were the common factor for deterioration. The increase in interest (payments 20C/lb) contributed 85-135% of the increases in production costs in the Philippines and Mexico. In all other countries interest payments declined. In the Philippines, interest payments increased over 200% in 1985 while interest payments rose 85% in Mexico. The copper industry in these countries may have suffered from the debt crisis environment. 132. The net production cost determines the competitiveness of the producer in the industry. The managements have very little control over co/by-product credits. In 1985, all countries except Zambia and Zaire showed a decline in co/by-product credits. Average co/by-product credits declined 9C/lb, or 25% in 1985. The decrease of 15% in average gross costs was thus offset substantially by decreases in co/by-product credits, leaving only a 6% improvement in the average net production cost, which decreased to 51/lb in 1985 from 570/lb in 1984. The countries which experienced the largest declines in co/by-product credits are Sweden, Canada, Mexico, PCN, Australia and Peru. For PNG and Sweden the improvements in gross production costs were offset by the decline in co/by-product credits, resulting in deterioration in their net production costs. As for the Philippines and Mexico, the decline in co/by-product credits and the deterioration in their gross production costs coincided to worsen their relative competitive positions. 133. Table 5.12 shows the ranking of the copper producing countries in 1985 on gross cost basis while the ranking based on net costs is shown in Table 5.10. The gross cost ranking among major copper producing countries did not change much between 1984 and 1985, except that PNG and the Philippines slipped into the higher-than-average group while Australia barely climbed up to the lower-than-average group. The net cost ranking in 1985 shows significant changes from 1984. Sweden, which is highly dependent on co/by- product credits, slipped from the lowest cost position to the higher-than- average position. PNG slipped down to sixth place among the 13 countries. The common factor for the worsening of the competitive position of Sweden and PNC is the massive declines in co/by-product credits which exceeded the decreases in their gross costs as mentioned above. The Philippines and Mexico slipped down to the highest net cost positions as the decline in co/by-product credit coincided with the increases in their gross costs. Their cost competitive positions are now even below the United States. Peru's relative net cost competitiveness improved mainly because of the large reduction in gross costs which in turn resulted from the devaluation of its currency. Canada maintained almost the same position as in 1984, by compensating the sharp decline in co/by-product credits with large improvement in gross cost. 134. Zambia and the United States remained in the below-average cost competitive group despite the substantial improvements in their net costs and gross costs from the previous year. In contrast, Chile maintained a strong competitive position in net costs as well as in gross costs. Chile achieved a 15% gross cost improvement in 1985 which was fully reflected in the net cost improvement as the impact of a decline in co/by-product credits was unimportant to Chile. - 77 - Table 5.10: ESTIMATED PRODUCTION COSTS IN MAJOR COPPER PRODUCING COUNTRIES IN 1985 /a (USc/lb) Direct Indirect Interest Gross By/co-product Net Country Costs Costs Expenses Costs Credits Costs So. Africa/ Namibia 49.3 3.4 2.1 54.8 26.2 28.6 Zaire 58.8 5.2 1.7 65.7 25.9 39.8 Peru 66.6 12.4 6.4 85.4 44.2 41.2 Chile 40.2 6.3 4.5 51.0 8.8 42.2 Canada 107.7 11.6 8.9 128.2 85.9 42.3 Papua New Guinea 78.3 1.2 0.8 80.3 37.1 43.2 Indonesia 62.9 2.1 0.2 65.2 15.5 49.7 Australia 69.4 5.9 1.8 77.1 25.2 51.9 Sweden 105.9 10.4 2.8 119.1 65.1 54.0 Zambia 46.3 13.0 6.5 65.8 10.0 55.8 United States 65.5 3.1 3.3 71.9 6.6 65.3 Mexico 63.7 19.9 39.2 122.7 43.2 79.5 Philippines 80.6 7.4 32.2 120.2 34.3 85.9 Weighted Average 64.3 7.4 6.6 78.2 27.7 50.6 /a Ranked in ascending order by net costs. Source: Brook Hunt Associates, Ltd., Western World Copper Costs, 1985/86 Edition, London, 1986. - 78 - Table 5.11: CHANGES IN PRODUCTION COSTS BETWEEN 1984 AND 1985 (US¢/lb) Direct Indirect Interest Gross By/co-product Net Country Costs Costs Expenses Costs Credits Costs Zambia -7.0 0.0 -1.0 -9.0 2.2 -11.2 Zaire 2.1 -1.7 -4.5 -4.1 1.3 -5.4 S. Africa/ Namibia -20.0 -1.6 -2.3 -23.9 -6.9 -17.0 Australia -21.6 -1.4 -1.3 -26.8 -12.4 -14.4 Papua, New Guinea -7.9 -0.3 -0.4 -8.6 -19.4 10.8 Philippines 2.6 2.0 22.8 27.4 -3.0 30.4 Indonesia -0.2 0.6 -0.8 -0.4 -4.1 3.7 Canada -42.8 -5.6 -5.7 -54.0 -40.3 -13.7 United States -13.1 -1.5 -2.2 -16.8 -4.0 -12.8 Mexico 0.4 -5.1 18.0 13.2 -28.4 41.6 Chile 0.7 -9.1 -0.5 -8.8 -2.3 -6.5 Peru -18.3 -7.4 -1.1 -26.8 -11.2 -15.6 Sweden -18.0 -2.7 -0.1 -20.8 -59.5 38.7 Weighted Average -10.7 -4.2 -0.6 -15.2 -9.2 -6.1 Source: Brook Hunt Associates Ltd., Western World Copper Costs, 1985/86 Edition, London, 1986. - 79 - Table 5.12: RANKING OF MAJOR COPPER PRODUCING COUNTRIES GROSS COST BASIS IN 1985 (US¢/lb) Direct Indirect Interest Gross Country Cost Cost Payments Costs Chile 40.2 6.3 4.5 51.0 South Africa/Namibia 49.3 3.4 2.1 54.8 Indonesia 62.9 2.1 0.2 65.2 Zaire 58.8 5.2 1.7 65.7 Zambia 46.3 13.0 6.5 65.8 United States 65.5 3.1 3.3 71.9 Australia 69.4 5.9 1.8 77.1 Papua, New Guinea 78.3 1.2 0.8 80.3 Peru 66.6 12.4 6.4 85.4 Sweden 105.9 10.4 2.8 119.1 Philippines 80.6 7.4 32.2 120.2 Mexico 63.7 19.9 39.2 122.7 Canada 107.7 11.6 8.9 128.2 Source: Brook Hunt Associates, Ltd., Western World Copper Costs, 1985/86 Edition, London, 1986 - 80 - VI. IMPACT OF CHANGES IN INDUSTRY ORGANIZATION ON SUPPLY 135. The purpose of this chapter is to review the changes during the last two decades in the ownership structure of the copper industry and to determine whether these changes have altered the ways the industry responds to the changing demand-supply balance. The chapter investigates whether or not increased state participation and increased ownership by energy-related companies have led to misplaced investments in capacity increases and uneconomic production in the face of depressed market conditions. A. Changes in Ownership and Control 136. Copper producers can be classified according to ownership and the extent of vertical integration. In terms of ownership, it is often considered useful to distinguish between state-owned or state-controlled companies and privately owned and operated companies. Privately owned companies can be divided into independent producers and those owned by multinational energy companies. These three groups of copper producers--state, private and multinational--can be classified further by the stages of production they engage in, namely, mining only, smelting and refining only and companies that intergrate all three activities. 137. Until the early 1960s, a large part of copper production was in the hands of a few multinational mining corporations that were mostly integrated, either domestically or internationally. The industry has become more diversified since that time. The change has occurred as a result of three major trends, namely, increased state participation in the ownership (or control) of copper producing activities (especially in Latin America and Africa), acquisition of major copper producing companies by energy-related multinational companies (especially in North America) and the increased importance of non-integrated mining companies in world mine production of copper. The last mentioned trend has occurred because demand for copper concentrates has increased as a result of the rapid capacity expansion of companies with smelting and refining capacities only, especially in Japan and Western Europe and lately in the Republic of Korea. 138. For a variety of historical, policial, and economic reasons, state participation in the copper industry increased significantly up to the - 81 - early 1970s. 1/ According to Sir Ronald Prain, at the beginning of the 1960s copper production in which governments held any sort of interest amounted to less than 100,000 tons a year, or 2.5% of capacity in the market economies; but by 1970 the total had risen to some 2.25 million tons, or about 43% of capacity. "[By 1970] more than a quarter of the world's copper was being produced by mines totally owned by government, 12 percent by companies in which the state had a majority interest, and 5 percent by companies in which government had minority interest." 2/ It appears that the trend towards state participation has slowed since the early 1970s. According to Marian Radetzki, as of 1981, 40.5% of copper mining capacity in the market economies had 5% or more equity holdings by governments (Table 6.1). 3/ 139. The first major step towards state participation in the copper industry was the complete nationalization of Gecamines properties in Zaire in 1967. This was followed by government purchases of a 51% interest in all mining properties in Zambia and in the large copper mines in Chile (CODELCO) in 1969. CODELCO subsequently (in 1971) took over the remainder of the interests in the Chilean mines. In 1974, the Peruvian Government nationalized the Cerro de Pasco mines, which have since been operated as a state enterprise (Centromin Peru S.A.). Also, a new state enterprise, Mineroperu, started copper production at Cerro Verde in Peru in 1977. In 1979, the Zambian Government-s equity shares in its copper mines were increased to 60%. Finally, in 1980, the new La Caridad mine of Mexicana de Cobre in Mexico started production. The company is 44% owned by the Mexican Government. 4/ 1/ For a discussion of the definitional problems and a survey of motivations for state participation in mineral industries, see Marian Radetzki, State Mineral Enterprises: An Investigation into Their Impact on International Mineral Market (Washington, D.C., 1985) Chapters 2 and 3. Radetzki's definition of a state-owned company is too broad for our purpose here. For the purpose of this paper, state-owned company is arbitrarily defined as a company with state ownership of about 40% or higher (as against Radetzki's 5% or higher). The difficulties relating to the definition of state-owned copper company can be illustrated by a reference to Kennecott Corp., which was taken over by a petroleum company, SOHIO, which in turn is owned and controlled by a state enterprise, British Petroleum. The precise definition of a state-owned copper company is not essential for the purpose of this paper. 2/ Sir Ronald Prain, op. cit., pp. 222-3. 3/ Radetzki, op. cit. 4/ Other copper companies which have state-participation are either small to medium-sized producers or large producers where state equity interests are rather small. All copper properties in Yugoslavia have long been government-owned (total capacity 100,000 tpy). - 82 - Table 6.1: GOVERNMENT OWNERSHIP POSITION IN THE COPPER INDUSTRY, 1981 /a Copper (Metal Content) Units Mining Smelting Refining Western World Total capacity (-000 tons) of which 7,820 8,780 9,120 Capacity with significant government ownership (Z) /b 40.5 30.6 25.9 Capacity proportional to government equity holding (%) 32.4 26.1 21.6 Capacity with majority government ownership (X) 34.7 29.7 24.3 Developing Countries Total capacity ('000 tons) of which 4,120 3,340 2,580 Capacity with significant government ownership (X) /b 73.0 75.5 82.6 Capacity proportional to government equity holding (%) 57.8 64.0 67.6 Capacity with majority government ownership (%) 62.0 72.9 77.0 /a Units are indicated, item by item. lb Defined as 5% and upwards of total equity. Source: Marian Radetzki, State Mineral Enterprises, Washington, D.C., 1985. - 83 - 140. The participation of energy (mainly petroleum) companies in the copper sector took place primarily in the 1976-82 period and was concentrated mostly in the United States. As of 1982, an estimated one-fifth of the market economies' copper mining capacity was directly controlled by these corporations. 1/ 141. Oil companies, especially international oil companies, were interested in diversifying away from petroleum long before the mid-1970s. Even in the 1960s, some oil companies showed interest in diversification into non-oil sectors (e.g., coal and uranium). The primary motivation was the anticipation of the growing limitations to their freedom of action in the petroleum sector arising from host government actions in oil producing countries. 142. As the steep increases in oil prices around 1973-74 led to sharp increases in profits, oil companies intensified the search for investment opportunities and began to regard the non-energy minerals sector as a potential area for diversification. The copper industry became a prime candidate for diversification. The prospects for copper seemed good at the time 2/ and the technology and marketing involved seemed not too difficult to learn. 143. The oil companies used four different approaches: (i) grass roots exploration; (ii) purchase of part or all of undeveloped, partially explored, claims; (iii) purchase of part or all of proven ore bodies at the development stage; and (iv) purchase of producing properties or companies. Texasgulf (a sulfur producer) took the first approach and discovered the Timmins deposit which subsequently became Texasgulf's most valuable asset. However, while the payoff with the first approach could be large, the lag to payoff was likely to be long. Given the market outlook at the time, the cost of acquiring existing producing assets seemed to be a bargain. Therefore, the majority of oil companies concentrated on the last three approaches. Examples of the third approach were: o Billiton (Royal Dutch Shell's subsidiary) bought into the Southern Peru Copper Corporation's development of the Cuajone project in Peru; and o British Petroleum bought into the Olympic Dam project in Australia. 1/ M.D. Fitzgerald and Gerald Pollio, Copper: The Next Twenty Years, Chemical Bank, New York, November 1981, p. 6. 2/ In the mid-1970s, the expectation that the copper market should recover from the depressed, oversupply conditions once the world economy recovers the momentum of the 1960s was prevalent. The "Club of Rome" concerns for supply of exhaustible resources were generally accepted. - 84 - 144. Major acquisitions of copper producing companies and properties (approach iv) by energy companies in this period were as follows: 1/ Acquiring Company Acquired Company Year Cities Service Tennessee Copper 1963 Cities Service Miami Copper 1963 Pennzoil Duval 1968 Shell Billiton 1974 Arco Anaconda 1977 Louisiana Copper Range 1977 Exxon Disputada 1978 Hudson Bay Inspiration 1979 Amoco Cyprus 1979 Superior Oil Falconbridge 1980 British Petroleum Selection Trust 1980 Sohio Kennecott 1981 Occidental Cities Service 1982 145. Table 6.2 shows the relative importance of state-owned and privately-owned copper companies in the total mine production of the 50 largest mines over the 1975-83 period. About 48-49% of total mine production was produced by state-owned companies. This share did not change much in the 1975-83 period. Production by privately-owned mines is broken down into oil- company-owned mines and privately-owned independent mines (see the note attached to the table for important qualifications). As of 1983, oil-company- owned mines accounted for about 10% of the total production of the 50 largest mines. 146. The rapid growth of the copper smelting and refining industry in Japan and, to a lesser extent, in the Federal Republic of Germany, in the 1960s and the early 1970s played an important role in changing the organization of the world copper industry. In the early 1960s, with copper consumption rising rapidly, the governments of Japan and F.R. Germany became painfully aware of their lack of domestic copper resources. They decided to offer financial assistance to their growing copper smelting/refining industries to secure overseas sources of supply of ores and concentrates. Copper smelter production in Japan grew from 187,000 tons in 1960 to 1,000,000 tons in 1973, and in F.R. Germany it increased from 62,000 tons in 1960 to 233,000 tons in 1973. As a result, imports of ores and concentrates in 1973 amounted to 832,000 tons (copper content) in Japan and 134,000 tons in F.R. Germany. Of the total world imports of ores and concentrates in 1973, Japan and F.R. Germany accounted for 68% and 13%, respectively. 147. In the 1960-73 period, Japanese and German copper smelting companies actively sought to secure new sources of ores/concentrates supply by offering 1/ A detailed, company-by-company account of the oil companies' involvement in the copper sector is available in CRU, Copper Studies, October 1984. - 85 - Table 6.2: MINE PRODUCTION OF COPPER BY STATE-OWNED AND PRIVATELY-OWNED COMPANIES, 1975, 1980 and 1983 /a ('000 tons) 1975 1980 1983 State Owned - 12 mines 1,982 2,217 2,288 Privately Owned 2,184 2,410 2,393 of which: Oil Company Owned - 10 mines 474 543 489 Independent - 31 mines 1,710 1,867 1,904 Total of Above 4,166 4,627 4,681 Market Economies Total 5,731 6,042 6,271 /a Output of 50 largest mines is identified by company from the Brook Hunt Associates data. This accounts for only 73-77% of market economies' production. For example, state owned mines in Yugoslavia, India and Burma are not included, nor are privately owned mines in Japan, Malaysia, Norway, Namibia, etc. just to name the important omissions. Another important qualification is that the oil-company-owned 10 mines were identified as such on the basis of the ownership as of 1983. Obviously, a substantially smaller number of these mines was owned by oil companies in 1975. Source: Brook Hunt Associates. World Bureau of Metal Statistics. - 86 - long-term purchase contracts and attractive financing facilities to potential new sources of supply. This was done aggressively because most of the traditional sources of supply were more or less securely tied to the established multinational copper companies based in Europe and the United States. As a result of the Japanese and (to a lesser extent) German smelting companies' aggressive search, many unintegrated mines 1/ were opened (with financial assistance from official financing agencies). 2/ These aggressive policies of Japanese and German copper smelting companies, may have had the effect of aggravating the chronic overcapacity in copper mining in the mid- 1970s. 3/ B. Producers' Investment Decisions 1974-84 148. This section focusses on how the recent changes in ownership structure of the copper industry have affected the pattern of capacity expansion (and reduction). The section starts with a review of countries/companies that expanded capacity in the midst of the chronic stagnation period. 149. As discussed in Chapter IV, Section C above, the copper industry suffered from chronic overcapacity since 1975. Even in this period, world mine capacity for copper expanded. Market economies' mine capacity increased from 6.61 million tons per year (tpy) as of the beginning of 1974 to 7.45 million tpy at the beginning of 1983, (although it declined somewhat during 1983-84 because of numerous permanent closures in North America). 4/ 150. Mine capacity expanded rapidly in the 1971-77 period because of investment decisions made in the late 1960s and early 1970s when copper prices were relatively high and the industry enjoyed abundant profits. Although the extended period of relatively high copper prices (1964-1973) ended in the first half of 1974, the market economies' mine capacity continued to expand 1/ Mines without their own smelting/refining facilities. 2/ The Export-Import Bank of Japan, for example, approved generous financing for development of the following copper mines: Toledo (Atlas, Philippines, 1970), Mamut (Malaysia, 1975), Valley Copper Bethlehem, Canada, 1964), Granile (Canada, 1965), White Horse (Canada, 1966), Brenda (Canada, 1968), Lornex (Canada, 1970), Ertzberg (Freeport Indonesia, 1970), Sagasca (Chile, 1969), Musoshi (Zaire, 1969), Gunpowder (Mammoth, Australia, 1975), Tennant Creek (Australia, 1979), and Bougainville (PNG, 1970). In F.R. Germany, Kreditanstalt fur Wiederaufbau provided similar financing for copper mine development overseas. 3/ ANNEX C describes the differences between the market for refined copper and the market for concentrates and blister, and explains why custom smelters are concerned with the long-term security of concentrate supply. 4/ See Table 6.3 - 87 - Table 6.3: COPPER MINING CAPACITY AT BEGINNING OF YEAR, SELECTED YEARS, 1970-1983 ('000 metic tons per year) Country or Net Increases Region 1970 1974 1977 1980 1983 1974-77 1977-83 United States 1,569 1,715 1,787 1,701 1,716 73 -71 Canada 621 930 898 880 907 -32 9 Australia 136 227 249 245 263 23 14 West Europe 122 172 195 195 187 23 -8 Japan 127 91 82 68 59 -9 -23 Total Industrial 2,576 3,134 3,211 3,089 3,133 77 -79 Mexico 64 82 100 272 272 18 172 Chile 744 862 907 1,075 1,184 45 277 Peru 218 222 376 399 404 154 27 Other Latin Amer. 33 23 23 18 54 0 32 Zambia 758 771 771 635 608 0 -163 Zaire 363 522 567 499 526 45 -41 S. Africa 132 191 181 218 222 -9 41 Namibia 33 36 54 41 52 18 -3 Other Africa 35 77 82 73 91 5 9 Philippines 136 227 259 372 353 32 94 Indonesia 0 45 64 64 73 18 9 Malaysia 0 30 27 27 27 -3 0 Other Asia 12 27 50 45 64 23 14 P.N. Guinea 0 181 181 172 185 0 4 Yugoslavia 91 132 154 145 168 23 14 Other S. Europe 79 52 50 32 36 -2 -14 Total Develop- ing /a 2,695 3,479 3,847 4,087 4,318 367 472 Market Economies 5,271 6,613 7,058 7,176 7,451 445 393 China /b 120 150 170 177 187 20 17 World Ex CPEs 5,391 6,763 7,228 7,353 7,638 465 410 /a Excluding China and CPE Asia. 7W Including other non-market economies in Asia. Source: Phelps Dodge Corporation and the World. Bank, EPDCS. - 88 - until 1977, albeit at a much slower pace. Between the beginning of 1974 and the beginning of 1977, net capacity additions in the market economies amounted to 445,000 tpy (Table 6.3). 151. About a dozen countries "contributed" toward the total net capacity increases in the period of 1971-1977. Most of these additions were made by privately-owned companies (often owned, at least partly, by foreign multinationals). By far the largest new mine completed in this period was the Cuajone mine of the Southern Peru Copper Corporation (SPCC) in Peru, with 154,000 tpy of rated capacity. SPCC is a 100% foreign-owned private mining company that had been operating another large-scale open-pit copper mine at Toquepala, Peru. The ownership included ASARCO (51.5%), Newmont Mining (10.25%), Phelps Dodge (16.0%) and Cerro Corporation (22.25%). 1/ 152. After Peru, the United States was the second largest contributor with net capacity additions of 73,000 tpy, all of which were undertaken by private corporations. The additions included, among others, Cities Service's Pinto Valley (57,000 tpy in 1974), Anamax's Twin Buttes (41,000 tpy in 1975), Phelps Dodge's Metcalf (55,000 tpy in 1975), and Hecla's Lake Shore (62,000 tpy in 1975). 2/ Most of the capacity additions elsewhere were also made by companies largely owned and operated privately (often with majority participation by foreign multinationals) in such countries as Australia, Zaire (the Sodimiza mine, owned and operated by a Japanese consortium led by Nippon Mining), Namibia, Philippines, Indonesia, Sweden and Norway. 153. The only net additions that could be considered to have been made by companies with significant government participation were 45,000 tpy in Chile, 18,000 tpy each in India and Mexico, and 22,000 tpy in Yugoslavia. Therefore, insofar as the capacity expansion in the 1974-77 period is concerned, only about 100,000 tpy out of the total net additions of 445,000 tpy could be traced to decisions made with significant state participation. 154. Between the beginning of 1977 and the beginning of 1983, market economies' mine capacity increased by almost 400,000 tpy (Table 6.3). The largest additions were in Chile and Mexico, which together accounted for 450,000 tpy of net additional capacity. In both countries, state-owned companies undertook most of the expansions, namely, CODELCO and Mexicana de 1/ Custom smelters in the United States, Japan and Western Europe played an important role in the decision to develop the Cuajone mine. IFC also played a role by offering a part of the total financing needed. The equity held by Cerro Corporation was later acquired by Billiton. 2/ US capacity increased by 145,000 tpy in 1974-75, but declined in 1976 as a result of ore depletions of old mines, according to the Phelps Dodge data. Other sources including the USBM data show different trends. - 89 - Cobre. 1/ Net additions in Chile, Mexico, Peru, Yugoslavia and Burma, considered to be state-ownership-sponsored, totalled 508,000 tpy. Other additions, considered private-enterprise-sponsored expansions, were located in Canada, Australia, Brazil, South Africa, Philippines and Indonesia--in total about 200,000 tpy. The development of new smelters in the Philippines and Korea influenced the mine production capacity growth around the world. They increased the demand for concentrates and encouraged the development of new supplies for Korea and for Japanese smelters who needed to replace Philippine mine production that was diverted to the new smelters there. 155. The capacity expansions of CODELCO came from expansions of its existing mines. These expansions were justified on economic grounds. Mexico's capacity increase was mainly due to the completion of the La Caridad mine. The operating cost net of by-product credits at this mine is higher than the average level of CODELCO mines, but it is below c60/lb today and is considered economically viable. 156. However, the cost of production at the Cerro Verde I mine in Peru which came onstream in 1977 has been relatively high. Although the economic viability of the additional capacities in Burma and Yugoslavia is not known, they are suspected to be marginal. 157. In the private sector operations in Canada, Australia, Brazil, South Africa, Philippines and Indonesia, the cost conditions and economic viabilities seem to range from very competitive to marginal. 158. This period of 1977-1983 also saw some significant net reductions in capacity in the United States, Japan, Zambia, Zaire, and Western and Southern Europe (in total, about 320,000 tpy). The large capacity losses by state- owned companies in Zambia and Zaire should be noted. 159. The copper companies controlled by oil companies did not expand their capacity, contrary to earlier expectations. One reason may be that the copper properties (whether they are already developed or still only potential mines) controlled by the oil companies are located in North America and are not the most economic properties. Despite the financial strength of most of the oil companies involved, capacity increases at these properties have rarely been economically justified. 160. However, there have been some exceptions. One example was the Disputada mines in Chile, acquired by EXXON in 1978. Production of copper in concentrates at Disputada mines rose from 26,460 tons in 1978 to 59,450 tons in 1983 and is expected to reach 77,000 tons in 1985. EXXON is currently expanding Disputada's El Soldad mine further, and may possibly expand the Los Bronces mine as well. Another example of capacity expansion by oil-company- owned copper producers is the rehabilitation of the Chino mine by the 1/ In Chile, CODELCO's production increased from 615,000 tons in 1973 to 893,000 tons in 1977, and to 1,033,000 tons in 1982. In the meantime, production in the small and medium mines, which are largely privately- owned, also increased from 120,000 tons in 1973 to 170,000 tons in 1977 and 208,000 tons in 1982. - 90 - Kennecott Corporation, owned by SOHIO, which in turn, is partly owned by British Petroleum and one-third owned by Mitsubishi. Kennecott and Mitsubishi also invested in modernization-cum-expansion of the Chino smelter. 1/ 161. In conclusion, mine capacity changes in the 1974-83 period may be characterized as follows. In the 1974-77 period, under the lagged influence of the high copper prices and optimistic market outlook of the 1964-73 period, capacity expansions materialized in a range of countries, primarily under the initiative of private mining companies (often involving multinationals). In contrast, in the 1977-83 period, the majority of capacity additions (notably in Chile and Mexico) was undertaken by state-owned companies; although significant amounts of capacity were also added by private sector initiative. In the latter period, there were also sizable capacity reductions by both state-owned and private companies. The decisions to increase capacity by producing companies, whether state-owned or privately-owned, were often induced by attractive offers of long-term purchase contracts and financing from buyers (often Japanese smelting and trading companies seeking to ensure adequate concentrate supplies). 162. In terms of the economic viability of the capacity added in this period, some of the projects turned out to be very competitive, while others were not justified on cost competitiveness alone. However, in terms of mistakes made in investment decisions, no clear difference appears between state-owned companies and privately-owned companies. C. Production Policies 1975-84 163. Under the conditions of chronic overcapacity, mine capacity utilization rates averaged only 86.6Z in the 1975-83 period (Table 6.4) compared with 93.5% in the 1960-74 period. In the face of declines in demand, some producers cut back production while others continued to produce at near full capacity. The question arises whether or not different producers acted rationally in either cutting back or maintaining capacity utilization. A corollary to that question is whether state-owned copper producing companies 1/ Although capacity additions by oil-company-owned copper producers have been rather limited so far, there is still a possibility that they could contribute capacity additions in the future. ANNEX D elaborates on the behavior of oil companies in the copper industry. Table 6.4: COPPER MINING CAPACITY UTILIZATION, SELECTED COUNTRIES/REGIONS, 1973-1983 (%) Country or region 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 United States 90.9 83.8 71.2 79.9 77.2 79.6 85.8 69.6 91.5 67.3 70.2 Canada 91.0 88.6 80.9 81.8 86.1 76.5 73.3 81.0 77.0 67.5 71.4 Chile 91.3 102.0 91.3 110.8 108.3 99.1 100.3 98.7 97.7 107.4 103.6 Zaire 96.9 91.8 87.3 78.4 86.7 77.9 76.6 92.1 100.3 97.2 93.9 Zambia 93.0 90.5 87.8 91.9 90.4 94.5 89.4 93.8 93.2 85.8 89.4 Yugoslavia 85.0 81.0 79.2 80.2 75.3 82.4 76.7 80.5 74.4 74.3 80.3 Western World 93.8 91.8 83.5 88.2 89.1 87.4 86.9 83.8 89.2 84.6 85.0 North America 90.9 85.5 74.4 80.5 80.2 78.6 81.5 73.5 86.5 67.4 70.6 West. World (Ex. N. America) 95.9 95.9 89.5 93.1 94.5 92.5 89.9 89.6 90.6 93.9 91.9 Source: Mine capacity data -- Phelps Dodge Corp. Production data -- World Bureau of Metal Statistics. - 92 - have been especially inflexible refusing to cut back their capacity utilization in the face of declines in demand and falling copper prices. 1/ 164. It has been argued that, to the extent that state-controlled, copper-producing companies pursue socially-imposed objectives other than profit maximization, they will be less flexible than private firms in cutting back operations in the face of depressed market demand and prices. Some costs that are clearly variable costs to private firms may be more or less fixed costs for state-controlled firms. Therefore, they can cover variable costs at a lower price. 2/ This is true of the cost of labor in the developing countries where alternative jobs are not readily available. However, it is also true of private companies located in the industrial countries where, by tradition, a large part of labor employed by private companies is considered fixed, e.g., Japan. This, of course, presents a sharp contrast to countries like the United States where labor costs are considered to be variable costs. 165. If the shadow exchange rate is higher than the market rate, the social value of the revenue from export sales would be higher than the private value of this revenue. If the state-controlled copper companies are an important foreign exchange earner, a decline in demand and price might increase the shadow exchange rate under a fixed exchange rate regime. This, of course, would tend to raise the social value of the output in terms of the domestic currency, thus offsetting partly the initial effect of the fall in the international price (measured in an international currency). 3/ 166. Sufficiently long time series data on capacity and production are not readily available on a company-by-company basis but are available on a country-by-country basis. Table 6.4 shows the capacity utilization rates for selected countries and regions for the 1973-1983 period, as calculated 1/ See the following quote from the Chairman of ASARCO Incorporated: "The public sector mining companies have not, to date, made a significant response to the economic imperative of reduced demand. They have continued to operate at full capacity, responding to social priorities and the need for foreign exchange." (Charles F. Barber, "Economics of Copper in the mid-80's, a Role for the Multilateral Financial Institutions," The 1983 Hamilton Lecture, Canadian Institute of Mining and Metallurgy, Toronto, Canada, December 8, 1983.) 2/ Radetzki, op. cit., p. 75. 3/ Radetzki, Ibid. - 93 - from capacity and production data. l/ During the period from 1964 to 1973 copper prices were at relatively high levels, and only decreased in the period since 1973. In evaluating the ups and downs in capacity utilization in North America, allowances must be made in 1974, 1977 and 1980 for the effects of production losses resulting from labor strikes in the United States. The countries where copper production has been primarily under state-control are Zaire, Zambia and Yugoslavia. In Chile and Mexico, state-owned companies account for more than 50% of the countries' production. In Peru, only 20-30% of production is state-controlled. 167. Table 6.5 shows the percentage changes in capacity utilization rates in the years 1975, 1977, 1981 and 1982, years in which copper prices declined substantially. In 1975 all countries except South Africa cut back capacity utilization, with Australia, the United States, Peru, Mexico, Chile and Canada experiencing the greatest cutbacks. Reductions in capacity utilization in Yugoslavia, Zaire and Zambia were much less than in the above mentioned countries. 168. In 1977, Peru, Zaire, Australia, Canada, Philippines and Papua New Guinea increased capacity utilization. The United States and South Africa hardly reduced it, but Mexico and Yugoslavia adjusted downward substantially and Chile and Zambia modestly. 1/ This study relied on the capacity data provided by the Phelps Dodge Corporation (PD). For the United States, US Bureau of Mines' unpublished data show magnitudes differing from the PD data. There are other sources for mine capacity data (e.g., RTZ data). However, for the sake of consistency in the definitions of mine capacity over a long period of time and across all countries, this study used primarily the PD data using other sources only to supplement the PD data. It should be emphasized that use of any capacity data to answer the question at hand presents some difficulties. Therefore, the analysis here is intended to obtain no more than broad indications. For example, while mine capacity is expressed in terms of copper content, actual capacity is determined by gross-weight throughput (usually at the concentrator stage) and the copper-content of this often changes over time. Although the capacity data used here comes from one of the best sources, some inaccuracies due to the above factor may well have crept in. In estimating a country's capacity during a year, we took a simple average of the capacity at the beginning of the year and at the beginning of the following year. This would not be accurate if a new large mine capacity came onstream either very early or very late. Another type of problem exists. A mine closure may be originally intended as a temporary step and is included in the capacity in that year, but it may be declared in the following year as "permanently closed", removed from capacity. Depending on how this type of situation is treated, the capacity utilization rates for the years concerned could change signficantly. In Table 4.19 over 100% capacity utilization was recorded for Chile for several years, and in Yugoslavia capacity was consistently underutilized. These figures suggest that there may be some problems with the capacity data used and/or the way they have been used here. - 94 - Table 6.5: CHANGES IN CAPACITY UTILIZATION RATES IN THE COPPER INDUSTRIES OF SELECTED COUNTRIES, IN 1975, 1977, 1981 AND 1982 (x) Country or Region 1975 1977 1981 1982 Average United States -15.0 -0.4 31.5 -26.4 -2.6 Canada -8.7 5.3 -4.9 -12.3 -5.2 North America -13.0 -0.4 17.7 -22.1 -4.5 Chile -10.5 -2.3 -1.0 9.9 -1.0 Zaire -4.9 10.6 8.9 -3.1 2.9 Zambia -3.0 -1.6 -0.6 -7.9 -3.3 Yugoslavia -2.2 -6.1 -7.6 -0.1 -4.0 Australia -16.6 8.5 -8.3 4.2 -3.1 Mexico -12.3 -13.5 31.4 8.8 3.6 Papua New Guinea -6.3 3.3 13.3 -2.9 1.9 Peru -27.4 18.5 -10.4 8.6 -2.7 Philippines -5.3 6.6 -1.3 -0.9 -0.2 South Africa 2.6 0.0 -1.5 -1.7 -0.15 Rest of the World /a -2.8 -4.8 4.0 9.3 1.4 Market Economies /b other than N. America -6.7 1.5 1.1 3.6 -0.1 /a The World (excluding CPEs) minus the countries specified above. 7b_ The world excluding CPEs, China, other non-market economies in Asia and Cuba. Source: Phelps Dodge Corp.; World Bureau of Metal Statistics; and The World Bank. - 95 - 169. During the 1981 recession, the United States recovered its capacity utilization dramatically (by 31.5%) up to 91.5% from the strike-affected 70% rate in 1980. Mexico, Peru, Zaire and Papua New Guinea also increased their utilization rates. The countries which reduced capacity utilization significantly in 1981 were Australia, Yugoslavia and Canada. In 1982, with the market declining further, the US industry reduced its utilization rate by 26%, while Canada, Zambia, Zaire, Peru and Papua New Guinea also reduced their utilization rates, though in a more modest way. In contrast, Chile and Mexico increased their utilization rates by 9-10%. 170. Thus, the results presented in Table 6.5 may be interpreted as showing that, at the time of a depressed market, the countries whose copper production is dominated by private sector producers tend to adjust capacity utilization more than those countries dominated by state-controlled companies, 1/ but the evidence is indeed weak. 171. In this context, it should be pointed out that a substantial portion of the capacity idled in the United States since 1980 is, in effect, permanently closed, and yet, still listed as parts of the assets of the books of the companies concerned. It is difficult to tell how many of these closed mines should be removed from the US capacity for the purpose of economic analysis. Similar problems exist in varying degrees for the capacity statistics of some other countries such as Canada, Australia and Philippines. 172. Capacity utilization behavior is expected to be affected by the cost competitiveness of the different producers, just as capacity increase/decrease decisions tend to be affected by such considerations. Figure 6.1 shows the locus of major copper producing countries in terms of their production costs and production quantity in the benchmark years of 1975, 1980 and 1984. 173. Over the period of 1975-84 Chile increased its mining capacity and production substantially. Chile's production expansion was achieved mostly by rationalization and expansion of existing mines and not by development of new mines. By this strategy, coupled with cost-conscious management through its corporate reorganization, CODELCO (state-owned) has successfully pursued a rehabilitation program to reduce its average production costs by maximum utilization of the capacity. Also CODELCO has been well-positioned to increase production because it has large existing mines which, with their high copper grades, can be expanded with less cost than development of new mines. An increase in production volume and a reduction in costs (in constant US dollar terms) are the twin results of Codelco's production strategy over the last decade. 174. In addition, part of the production from major producing countries such as Chile, Peru, PNG, Canada and Indonesia are sold in concentrates largely on the basis of long-term contracts covering at least one year (sometimes, up to fifteen years). Because the buyers are concerned with long- term security of supply, they try to accept the full contracted quantities. 1/ The countries where copper production is dominated by state-owned enterprises are Zaire, Zambia, Yugoslavia, India, Burma, Chile, and, to lesser extents, Mexico and Peru. - 96 - Figure 6. 1: CHANGE IN NET COST BY COUNTRY 1975-1964 go - 7 9A7 75 064 V1, TNS4 as 97s Wm so 7W70 3%, . 75 40 - &s So i le -.. s an 1280 1'm TONS 19711- 1964 ttU4 rem lut K FM J75 Z Country Keys to Figure 6.1 Source: Brook Hunt & Associates. A - Zambia F - Philippines K - Chile B - Zaire G - Indonesia L - Peru C = S. Africa/Namibia H - Canada M - Sweden D - Australia I - United States E - PNG J - Mexico - 97 - Such long-term minded buyers of copper concentrates have tended to get contract terms that ensure the continued operation of mines to supply their needs even during depressed price periods. 175. In the cases where the long-term buyers have provided substantial financing for development of mines in exchange for long-term marketing agreements, the borrowers' repayment obligations are often planned to go hand in hand with shipment plans. Because of these ties, producers try to produce as much as possible, as long as they can cover their cash operating costs. 176. By contrast, in Zambia (all of whose production comes from a state- owned company) production declined substantially over the 1975-1984 period (Figure 6.1). The production decline in Zambia was not a result of intentional cutbacks. Country specific factors contributed to Zambia's decline in production, including (i) transport difficulties, (ii) shortage of foreign exchange for the mining company to purchase spare parts and equipment and (iii) shortage of skilled labor. While "uneconomic" production continued in Zambia during loss-making periods, these financial losses were, for a large part, a result of the overvalued currency. After the overvaluation of the Zambian currency was corrected through massive devaluations, the Zambian copper industry began to show profits, partly reflecting the results of the state-owned copper company's genuine cost-cutting effotts. Although Zambia belatedly embarked on a rehabilitation program in 1982, it still encounters the various practical difficulties mentioned above. In addition, since the mining sector provides over 90% of the country's foreign exchange earnings, large demands made for foreign exchange by the rest of the economy resulted in insufficient foreign exchange for operating and maintenance needs of the copper company--a situation that complicated the already difficult financial problems of the company. These problems, in turn, led to gross underinvestment in development of the reserves and maintenance of the facilities and to a steady decline in copper production. 177. The 1981-84 period was notable for extensive cutbacks and closures, especially in North America. In 1982, cutbacks involved some 750,000 tpy of nominal capacity. Well over 500,000 tpy and 150,000 tpy of capacity was affected in the United States and Canada, respectively. Cutbacks also occurred in Philippines, Australia and Spain, and even in Chile and Peru. In Chile and Peru, the cutbacks and closures were concentrated in the small mines sector, where most mines were privately-owned. These cutbacks and closures continued in 1983 and 1984, principally in the United States. Mines in Type I producing countries (Figure 5.1), were under particular pressure. Mines in Type II producing countries, where co/by-product credits are not sufficiently large to offset their high gross costs, were also put under pressure because of depressed co/by-product prices, e.g., mines in Canada, the Philippines and Australia. 178. A further reason for the especially heavy concentration of capacity closures and production cutbacks in North America, and particularly in the United States, may be related to the short-term nature of sales contracts used by these companies. Buyers can react quickly to a lower demand for products made from copper. Captive mines operated by companies fully integrated downstream also are sensitive to changing market conditions and can react quickly. - 98 - D. CIPEC 179. The Intergovernmental Council of Copper Exporting Countries (CIPEC) was created by Chile, Peru, Zaire and Zambia in 1967, to coordinate policies of copper marketing and investment, among other objectives. 1/ The influence of CIPEC in the areas of market stabilization and supply management, for the purpose of supporting price levels, has been rather modest. Its first tangible success was achieved in 1974 when Japan complied with CIPEC's request to stop Japanese copper smelters from selling surplus refined copper on the LME. In November 1974, CIPEC announced a joint cutback of 10% of production and/or exports. When this action seemed to have little effect, and prices continued to fall, the announced cutback was raised to 15% of the members' production in April 1975. Actual reductions in production in these countries appear to have been less than those agreed to. Whether these actions had any effect is hard to determine because prices fell throughout the period. The export quotas were abandoned in 1976. 180. In 1977, in the face of falling prices, Peru, Zaire and Zambia agreed to reduce copper production and sales. Chile not only refused to join the action but announced a plan to increase capacity. No further attempt by CIPEC has been made to cut back production/exports in a coordinated fashion. l/ The membership has grown since 1967; it now includes Indonesia as a full member and Australia, Papua New Guinea and Yugoslavia as associate members. In 1982, CIPEC's full members accounted for 44% of mine production, 40% of smelter production and 26% of refining production in market economies. For a detailed discussion of the scope of CIPEC's objectives, history and an assessment of its role in copper market, see Gerhard Wagenhals, The World Copper Market, Structure and Econometric Model, Springer-Verlag, 1984, pp. 40-47; also see Stephen A. Zorn, "Producers' Associations and Commodity Markets: The Case of CIPEC," in F. Gerard Adams and Sonia A. Klein (eds.), Stabilizing World Commodity Markets (Lexington Books, 1978), pp. 215-234. - 99 - VII. TRADE AND PRICES A. Changing Patterns in Copper Trade 181. Total exports of copper from industrial and developing countries increased at an average rate of 2.9% per annum during the 1960-84 period (Table 7.1). However, the growth rate has slowed down markedly since 1973. The average growth rate in the 1973-84 period was 1.0% as compared with 3.4% in the 1960-73 period. This sharp slowdown reflects the slowdown in world copper consumption since 1973. 182. In the 1960-73 period, exports in the form of copper ores/concentrates increased more rapidly than other forms of copper (Table 7.1). As a result, the share of ores/concentrates of industrial and developing countries increased from 9.7% in 1960 to 25.3% in 1973. However, the share stabilized during the 1973-84 period. The share of blister copper after having declined sharply in the 1960-73 period, has also stabilized in the period since 1973. Indeed, the volume of blister exports hardly increased between 1960 and 1984. 183. The main reason behind the increase in the exports of copper in ores/concentrates form during the 1960-73 period was the rapid growth of smelting capacity and smelter production in consuming countries that had limited domestic mine production of copper, notably Japan. Japan's imports of ores and concentrates increased from 125,000 tons in 1960 to 832,000 tons in 1973, increasing its share in the industrial and developing countries' imports of ores and concentrates from 50% to 74% (Table 7.2). F.R. Germany also rapidly increased its imports of ores and concentrates in this period, but its share in the total imports of industrial and developing countries remained at around 12%. 184. In the early 1960s the Japanese Government became increasingly cncerned about the long-term security of supply of raw materials including copper, and encouraged the Japanese non-ferrous metal smelting/refining industry to increase its capacity rapidly and to make long-term arrangements for increasing imports of raw materials. 1/ For this purpose, ready financing on favorable terms was provided, e.g., through loans by the Export-Import Bank 1/ For an assessment of the role of government in Japan for development of overseas supply sources of nonfuel minerals as well as fuel minerals, see Raymond Vernon, Two Hungry Giants, The United States and Japan in the Quest for Oil and Ores, (Harvard University Press, 1983), pp. 98-106; and T.E. Kolenda, "Japan's Develop-For-Import Policy," Resources Policy, December 1985, pp. 257-266. Table 7.1: INDUSTRIAL AND DEVELOPING COUNTRIES' COPPER /a EXPORTS, 1960-1984 Average Growth /b Share in Total 1960 1973 1984 1960-73 1973-84 1960 1973 1984 ---'000 tons---- - per annum-- Ores/Concentrates 282 1,129 1,297 11.3 1.3 9.8 25.3 26.1 Blister Copper 788 781 794 -0.1 0.2 27.2 17.5 16.0 Refined Copper 1,824 2,549 2,878 2.6 1.1 63.0 57.2 57.9 Total 2,895 4,459 4,969 3.4 1.0 100.0 100.0 100.0 /a Includes copper content of ores/concentrates, blister and refined copper. /b Average compound rate between end points. Source: World Bureau of Metal Statistics. The World Bank Commodity Data Bank. - 101 - Table 7.2: IMPORTS OF COPPER IN ORES AND CONCENTRATES, SELECTED COUNTRIES, 1960, 1973 AND 1984 Share in Total Country or Region 1960 1973 1984 1960 1973 1984 -----'000 tons-- ----- - Germany, F.R. 31 134 150 12.5 11.9 12.0 Spain and Sweden 14 45 68 5.6 4.0 5.5 Japan 125 832 774 50.4 74.0 62.2 United States 73 38 36 29.4 3.4 2.9 Korea, Republic of 0 5 104 0.0 0.4 8.4 Yugoslavia 0 13 0 0.0 1.1 0.0 Industrial & Developing 248 1,125 1,245 100.0 100.0 100.0 Countries Source: World Bureau of Metal Statistics. World Bank Commodity Data Bank. - 102 - of Japan. 1/ A number of new mines, without smelting facilities, were opened up in the 1960s and early 1970s with Japanese financing and long-term marketing agreements to provide copper concentrates to Japanese smelting companies. 2/ 185. In the 1973-84 period, however, Japanese imports of copper concentrates decreased, while total imports of ores/concentrates by the industrial and developing countries increased modestly. Hence, Japan's share declined from 74% to 62%. Germany's share remained at around 12X. A feature of this period was the emergence of new importers of copper in ores/concentrates such as the Republic of Korea. 186. Another notable feature of world copper trade has been the increase in the imports of refined copper by the United States since the early 1970s (Table 7.3). United States imports increased from about 100,000 tons per year in the early 1960s to around 350,000 tons per year in the 1978-82 period, and increased further in 1983-84. United States imports increased at 6.7Z per annum in the 1960-84 period. As a result, the share of imports in total consumption in the United States rose to 25-27% in 1983-84, as compared with less than 10% in the early 1970s (Table 7.3). The increasing trend in the share of imports in the US consumption since the early 1960s reflects US copper industry's gradual loss of cost competitiveness over the last two decades. 187. The gradual loss of the "market power" of US copper industry over the last two decades is also reflected in the changes in its pricing policy. US domestic copper producers have miaintained a producer price system since the 19th Century. US producer price was maintained at somewhat higher levels than the LME price before World War II (on the average, about 6% higher than the LME price in the 1900-1945 period). 3/ The difference narrowed substantially in the 1950s and the early 1960s (Table 7.4). During the "high price" period 1/ Furthermore, import duties on copper products in Japan have been high and "escalated." Around 1980, import duties (MFN) on blister copper, and copper cathodes were 8.5% and 5.6%, respectively, while copper concentrates were duty-free. This structure of tariffs implies an effective protection rate of 50% for copper smelting in Japan. In contrast, import duties on these copper products in the EEC and the United States have been nil. For details, see UNCTAD, "Processing and Marketing of Copper: Areas for International Cooperation," TD/B/C.1/PSC/30, United Nations, 1982, mimeographed; and GATT, "Problems of Trade in Certain Natural Resource Products: Background Study on Copper and Copper Products," MDF/W/10, October 1984, mimeographed. 2/ See ANNEX C. 3/ Jose Luis Mardones, Enrique Silva and Cristian Martinez, "The Copper and Aluminum Industries," Resources Policy, vol. L1, no. 1, March 1985, pp. 3-16. - 103 - Table 7.3: IMPORTS AND CONSUMPTION OF REFINED COPPER IN THE UNITED STATES, 1960-1984 Share of Imports in Period Imports Consumption Consumption -------'000 tons per annum----- 1960-64 102 1,466 7.0 1971-75 185 1,985 9.3 1978-82 345 1,984 17.3 1983 483 1,775 27.2 1984 504 2,041 24.7 Source: World Bureau of Metal Statistics. - 104 - Table 7.4: COPPER PRICES, 1950-1984 Current $ 1983 $ Terms (MUV Deflated) LME Price US Producer Price LME Price US Producer Price Year $/ST US¢/lb $/MT USC/lb $/MT US¢/lb $/MT USo/lb US$/ton US$/ton US$/ton US$/ton 1950 493 22.4 468 21.2 2,107 95.6 2,000 90.7 1951 607 27.5 534 24.2 2,240 101.6 1,970 89.4 1952 715 32.4 534 24.2 2,527 114.6 1,887 85.6 1953 664 30.1 635 28.8 2,406 109.1 2,301 104.4 1954 686 31.1 655 29.7 2,541 115.2 2,426 110.0 1955 968 43.9 827 37.5 3,520 159.7 3,007 136.4 1956 906 41.1 922 41.8 3,190 144.7 3,246 147.3 1957 605 27.4 652 29.6 2,086 94.6 2,248 102.0 1958 545 24.7 658 25.8 1,841 83.5 1,919 87.0 1959 655 29.7 687 31.2 2,251 102.1 2,361 107.1 1960 677 30.7 707 32.1 2,272 103.0 2,372 107.6 1961 633 28.7 660 29.9 2,089 94.8 2,178 98.8 1962 644 29.2 675 30.6 2,091 94.8 2,192 99.4 1963 646 29.3 675 30.6 2,132 96.7 2,228 101.0 1964 968 43.9 705 32.0 3,133 142.1 2,282 103.5 1965 1,290 58.5 772 35.0 4,148 188.1 2,482 112.6 1966 1,530 69.4 797 36.2 4,737 214.9 2,467 111.9 1967 1,138 51.6 843 38.2 3,491 158.3 2,586 117.3 1968 1,241 56.3 923 41.9 3,830 173.7 2,849 129.2 1969 1,466 66.5 1,048 47.5 4,312 195.6 3,082 139.8 1970 1,413 64.1 1,272 57.7 3,914 177.5 3,524 159.8 1971 1,080 49.0 1,134 51.4 2,835 128.6 2,976 135.0 1972 1,071 48.6 1,116 50.6 2,581 117.1 2,689 122.0 1973 1,786 81.0 1,298 58.9 3,713 168.4 2,699 122.4 1974 2,059 93.4 1,690 76.7 3,514 159.4 2,884 130.8 1975 1,237 56.1 1,401 63.5 1,900 86.2 2,152 97.6 1976 1,401 63.5 1,517 68.8 2,123 96.3 2,298 104.3 1977 1,309 59.4 1,451 65.8 1,806 81.9 2,001 90.8 1978 1,365 61.9 1,444 65.5 1,635 74.2 1,729 78.4 1979 1,985 90.0 2,034 92.3 2,098 95.2 2,150 97.5 1980 2,183 99.0 2,257 102.4 2,105 95.5 2,176 98.7 1981 1,742 79.0 1,846 83.7 1,670 75.8 1,770 80.3 1982 1,480 67.1 1,607 72.9 1,462 66.3 1,588 72.0 1983 1,592 72.2 1,716 77.8 1,592 72.2 1,716 77.8 1984 1,377 62.5 1,491 67.6 1,402 63.6 1,518 68.9 Source: Metal Bulletin; Metals Week; The World Bank. - 105 - from the mid-1960s to the mid-1970s, US producer price was generally maintained below the LME level, partly because of keener awareness of threats from substitutes and partly because of the pressure from the government to contain price increases (for the sake of controlling inflation). 188. In the 1975-78 period, however, US producer price averaged 11% higher than the LME price. In 1978, major US producers decided to base their price quotes on the price at the New York Commodity Exchange (COMEX) which in turn is closely related to the LME price. Thus, since 1978, the US producer price has been changed more frequently than at any time before. The US producer price in the 1979-84 period was, on the average, 6.1X higher than the LME price. B. Impact of the Changing Exchange Value of the US Dollar on Copper Prices 189. As discussed in Chapter III, in the early 1970s, the world economy experienced rapidly accelerating inflation and a fall in the relative value of the US dollar vis-a-vis other major currencies, and in the early 1980s, control of inflation and sharp increases in the value of the dollar. While it is taken for granted that inflation affects current-dollar prices of commodities, it is not so well known how exchange rate changes affect commodity prices. 1/ 190. Theoretically, currency appreciation raises the foreign price and lowers the home price of traded goods. The extent of this impact depends on the relationships of the home to foreign market size, and of home and foreign elasticities of supply and demand for the goods concerned. Under free market assumption, the size of the elasticity coefficient for the exchange rate variable should be less than, or at most equal to, unity. 2/ As shown in ANNEX E, however, the impact of changes in the US dollar exchange rate on the current US$ price of copper at LME seems to have been much more significant than commonly believed, with the estimated elasticity at slightly greater than unity. One plausible interpretation is that, in the case of copper during the 1/ The following studies have analyzed the relationships between exchange rates and commodity prices: Ridler, D. and C.A. Yandle, "A Simplified Method for Analyzing the Effects of Exchange Rate Changes on Exports of a Primary Commodity," IMF Staff Papers, 19, 1972, pp. 559-578; C.A. Enoch and Panic, "Commodity Prices in the 1970s, Bank of England Quarterly Bulletin, March 1981, pp. 42-53; E.R. Crilli and M.C. Yang, "Real and Monetary Determinants of Non-Oil Primary Commodity Price Movements," in J.A. Kregel (ed.), Distribution, Effective Demand and International Economic Relations, MacMillan, 1983; K.Y. Chu and T.K. Morrison, "The 1981-82 Recession and Non-Oil Primary Commodity Prices," IMF Staff Papers, vol. 31, no. 1, March 1984, pp. 93-140; Marian Radetzki, "Effects of a Dollar Appreciation on Dollar Prices in International Commodity Markets," Resources Policy, September 1985, pp. 158-159; Christopher L. Gilbert, The Impact of Exchange Rates and Developing Country Data on Commodity Prices, World Bank Division Working Paper No. 1986-4, March 1986. 2/ For example, see Radetzki, ibid. - 106 - data period covered, the copper price tended to react speculatively to changes in the relative value of the US dollar, i.e., over-shooting. 191. In 1984, consumption of refined copper in the market economies rose sharply as a result of the rapid recovery of the world economy in apparent defiance of the popular thesis of structural stagnation in copper demand. Industrial countries' real GDP is estimated to have grown by 4.7% in 1984, while market economies consumption of refined copper increased by over 10%, implying an apparent "elasticity" of two! 1/ In contrast, market economies production of refined copper declined by 1.5% in 1984. While net exports of copper to CPEs and China decreased, the changing balance in refined copper in the market economies resulted in a sharp reduction in commercial stocks. Total commercial stocks, as reported by the World Bureau of Metal Statistics (WBMS) declined from 1.7 million tons at the beginning of 1984 to 1.2 million tons by the end of the year, or a decline of almost 30% for the year. Yet, the price of copper at LME in January 1985 averaged only 61.6C/lb. as compared to 62.4¢/lb. in January 1984. Indeed, the average price declined from 72.2c in 1983 to 62.5¢ in 1984--in the face of a sharp decline in commercial stocks. 192. As shown in ANNEX E, the US dollar price of copper is affected significantly by the change in the ratio of stocks to consumption, inflation and the US dollar exchange rate. 2/ In 1984 inflation in industrial countries was negative 1% in terms of US dollar GNP deflator, but about 3.3% in national currency GDP deflators. The exchange rate of the US dollar against the basket of major currencies (excluding the US dollar) appreciated by 11.25% between 1983 and 1984. Thus, the plausible explanation of what happened in the copper market in 1984 is that the very favorable change in the stock situation and the push from local currency inflation were more than offset by the "adverse" US dollar exchange rate movement. The regression results presented in ANNEX E 1/ It is no surprise that reported consumption of refined copper fluctuates much more than industrial production at turning points of business cycles. This fluctuation occurs because of changes in "invisible" stocks of copper which are in the form of stocks of semis, fabricates and copper containing manufactures. See M. Radetzki, "Fluctuations in Invisible Stocks: A Problem for Copper Market Forecasting", The World Bank, Commodity Paper No. 27, October 1977. 2/ Gilbert's recent study (cited above) concludes: "The long-run elasticity of the World Bank's commodity price indices with respect to a change in the value of the dollar was found to average 0.9. This response was analyzed into two components. The direct response averages 0.6. The indirect response, resulting from the impact of dollar appreciation or depreciation on dollar-denominated debt, accounts for the remainder." - 107 - tend to imply that, had the US dollar exchange rate remained the same in 1984 as in 1983, the price of copper might have been about 70¢/lb or some 12X higher than the actual average of 62.1¢/lb. - 108 - VIII. LONG-TERM MARKET PROSPECTS FOR COPPER 193. This chapter assesses the long-term prospects for the world copper industry. A discussion of the likely developments in the world economy is followed by a brief description of the projection techniques used and a summary of the base case projections. Then, market prospects for copper are discussed under four headings--i.e., future demand growth, capacity and supply scenarios, production cost trends and price outlook. A. Likely Trends in the World Economy 194. The recent return to economic growth (since 1983) has eased economic adjustments in industrial and developing countries alike, reducing the gravity of the "debt crisis," and keeping the political tensions arising from trade frictions within manageable bounds. A majority of developing countries, however, has yet to regain internal and external balance and to resume solid growth. Some key industrial countries have developed new disequilibria in their domestic and external financial relationships (e.g., high real interest rates and growing fiscal deficits). Thus, while the ongoing recovery has raised hopes for improved longer-term growth of the world economy, remaining hardcore adjustment problems require profound solutions. Disturbing questions abound in a period of high uncertainty. Are further reductions in unemployment likely? Will real interest rates return to more normal levels? What will happen to relative values (exchange rates) among major currencies? How long is the "debt problem" likely to retard the process of economic growth for developing countries? Will pressures for increased protection be contained? 195. Table 8.1 presents economic growth projections for industrial countries over the 1986-95 period. Industrial countries as a whole are expected to grow at 3.1% p.a. in the 1985-90 period, or at about the same rate as in the 1970s, and to grow somewhat faster in the 1990-95 period. These World Bank projections (prepared in early 1986) assume some changes in present policies. For the United States they assume that the government tightens fiscal policy so that the budgetary deficit as a percentage of total GNP would come down from the 1984 level of 5% to 3% by 1990. It is also assumed that the US Federal Reserve raises the monetary growth targets. For other industrial countries, the projections assume that they marginally ease fiscal policy in ways aimed at increasing the supply responses of their economies, with F.R. Germany and the United Kingdom cutting taxes and Japan increasing government expenditures. It is also assumed that these countries ease monetary policy. 196. Projected inflation in industrial countries for the 1985-95 period is shown in terms of various measures in Table 8.2. The inflation in the United States, as measured by the GDP deflator, is expected to rise from the recent level at 3.5% to 4.5% p.a. in the 1988-95 period. 197. Projections for interest rates and for the US dollar exchange rate vis-a-vis SDR for the period 1984-95 are also shown in Table 8.2. Because the Table 8.1: INDUSTRIAL COUNTRIES: REAL GDP GROWTH RATES, 1983-85 (ACTUAL) AND 1986-1995 (PROJECTED) (percent per annum) Country or Actual Projected Region 1983 1984 1985 1986 1987 1988 1989 1990 1991-95 North America 2.9 6.6 2.6 2.5 3.0 4.0 3.5 2.5 3.5 Europe 1.4 2.2 2.2 2.4 2.4 2.5 2.6 2.7 2.9 Japan 3.4 5.8 4.8 3.8 4.0 3.9 4.7 3.4 4.2 Industrial Countries: total 2.3 4.5 2.8 2.8 3.2 3.2 3.2 3.2 3.4 Source: EPDGL, December 1986. Table 8.2: NATIONAL INFLATION RATES, SDR/US$ EXCHANGE RATE AND INTEREST RATES, 1973-85 (ACTUAL) AND 1986-95 (PROJECTED) (x) 1973-80 /a 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1990-95 /b Average Annual Rate of Change National GNP/GDP deflators Europe 9.1 6.4 7.9 5.8 4.4 4.1 2.3 3.0 3.3 3.6 3.9 3.9 Japan 6.9 2.9 1.7 0.5 0.6 1.2 0.8 2.2 1.9 1.6 2.0 2.0 United States 7.5 9.2 6.0 3.8 3.8 3.5 3.5 3.9 4.5 4.8 4.5 4.5 SDR/US$ Exchange Rate -1.3 9.9 6.6 3.2 4.2 1.2 -6.8 -3.8 -3.9 -4.1 -4.2 -0.5 U.S. Real GNP 2.3 2.5 -2.2 3.6 6.8 2.5 2.5 3.0 4.0 3.5 3.5 3.5 Average Annual Rate U.S. short-term interest rate /c 9.3 16.6 13.5 9.8 11.2 8.7 9.0 9.0 9.0 9.0 8.0 7.5 U.S. Real Interest Rate /d 1.8 7.4 7.5 6.0 7.4 5.2 4.5 4.0 4.0 4.0 3.5 3.0 /a Average of 10 changes, from 1974 through 1980, inclusive. /b Figures have been reduced to the nearest half point. Tc Six-month dollar LIBOR, for projections and historically back to 1978. Prior to 1978, three-month Eurodollar rate. 7-d U.S. short-term interest rate less the rate of change of the U.S. GNP deflator. Source: The World Bank, Economic Analysis & Projections Department. combined effect of the reduced budgetary deficit on expected credit demands demands and the easing of monetary policy in the United States, the 6-month London Inter-Bank Offer Rate (LIBOR) is projected to average 7.5% in the 1990- 95 period. The real short-term interest rate in the United States is expected to continue to decline to 3.0% in the 1991-95 period. The value of US dollar in terms of SDR is expected to depreciate by 20% between 1984 and 1990 and to depreciate further, though slightly, during the 1991-95 period. 1/ The projected depreciation of the US dollar is expected to reduce inflation in other industrial countries. 198. Petroleum prices are expected to recover from the low levels prevailing in the first quarter of 1986. The average OPEC official selling price, which is projected to average US$13.50/bbl in 1986, is expected to recover steadily in the 1987-95 period, reaching US$17.90/bbl in 1984 constant dollars in 1995. This projection assumes, among other things, that: (a) OPEC and non-OPEC oil-exporting developing countries will decide to manage supply with the requisite political commitment; (b) the decline in prices is allowed to be partially passed through to the final consumers; (c) oil prices are targeted at such a level as to stop further erosion in the share of oil in the under-the-boiler market; and (d) Saudi Arabia, while intending to produce whatever share it is allocated, would implicitly agree to support the price structure. 199. Developing countries as a group are projected to grow at 4.3% p.a. in the period 1985-90 (Table 8.3). Many developing countries have been suffering from high inflation. Some of them have adjusted their exchange rates substantially. It is assumed that most developing countries that are important in the world copper market will succeed in controlling their inflation in the next 5-10 years and that any overvaluation or undervaluation of their currencies, if it exists today, will be corrected over the next few years. B. Projection Methodology and the "Base Case" Projections 200. An econometric model of the world copper market was used to project the evolution of the market in the medium to long-term. The model covers the industrial and developing countries only, but it allows for net trade in copper with the CPEs. A brief description of the model is provided in Annex F. The use of the model provides a framework for forecasting the copper market, allowing for easy comparison of projections made on the basis of different sets of assumptions (about income growth, exchange rate changes, substitution possibilities, etc.) Different response parameters can also be evaluated. 201. The model used has the following special characteristics. First, given the concern with stagnant consumption trends, emphasis is placed on the l/ A large part of this assumed depreciation already took place between September 1985 and March 1986. - 112 - Table 8.3: PAST AND PROJECTED WORLD GNP GROWTH RATES, 1973-95 Billion Country Group US$ Percent Change per Annum or Region 1980 1973-80 1980-85 1985-90 1990-95 ALL DEVELOPING COUNTRIES 2,257 5.0 3.3 4.3 4.7 90-COUNTRY SAMPLE 2,091 5.4 3.3 4.4 4.9 Low Income Countries 549 4.7 7.2 5.3 4.9 Africa 53 2.7 0.7 3.5 3.6 Asia 495 5.0 7.8 5.4 5.0 Middle Income Countries 1,543 5.6 1.8 4.0 4.8 Oil Importers 977 5.5 2.0 4.5 5.1 Major Exporters of 746 5.9 2.1 4.7 5.3 Manufactures Non-oil 231 4.5 1.6 4.0 4.3 Oil Exporters 566 5.8 1.3 3.2 4.3 OTHER DEVELOPING COUNTRIES /a 166 1.1 3.3 2.7 3.0 HIGH INCOME OIL EXPORTERS 230 7.6 -2.3 4.1 5.6 INDUSTRIAL COUNTRIES 7,444 2.8 2.1 3.1 3.4 North America 2,841 2.9 2.3 3.3 3.5 Pacific 1,213 3.9 4.0 3.9 4.2 Western Europe 3,390 2.4 1.3 2.5 2.9 WORLD (Excluding CPEs) 9,931 3.4 2.3 3.4 3.8 Centrally Planned Economies /b 2,212 3.2 2.4 2.4 2.6 USSR 1,719 3.2 2.7 2.4 2.6 OTHER E. EUROPEAN N.M.E. /b 493 3.1 1.4 2.3 2.4 WORLD (Including CPEs) 12,143 3.3 2.3 3.2 3.6 /a Excludes Bahamas, Cuba, Kampuchea, Korea D.R., Laos, Mongolia, Namibia, W. Samos and Vietnam. /b Excludes Albania. Source: The World Bank, EPDGL (January 1986). - 113 - demand side of the model. Consumption in six major consuming countries (United States, Japan and four major EEC countries) is modelled for each of five major end-use sectors (i.e., electrical, construction, transport, general engineering and domestic use sectors). This approach allows analysis of the future trends in consumption in each sector, including the impact of substitution. In view of the changes in the structure of the industry over the last 10-15 years, it was decided to estimate demand relationships on the basis of data only from the late 1960s onwards. 202. Second, in view of the failure on the part of other econometric models to present mine capacity investment behavior in a form that is useful for predictions (especially, on a country-by-country basis), it was decided to treat mine capacity exogenously. Mine capacity is projected judgmentally on the basis of available information on planned and potential capacity increases/decreases, taking into account possible future price developments. In other words, to the extent that the impact of projected price trends on capacity changes is taken into account, the process of projecting mine capacity is done "interactively." Mine production is then determined by projected mine capacity and the price of copper adjusted for inflation (which is assumed to reflect the trends in costs.) 203. "Base case" projections for refined copper consumption, mine production and the LME price are summarized in Table 8.4, along with assumed mine capacity estimates and net trade with CPEs. 1/ These projections were derived on the basis of the macroeconomic assumptions set out in the previous section. In short, world consumption is projected to grow at 1.3% p.a. between 1984 and 1995, and mine production at 0.8% p.a. The LME price in 1984 dollar terms is projected to remain near present low levels until 1990 and then to rise moderately to 74¢/lb. by 1995. Detailed discussions of these projections are presented below. However, please note that, use of the model (which was constructed on the basis of historical data) for the purpose of forecasting, required the introduction of some modifications to the model. These modifications will be noted below at appropriate places. C. Demand Prospects 204. As pointed out in Chapter IV, critical factors influencing copper consumption are; (a) economic growth (in particular, the level of real per capital income or industrial prodcution), (b) the relative share of industrial production in GNP, (c) technological changes resulting in material substitution and material saving, and (d) the price of copper. In this section, the most likely scenario, the "Base Case" (Case A) for consumption of refined copper in the next decade is examined in terms of the likely developments in the above-mentioned key parameters. The "low-economic growth" case (Case B) follows along with a discussion of possibilities for higher copper consumption. 1/ CPEs = East European Nonmarket Economies. - 114 - Table 8.4: SUMMARY OF "BASE CASE" PROJECTIONS FOR WORLD (EXCLUDING CPEs) COPPER MARKET, 1983 AND 1984 (ACTUAL), 1990 AND 1995 (PROJECTED) Actual Projected Unit 1983 1984 1990 1995 Mine Capacity, Beginning of Year '000 tpy 7,668 7,619 7,915 8,000 Mine Production (Cu content) '000 tons 6,529 6,715 7,160 7,340 Refined Copper Production 000 tons 7,664 7,546 8,500 9,100 Refined Copper Consumption 000 tons 7,190 7,886 8,375 9,130 Net Imports From CPEs /a 000 tons 85 129 120 120 Copper Price, LME 1984$ ¢/lb. 70.9 62.5 63.4 74.1 Current$ ¢/lb. 72.2 62.5 90.0 128.0 /a Net imports of refined copper from East European Nonmarket Economies. Source: 1983-84 (Actual) - World Bureau of Metal Statistics. 1990 and 1995 (Projected) - The World Bank. "Most Likely" Scenario (Base Case) - Case A 205. Table 8.5 shows the "most likely" levels of refined copper consumption in 1990 and 1995 for selected key countries and for the world. These projections assume, among other things, the GNP projections presented above. It is also assumed that the long-term trend of the share of industry in GDP in industrial countries will continue to decline. I/ 206. The above projections also reflect likely effects of technological changes in the end-use markets. As reviewed in Chapter IV, copper end-use market losses resulting from substitution and material saving were substantial in the 1950s and 1960s but appear to have slowed down somewhat in the 1970s. Further losses are expected to occur in the projection period. 207. There are many areas, such as automobile radiators, where further substitution against copper is possible. Furthermore, substitution by plastics in plumbing and tubing for construction and industrial applications is expected to accelerate. The likely impact of fiber optics technology on copper use in telecommunications is substantial. This will be discussed below. Because of technological "leap frogging", the rising trend in the intensity of use in developing countries will probably moderate more quickly than the historical pattern indicated by the now industrially mature economies. 208. In some areas, such as high tension power transmission cables, substitution against copper is virtually complete. In other areas, innovations and changing market preferences are favoring copper--e.g., the trend toward prefabrication in the building industry. Indeed, there are areas where copper is likely to find expanding uses--e.g., the trend toward increased electric/electronic gadgetry for automobiles. On balance, it is likely that in the industrial countries, market losses due to substitution and material-saving innovations will continue at a pace similar to that in the last 15 years. 209. One new technological development expected to cause significant displacement of copper use in the coming decade is the application of fiber optics in the telecommunications sector. Substitution by optical fiber cables for traditional copper cables represents a new source of market loss for copper. Therefore, a separate projection exercise was carried out to estimate the likely losses for copper in this area in the six largest copper-consuming countries--i.e., the United States, Japan and four EEC countries (F.R. Germany, France, Italy and the United Kingdom) which account for about two- thirds of total consumption in the world. I/ This is by no means an assumption on which consensus exists. For example, Data Resources, Inc. (DRI), expects that "industrial output increases dramatically as a share of economy-wide output" in the United States in the 1985-95 decade. See DRI U.S. Long-Term Review, Winter 1984-85, p. 116. - 116 - Table 8.5: CONSUMPTION OF REFINED COPPER - "BASE CASE" PROJECTIONS FOR SELECTED COUNTRIES AND THE WORLD (EXCLUDING CPEs), 1990 AND 1995 Actual Projected Country Group 1983 1984 1990 1995 -------------'000 tons------- --- Industrial Countries United States 1,775 2,041 2,055 /a 2,075 /a Canada 195 231 235 235 EC-4 /b 1,810 1,890 1,790 /a 1,900 /a Japan 1,216 1,369 1,340 /a 1,480 /a Developing Countries Brazil 148 181 325 440 China /c 487 540 /c 715 835 Korea, Rep. 152 170 225 320 India 96 72 110 125 "Rest of World (excl. CPEs) /d 1,271 1,400 1,575 1,720 Industrial & Developing Countries /e 7,189 7,880 /e 8,370 9,130 /a These projections reflect estimated losses due to substitution by optical fiber cables in telecommunications sector, shown in Table 8.6 below. /b Germany (F.R.), France, Italy and the United Kingdom. 7TW China in this table includes Taiwan. /d The world excluding East European Non-Market Economies and all the countries listed separately above. Thus, the rest of the world includes some industrial as well as some developing countries. /e World excluding East European Non-Market Economies. Source: Actual (1983 and 1984) - World Bureau of Metal Statistics. Projected (1990 and 1995) - The World Rank. - 117 - 210. Annex H contains a brief description of the current situation with fiber optics technology and a discussion of the way the estimate of the prospective copper market losses was formulated. Table 8.6 summarizes the estimation results. For the six major countries, the total loss of copper consumption due to fiber optics technology is estimated at 235,000 tons in 1990 and 325,000 tons in 1995. 211. There is a greater chance that estimates in Table 8.6 may turn out to be too low rather than too high. First, a recent estimate by a communications industry consultant suggests that, for the United States, the loss in the feeder loop of the telecommunication system alone could be as high as 90,000 tons by 1990, 1/ compared to 30,000 tons given in Table 8.6 for the entire telecommunication sector of the United States. Second, estimates given in the table cover only six major consuming countries and other countries, both industrial and developing, will experience additional losses. Such losses may be "guesstimated" ranging up to 50 thousand tons in 1990 and 75 thousand tons in 1995. Third, it should be remembered that the speed of technological advance and adoption in this area has been consistently underestimated. 212. Broadly speaking, the "base case" projections for consumption indicate the following: (a) World copper consumption increases at 1.3X p.a. between 1984 and 1995. (b) In industrial countries, consumption levels in 1995 are not likely to be higher than the levels achieved in 1984, except for Japan. (c) In the selected industrializing developing countries, consumption will grow significantly if these countries manage to achieve the relatively high economic growth rates assumed. 213. The intensities of copper use in selected countries implied by the "base case" projections are illustrated in Figures 8.1 and 8.2. Figure 8.1 shows that the declining trend in the intensity of use in the industrial countries, i.e., Canada, United States, Japan and the four major EEC countries, will continue in the coming decade. Figure 8.2 shows that historical trends in the intensity of use in the selected developing countries have been divergent and that future trends are expected to take divergent directions also. 1/ Harvey Blustain, The Impact of Fiber Optics on the Copper Wire and Cable Industry, a multi-client study by IGI Consulting, Inc., Boston, Mass., 1985, p. 88. - 118 - Table 8.6: IMPACT OF FIBER OPTICS TECHNOLOGY ON COPPER CONSUMPTION IN MAJOR INDUSTRIAL COUNTRIES, 1990 AND 1995 ('000 tons) Consumption Without Market Loss Due Impact of Fiber Optics to Fiber Optics 1990 1995 1990 1995 United States 2,085 2,125 30 50 Japan 1,405 1,560 65 80 EC-4 /a 1,930 2,095 140 195 Total of Above 5,420 5,780 235 325 /a France, Germany (Federal Republic of), Italy and United Kingdom. Source: The World Bank. Figure 8.1 REFINED COPPERtINTENSITIES OF usE - CANADA { - - - C-4 2.0 1 -2.0 o0 A 00 0\ .1.5 lYea ouc T 00 000 do .-* 1. .1-I1501 O.-40. Oha-4 O sO117 7i r0ub 4 Year Source: The World Bank. Figure 8.2 EFrINED COPPER:[NTENSITIES OFr USE 7. ,- BRZIL 2.5 Ls co S4 . 0 v> a0 .-. . 0 1.5 1.0 i"+''>-r - 0 1- 1.0 r1. 0.5- - - - - - - - - - - - - 5 so as, 2 75Ug js w* Year Source: The World Bank. - 121 - "Low Economic Growth" Scenario - Case B 214. What would be the impact on copper demand of slower world economic growth than assumed in the "base case"? To assess the possible consequences of such a macroeconomic scenario, a simulation based on the model has been carried out. For all separately identified consuming countries, GDP growth rates in the 1984-95 period are reduced by 20% from the growth rates assumed for the base case. 1/ The results show that: (a) World consumption in 1995 would be 8.6 million tons compared with the base case 9.1 million tons, or reduced by 5.5%. (b) The LME price in 1995 would be reduced down by 22%. 2/ 215. For the above simulation, only the GDP growth assumptions for the specified countries have been changed from the base case. Mine capacity projections are assumed to be the same as the base case, although logically they should be reduced. It may also be argued that such a low economic growth scenario should logically be combined with higher or lower inflation for these countries and perhaps higher or lower real interest rates, which in turn may imply a stronger or weaker US dollar exchange rate. However, there is no consensus view on how all these macroeconomic assumptions should hang together. Thus, depending on what other macroeconomic assumptions should be changed along with the GDP growth assumptions, the outcome of the combined effects on copper demand and price would be different. Possibilities of Higher Consumption Growth 216. Copper consumption could grow more rapidly than in the base case, if public and private sector investment picks up significantly in industrial countries during the projection period. Because of the prolonged economic turmoil in these countries in the last decade (the need to control inflation, the need to control fiscal deficits, high interest rates, capacity underutilization, etc.), social infrastructure and private sector production facilities have not been maintained well. It is sometimes suggested that, as the world economy recovers, in future time, physical deterioration trends in these areas will have to be reversed. If or when that should occur, 1/ These countries accounted for 80% of world consumption in 1983. 2/ The above simulation results underestimate the possible impact of assumed reductions in world economic growth on the copper market. The GDP assumptions for only the specified countries have been changed, i.e., assuming no change in economic growth in the rest of the world. Because of the structure of the model, consumption in the latter region is not responsive to changes in GDP growth assumption. However, in reality, if all the specified countries are experiencing slower economic growth, it is fair to assume that this should reduce the economic growth of the rest of the world as well, thus reducing copper demand growth in the latter group. Therefore, the above simulation results underestimate the reductions in world consumption growth and price. - 122 - consumption of copper will be stepped up significantly as production of investment goods tends to be relatively copper-intensive. 217. Another possible scenario for higher copper consumption growth would be that somehow copper producers and fabricators will be successful in promoting the use of copper. Traditionally, the world copper industry has spent little on promoting consumption of copper. As of 1985, estimates show that the total annual budgets of various copper promotion organizations amounted to a little over US$10 million. 1/ No good estimates exist to show how much individual copper producing and fabricating firms have expended for demand creation on their own. That they spend very little for research, development and advertizing aimed at promoting copper consumption is well known. This situation presents a stark contrast to other metals. The aluminum industry has traditionally spent far more aggressively for promoting the consumption of aluminum products. 2/ 218. Individual copper mining companies do not spend much on promoting consumption of copper for good economic reasons. First, copper, unlike aluminum, is primarily used for production of capital goods. Second, even when copper ends up in consumer goods such as residential buildings and automobiles, products used in production of these goods are in some forms of intermediate products where copper accounts for only a small portion of the total value of the products. Third, and most important, copper producers are seldom vertically integrated downstream far enough to include the production of these intermediate products. Semi-fabricated and fabricated products are essentially homogeneous products. Therefore, developing strong brand loyalty among customers is difficult. If use of certain copper containing intermediate products is promoted successfully by an individual producer, the benefits of sales volume increases are likely to be shared by other producers. 219. Because of the externality mentioned above, individual copper producers or semi-fabricators are willing to incur expenditures for promoting consumption of copper products only if it is done jointly with other producers. However, for a large copper producer such as CODELCO (with annual copper production at over one million tons), allocation of funds to promote consumption of copper products may be economically justified, whether or not other producers join in the action, provided that specific opportunities for promoting products made from copper are identified. 1/ Information provided by the International Copper Research Association. 2/ For a review of contrasting attitudes of copper and aluminum industries toward creation of demand, see Mardones, Jose Luis, Enrique Silva and Cristian Martinez, "The Copper and Aluminum Industries," Resources Policy, March 1985, pp. 3-16. - 123 - 220. The rationale for promotion of copper consumption by an individual producer (if it has a large share of world production) is as follows. Suppose specific opportunities to promote products made from copper (e.g., a new type of copper-based radiator for automobiles, copper roofing, and copper pots and pans) are identified. Suppose that a large copper producer such as CODELCO expends five million dollars a year for promoting these copper products while some other large producers (perhaps two or three mining companies or semi- fabricators) jointly spend a similar amount. If, as a result of promotion efforts, copper consumption is raised sufficiently to increase the price of copper by one cent per lb, the "rate of return" on the promotion expenditures would be spectacular (a lC/lb increase in the price of copper implies an increase in revenue of 22 million dollars on one million ton production). 221. Recent activities suggest that major copper producers and fabricators have become increasingly aware of the need to focus on promotion of the use of copper. Various initiatives for cooperative market development efforts have been launched recently. 1/ If these efforts prove to be successful, copper consumption could grow marginally more rapidly than the base case projections. D. Supply Prospects 222. Supply of refined copper derives from two sources: (a) newly-mined copper processed into refined copper either through smelting and refining or through leaching and electrowinning (hydrometallurgical process); and (b) scrap copper refined into refined copper. In assessing supply prospects, attention is usually focussed on mining capacity developments because the availability of copper over the long-term is limited by mine capacity. Copper supply could be increased in the short run by increased scrap collection and processing. In the longer run, however, this source tends to "dry up" (because scrap is a form of stocks). Similarly, although using up (or drawing down) of stocks represents additional supply of refined copper, it is so only in the short run. 223. The process of building new mining capacity normally takes longer and costs more than building smelting or refining capacity. Building new mining capacity typically takes 4-5 years (2-3 years for expanding existing mines) after investment decisions are made. Building smelting and/or refining facilities typically takes only 2 years. The capital cost of establishing one annual ton (tpy) of capacity is typically higher for mining capacity (e.g., $4,000-6,000 per tpy for a greenfield project and $3,000-4,000 for expansion of existing mines) than for increases in smelter and refinery capacity (e.g, $3,000 or less per tpy). Furthermore, because processing facilities need not be located close to mining sites, there is a greater scope for choice of locations. 1/ Examples of recent initiatives for copper promotion are the joint effort by Kabelmetal and RTZ and the increase in the promotion budget by the CIPEC. - 124 - 224. Copper deposits tend to be concentrated geographically. Currently, world copper reserves are estimated at 525 million tons (Table 8.7). Eight countries with the largest reserves (Chile, the United States, the USSR, Australia, Zambia, Canada, Peru and Zaire) account for 73% of the existing world reserves. The next 4 largest countries account for another 12%. Assuming that copper consumption grows at 1.5% p.a., the existing reserves would last 50 years. Over time, world copper reserves have tended to increase steadily as a result of exploration, and they are likely to increase in the future at more or less the same pace as consumption. 1/ 225. Table 8.8 shows mine capacity existing at the beginning of 1984 and projected for each year of the period 1986-95. These mine capacity projections are assumed for the base case projections of mine output and refined copper production. This set of mine capacity projections has been derived judgmentally, after running dozens of simulations based on alternative capacity projections. The projections take into account (a) long-run marginal costs of the mines involved, (b) the forecast price trend (which affects the viability of existing and potential mines), (c) the investment climate and (d) likely depletions of reserves. For greater detail, including some methodological considerations, see ANNEX I. Base Case Supply Projections 226. Under the base case scenario, world mine capacity increases by about 300,000 tpy from 7.6 million tpy in 1984 to 7.9 million tpy by 1990, and by an additional 100,000 tpy by 1995. Main expected developments are as follows: (a) United States' overall mine capacity declines further as depressed prices continue in the next few years, in which (i) many temporary closures transform into permanent ones; (ii) some high-cost mines currently operating cannot sustain production; (iii) depletion of reserves continues to affect capacity, forcing closure of some mines that are competitive today; and (iv) new capacity is virtually limited to a few low-cost small leaching operations. However, some currently closed mines may be renovated by new investments. 1/ World copper reserves increased dramatically between 1960 and 1976, increasing the "lifetime" of reserves. Wolfgang Gluschke, Joseph Shaw, and Bension Varon, Copper: The Next Fifteen Years, published by the United Nations, by D. Reidel Publishing Company, 1979, Chapter IV. - 125 - Table 8.7: WORLD COPPER RESERVES Reserve Base /a Country Group (Million TontsT Industrial & Developing 470 Australia 41 Canada 23 United States 90 Yugoslavia 7 Chile 97 Peru 32 Mexico 23 Argentina 7 Panama 12 South Africa 4 Zaire 30 Zambia 34 Papua New Guinea 14 Philippines 18 Iran 5 China 3 Other Industrial & Developing 31 Centrally Planned 55 Poland 15 USSR 36 Other Eastern Europe 4 World 525 /a That part of an identified resource that meets specified minimum physical and chemical criteria related to current mining and production practices, including those for grade, quality, thickness, and depth. The reserve base includes those resources that are currently economic, marginally economic, and some of those that are currently subeconomic. Source: US Bureau of Mines, Mineral Commodity Summaries 1986; Details through direct communication with USBM. Table 8.8: COPPER MINING CAPACITY AT BEGINNING OF YEAR, 1984-96 (BASE CASE) (O000 metric tons per year) Economies 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Major Producers United States 1,550 1,450 1,300 1,250 1,250 1,250 1,250 1,250 1,200 1,150 1,100 1,050 1,050 Canada 850 850 880 880 880 880 880 880 880 880 880 880 880 Mexico 250 250 300 330 340 340 340 340 340 340 340 340 340 Chile 1,271 1,300 1,350 1,400 1,450 1,500 1,600 1,650 1,670 1,700 1,750 1,800 1,850 Peru 395 395 400 400 420 430 435 435 450 460 470 470 480 Zambia 600 570 570 560 540 530 520 500 480 470 460 440 420 Zaire 544 520 520 510 520 530 540 550 550 550 560 570 570 S. Africa 222 215 215 215 215 210 210 200 200 190 190 190 180 Philippines 371 350 350 350 360 370 370 370 370 380 380 380 380 Australia 263 260 260 260 280 300 310 320 330 340 350 350 350 Papua New Guinea 185 185 180 180 180 180 190 200 210 220 220 230 230 TOTAL 6,501 6,345 6,325 6,355 6,445 6,500 6,620 6,640 6,630 6,600 6,560 6,500 6,530 Other Producers (Not projected separately except for China + CPE Asia) Algeria, Morocco 32 Bolivia 9 Botswana 27 Brazil 36 Burma 18 Finland 38 France/Greece 3 Germany, F.R. 2 India 36 Indonesia 91 Ireland 4 Iran 45 Israel 5 Italy 2 Japan 59 Korea, Rep. 2 Malaysia 27 Namibia 52 Norway 30 Portugal 5 Spain 81 Sweden 85 Taiwan, China 3 Turkey 27 Yugoslavia 181 Zinbawe 32 China + CPE Asia 187 187 187 190 200 200 210 215 220 225 230 235 240 TOTAL 1,117 1,151 1,180 1,200 1,230 1,250 1,270 1,300 1,300 1,300 1,300 1,300 1,300 World Ex CPEs 7,619 7,496 7,505 7,535 7,665 7,770 7,915 7,995 7,980 7,980 8,000 8,000 8,030 Market Economies 7,432 7,309 7,318 7,345 7,465 7,570 7,705 7,780 7,760 7,755 7,770 7,765 7,790 Source: The World Bank. - 127 - (b) Capacity in Zambia and South Africa falls significantly; no new capacity is expected in these countries during the projection period while capacity losses result from deterioration of ore grade, depletions of reserves and, in Zambia, from closing of uneconomic mines. 1/ (c) The majority state-owned Cananea expansion in Mexico reaches full production in the next three years. (d) Chile's capacity will expand by 300,000 tpy by 1990 and by a further 200,000 tpy by 1995. Notable increases envisaged are expansions of existing CODELCO and EXXON's Disputada mines, and the privately owned Mantos Blancos, Cerro Colorado and/or Escondida projects. The Escondida project is owned by Utah International, RTZ and Mitsubishi Corp. and is expected to reach 300,000 tpy by mid-1990s, if undertaken. (e) Peru's capacity increases as the almost-completed Tintaya project gradually reaches full production (40,000 tpy). MINEROPERU's Cerro Verde II (50,000 tpy) is uncertain but is assumed to be completed sometime in the 1990s. (f) Philippines is assumed to regain some of the capacity lost as the economic climate improves. (g) In Australia, the Olympic Dam mine is assumed to be developed by Western Mining and BP Australia on a phased schedule, to reach an eventual capacity of 150,000 tpy by the mid-1990s. The mine is rich in by-products (gold, silver and uranium) as well as in copper (1.67Z grade). (h) In Papua New Guinea, as the capacity at the existing Bougainville mine declines because of falling ore grade, copper production from the Ok Tedi II project (owned by Broken Hill Pty, AMOCO, etc.) is assumed to increase. (i) In Portugal, the Neves-Corvo project, with a majority (51%) ownership by the Portugese-Government-owned EMMA, has a rich ore-body grading of over 8% and is to be developed as an underground mine by RTZ Ltd. It is expected to reach an eventual capacity of 120,000 tpy by the mid-1990s. See "Zambia Bites the Bullet", Mining Journal, vol. 306, no. 7851 (February 7, 1986), pp. 89-91; and Zambia ZCCM Rationalization," CIPEC Quarterly Review, October-December 1985, pp. 89-90. - 128 - (j) In the rest of the world a number of closures will occur due to depletion, but some capacity additions will be made as well. Specifically, capacity losses are likely in Botswana, Indonesia, Malaysia and Sweden. Iran's Sar Chesmeh mine, which was virtually completed before the 1979 revolution in Iran, is assumed to increase its effective capacity to reach the original design capacity (145,000 tpy) in a few years. Some capacity increases may take place in Burma (Monywa expansion by a state mining company), Spain (Minas Riotinto) and Yugoslavia (Veliki and Bucin/Cukor expansions). Additional increases are possible in Argentina and Brazil but these are not very likely to come onstream before 1995. 227. The above "base case" set of mine capacity projections is associated with the "base case" price projections which indicate that the LME price in constant 1984 dollars, after remaining flat, will rise moderately from 63¢/lb in 1990 to 74¢/lb in 1995. In the 1985-90 period the price is expected to remain near present low levels in constant dollar terms because the projected net capacity increase (300,000 tpy) will sustain the present surplus capacity overhang. This increase in mine capacity is likely to materialize because most of the planned additions are either already at advanced stages of development or their projected production costs are so low that the projects are viable at prices no higher than the current level. 1/ Partly because of the continuation of the relatively low price levels in the 1986-90 period, no more than a few new projects are likely to be planned for the period 1990-95; or their aggregate increase is likely to be only 100,000 tpy more than needed to offset the likely capacity losses resulting from ore depletions and closings of high-cost operations during the period. 228. Under the base case assumptions, world mine production grows from 6.72 million tons in 1984 to 7.16 and 7.34 million tons by 1990 and 1995, respectively. Table 8.9 shows projected mine production levels for major producing countries. Capacity utilization rates in the United States and Canada are expected to remain low during the 1986-90 period. In the United States the rate is likely to remain at around 75% (or about the same level as in 1983) until the early 1990s and then improve to reach 85-86% by 1995. In Canada, the utilization rate is expected to improve slowly from 72% in 1983 to approach 80% by 1995. In other countries, capacity utilization is projected to increase to and remain above the 90% level in most cases, except for Philippines and Papua New Guinea where it is likely to be in the 80-85% range. 1/ Capacity additions planned in Burma, China and Yugoslavia may be carried out by state enterprises regardless of their international competitiveness. - 129 - Table 8.9: PROJECTED MINE PRODUCTION OF COPPER, 1990 AND 1995 "BASE CASE" (CASE A) Actual Projected Country Group 1983 1984 1995 Industrial Countries of which 2,169 2,266 2,175 2,135 Australia 265 246 300 330 Canada 625 706 685 700 United States 1,038 1,087 975 890 Developing Countries of which 4,359 4,449 4,975 5,200 Chile 1,257 1,290 1,550 1,740 Mexico 206 186 305 305 Peru 322 364 405 445 Philippines 271 220 310 320 Papua New Guinea 183 163 170 190 South Africa 212 215 195 175 Zaire 502 501 505 540 Zambia 578 565 500 425 Rest of World /a 1,068 1,172 1,250 1,275 World Total 6,465 6,715 7,150 7,335 /a All countries not separately listed above, excluding CPEs. Source: Actual - World Bureau of Metal Statistics. Projected - The World Bank. Table 8.10: COPPER MINING CAPACITY AT BEGINNING OF YEAR, 1985-96 - CASE C ('000 metric tons per year) Country Group 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 Major Producers United States 1,450 1,300 1,200 1,100 1,100 1,100 1,100 1,000 1,000 1,000 900 900 Canada 850 850 880 880 880 880 880 880 880 880 880 880 Mexico 250 300 330 340 340 340 340 340 340 340 340 340 Chile 1,300 1,350 1,400 1,420 1,450 1,500 1,530 1,550 1,570 1,600 1,630 1,650 Peru 395 400 400 420 430 435 435 450 460 470 470 480 Zambia 570 550 540 530 520 500 480 460 440 420 400 400 Zaire 520 520 510 520 530 540 550 550 550 560 570 570 South Africa 215 215 215 215 210 210 200 200 190 190 190 180 Philippines 350 350 350 360 370 370 370 370 380 380 380 380 Australia 260 260 260 280 300 310 320 330 340 350 350 350 Papua New Guinea 185 180 180 180 180 190 200 210 220 220 230 230 0 TOTAL OF ABOVE 6,345 6,275 6,305 6,245 6,310 6,375 6,405 6,340 6,370 6,410 6,340 6,360 Other Producers (Not specified) Total Other 1,151 1,180 1,200 1,230 1,250 1,270 1,300 1,300 1,300 1,300 1,300 1,300 World Total 7,496 7,455 7,505 7,475 7,560 7,645 7,705 7,640 7,670 7,710 7,640 7,660 Market Economies 7,309 7,268 7,315 7,275 7,360 7,435 7,490 7,420 7,445 7,480 7,405 7,420 /a World Total minus China. Source: The World Bank. - 131 - Alternative Supply Projection - Case C 229. An alternative scenario regarding mine capacity suggests that total capacity increases during the 1985-90 period somehow fall short of what is envisaged in the "base case". Thus, in Case C 1/ it is assumed that mine capacity in the United States and Zambia declines more rapidly and that capacity expansion in Chile proceeds more slowly than in the base case (Table 8.10). With all macroeconomic and other assumptions remaining the same as in the base case (Case A) the Case C mine capacity scenario results in somewhat higher prices for 1990 and for 1995, hence somewhat lower consumption. The price projected in 1984 constant dollars would be 68C/lb in 1990 and 78¢/lb in 1995, as compared with 63¢/lb and 74c/lb, respectively, in the base case. 230. The share of secondary copper in total refined copper production is expected to rise modestly. When the long-run rate of demand growth declines from one period to the next, the share of secondary metal (made from scrap) in total metal consumption will probably rise. 2/ The share of secondary copper in world refined copper production is expected to rise from about 15% in 1983 to 18-19% by 1995. 231. The Role of State Companies: Despite the closure of a number of private mines in the United States, the role of state-owned copper companies in world copper supply is not likely to increase. As discussed in ANNEX I, Chile's CODELCO is likely to contribute significantly to expansion of world capacity, but Chilean private mines are also likely to expand at a similar rate. CODELCO has by far the largest potential for low cost expansion. Outside Chile, however, few state-owned copper mines are expected to expand capacity. Moreover, a significant reduction in Zambia's capacity will likely contain state-owned companies' aggregate expansion worldwide. 232. If copper consumption happens to fall rather than increase as forecasted, the copper industry could experience a further dramatic change in structure, with a growing preponderance of state production, because additional capacity cuts will come from privately-owned mines. With the exception of Zambia, state-owned companies are generally at the low end of the cost curve and would be in a much better position to survive without any government support. 233. The Role of Oil Companies: The possible future role of copper mines controlled by oil companies is unclear. No clear-cut pattern has evolved out of the experience of the incursion of such companies in the copper industry. In the recent years of depressed market prices, some of these companies were 1/ Case B is the low economic growth case with "base case" capacity projections (already discussed). 2/ See Marian Radetzki and Carl Van Duyne, "The Demand for Scrap and Primary Metal Ores After a Decline in Secular Growth," Canadian Journal of Economics, XVIII, no. 2 (May 1985), pp. 435-449. - 132 - quick to cut production and losses, while others persisted and sustained losses that would have been unbearable without the financial backing of the parent oil companies. While some oil companies have divested or seem determined to divest themselves of their copper interests, others are poised to expand them. Although oil companies still have the resources for a strong potential influence on the copper industry, it is increasingly evident that copper is not viewed as an essential business for them. Their recent behavioral pattern suggests that: (a) they are quick to withdraw from uneconomic projects, (b) they are internationally orientated and (c) they are ready to inject their funds into promising projects wherever they can be found. The closures and divestitures of copper producing units by oil companies have occurred primarily in the United States. Oil companies have retained their interests in promising copper properties located in such countries as Australia, Chile, Peru and Papua New Guinea. 1/ E. Price Outlook 234. Given the macroeconomic assumptions (GDP growth, inflation, exchange rates, etc.) in the base case and the most likely developments on the mine capacity side, the LME price of copper is likely to remain flat in constant dollar terms during the rest of the 1980s, but to increase moderately in the 1990s. The projected slow growth in capacity reflects partly the investment response to the low prices in the 1980s. 235. In the Case B scenario, where economic growth is presumed to average 20% below what is assumed under the base case, the price in 1995 would be 58C/lb in 1984 dollars, or down by 22% compared with the base case projection. 236. In the Case C scenario, where mine capacity does not increase as rapidly as envisaged in the base case, the price could rise to 68¢/lb. in 1984 dollars by 1990 and 78-80C/lb by 1995. The implication of the Case C scenario for the industry is important. If the growth of world mine capacity is held back somehow in the next few years, the implied improvement in the price would be substantial. Comparing Case C with Case A (base case), mine capacity in 1990 and 1995 under Case C is, respectively, 3.4% and 4.5% less than in Case A (base case) and the price in 1990 and 1995 under Case C would be 6-7% higher than in Case A. 237. In the base case projections, the US producer price is assumed to remain 7% higher than the LME price throughout the projection period. However, given the changing market structure and the resulting competitive pressures, the US producer price could be forced to move towards the LME price for the entire projection period. In that event, the LME price for 1990 and 1995 would be about 4% higher than the prices projected in the base case (Case A). 1/ Annex D discusses the role of oil companies' involvement in the world copper industry. - 133 - 238. Copper prices over the next ten years may be expected to be both lower in real terms and less volatile than they have been over the last two decades. Reopening retired mines and facilities takes less time and cost than developing new mines. Over the next ten years, most of the mines already closed would remain closed if prices behave as forecasted. Some of these, however, could be brought back into production if prices start rising above the projections, thus moderating any price increase. Indeed, for prices to remain long at a level significantly above the forecast, it would be necessary for consumption to rise considerably faster than expected and for output to experience significant shortfalls (a situation that could arise if a large producing country such as Chile, where an increasing share of world copper is being produced, experiences protracted production difficulties). 239. Future industry production and investment decisions require great attention to cost competitiveness to generate earnings and avoid possible large losses. The various scenarios provided by the model underline the vulnerability of the industry's future prospects to world economic trends. Over the past decade, the restructuring of the industry has occurred in the face of persistent over-supply. It has resulted in shifts in the geographical location of the industry, as well as changes in industry ownership structure, including divestment of copper (and other) mining operations by multi-national parent companies. 240. A possible turnaround in the industry's financial fortunes is indicated if a sustained economic recovery occurs or a large segment of high cost producers close permanently. However, it is possible that the industry will face several more years of intense competition and low prices if lower economic growth occurs and/or untimely investments are made. Under such circumstances, only low cost operations with cost conscious management will be able to remain profitable. State-owned and private producers alike will need to assess their competitive strategy carefully in the light of evolving demand trends. They will need contingency plans to minimize the costs of adjustments in the event that low prices persist. Finally, if the evolving initiatives among the copper producers and fabricators for cooperative market development for copper uses should prove successful, copper consumption growth could be moderately stepped up, resulting in somewhat higher prices in the projection period. - 134 - ANNEX A. COPPER CARTELS, 1870-1940 1/ 1. The history of the copper industry abounds with episodes of repeated attempts to form cartels, either to squeeze extra profits at times of shortages or to minimize financial losses at times of depressed markets. One of the earliest attempts was a loose combination of US producers formed in the 1870s to increase profits behind a tariff wall. Herfindahl declares, "...from the point of view of the US producers it is hard to see how this period and its restrictionism could be viewed as anything but a spectacular success." 2/ This was followed by another famous episode, the so-called Secretan operation in 1887-89. The French Societe Industrielle et Commerciale des Metaux controlled 80% of world copper supply by means of contracts with all major producers. When output by non-participants increased dramatically in response to the increase in prices (especially in the United States) the scheme collapsed. 2. Other notable instances of attempts to form a cartel before World War I include the American Producers' Association (active from 1892 to 1903), a European Producers' Committee collaborating with the above, supply management in 1904-05 by the Amalgamated Copper Company (a holding company) organized by Standard Oil which controlled about one-fifth of world copper production, alleged collusive actions that took place in 1906-07 and 1908, and the activities of the Copper Producers' Association in the United States in the 1909-15 period. Generally speaking, some of these attempts were successful in increasing prices for short periods (up to 2-3 years) but none were successful over longer periods. 3. At the end of World War I, with copper prices plunging and unwanted stocks piling up, copper demand collapsed. In the United States, the Copper Export Association (CEA) was formed in December 1918 under the 1918 Webb- Pomerene Act. This act gave an antitrust law exemption to export associations. In 1919 CEA members controlled 89% of US production and 65% of world production. CEA allocated export sales and engaged in more direct pooling activities. For all practical purposes, however, CEA broke up in 1923 when the copper market recovered to a "normal" condition. Herfindahl's judgement about the success of CEA is as follows: "That CEA engaged in a successful collusive restriction of production in 1921 and 1922 is hardly open 1/ This brief review of attempts at collusive control by copper producers is primarily based on the detailed studies by the U.S. Federal Trade Commission and Orris C. Herfindahl. U.S. Federal Trade Commission, Report on the Copper Industry, Part I, "The Copper Industry of the United States and International Copper Cartels," US GPO, 1947, and Orris C. Herfindahl, Copper Costs and Prices: 1890-1957, The Johns Hopkins Press, 1959. 2/ Orris C. Herfindahl, Op. cit., p. 71. - 135 - to question. Whether CEA was effectively influencing market price before 1921 is a more difficult question." 1/ 4. The years 1923 to 1925 appear to be free of output restrictions, but another producer organization, Copper Exporters, Inc. (CEI), was formed in 1926. It included non-US as well as US producers, and accounted for 63-65% of world output in the 1926-29 period. The general rules were that members would sell copper at a price suggested by the CEI committee according to agreed percentages of total export sales. CEI was active during the period 1926- 32. The impact of CEI on copper prices in the 1926-29 period is unclear. However, after 1929, in the face of unprecedented depressed market conditions, the degree of community of interest among CEI members that had existed was weakened considerably and it came to a de facto end in 1932. After 1929 the main burden of restricting production fell on the US members of CEI. 5. The extensive interventions in the US copper market in 1933-34 under the National Industrial Recovery Act was followed by an international copper cartel initiated by Roan Antelope (a Rhodesian subsidiary of The American Metal Co.) in 1935. It included only five members (Anaconda, Kennecott, Roan Antelope, Rhokana and Katanga) and two "co-operating" firms (Bor of Yugoslavia and Rio Tinto). Canadian producers and Cerro de Pasco in Peru were also "friendly" to the cartel. The members accounted for about one-half of world output. Herfindahl concludes: "The cartel clearly succeeded in getting some restriction of output, especially at certain periods; but looking at the whole period from 1935 to 1939, it appears that the effect of the cartel on price was not very great." 2/ l/ Herfindahl, op. cit., p. 95. 2/ Herfindahl, op. cit., p. 114. - 136 - ANNEX B. ADJUSTED NET COSTS OF COPPER PRODUCTION: METHOD OF CALCULATION 1. The precise method used to calculate the adjusted net production costs is presented below. In calculating the adjusted net production costs, the following formula was used: N' = m . (D + I ) + (I-m) . (D + I )F(t t t ~Rj Ft -C where N't = net costs in year t adjusted to the base-year exchange rate and the differential rates of inflation between the country in question and the United States in the period between the base year and the year t. Dt = direct cost in year t It = indirect cost in year t m = the Z share of foreign cost component in total gross production cost (1-m) = the Z share of local cost component Ft = the amount of interest included in the gross cost Ct = co/by-product credits Rt the actual exchange rate (in local currency per US$) Rt = the adjusted exchange rate for the year t Hlocal H US R t t t X local US H H e e where Htlocal = the value of the inflation index for the year t in the non-US country concerned Helocal = the value of the inflation index for the non-US country concerned for the year chosen as the "equilibrium" exchange rate year HtUS = the value of the US inflation index for the year t - 137 - He US = the value of the US inflation index for the designated "equilibrium" exchange rate year The reason for deducting interest costs and co/by-product credits without adjusting for exchange rates is that both of these are usually determined in U.S. dollars and can be assumed to be unchanged in US dollars regardless of exchange rates. The assumption would be broadly true only if the devaluing copper exporting countries are only minor producers of the co/byproducts. This assumption fails when these countries are significant producers of these co/byproducts--e.g., gold in South Africa, silver in Mexico and Peru, nickel in Canada, and cobalt in Zaire. - 138 - ANNEX C. THREE FORMS OF COPPER IN TRADE 1. Broadly speaking, three different markets for copper can be distinguished by the copper processing stage, i.e., the market for copper concentrates, for blister copper and for refined copper. The distinction is important in understanding the dynamic relationship between suppliers and customers in the international trade for copper. Each of these three categories of products has its own market structure in terms of (1) the degree of flexibility in demand and supply, (2) the degree of homogeneity of product and (3) the nature of suppliers and customers. Table C.A shows the characteristics in marketing of each product category in the international trade. 2. Flexibility in supply here refers to the degree of ease with which customers are normally able to acquire the product in the quantity desired and at the time desired. Flexibility in demand refers to the degree of ease with which suppliers are able to find customers for their products. The difference in behavior of suppliers and customers in each product market can be explained by the difference in flexibility in supply and demand of each product. Supply and demand for copper concentrates and blister copper are less flexible than those for refined copper because of the following factors; lower degrees of homogeneity in quality, lack of ready, daily auction markets, and mining producers' preference for downstream integration. 3. Product Hetrogeneity in Concentrates and Blister: Quality differences in copper concentrates and blister copper are distinctly greater than in refined copper. The mineral composition and assay of the copper concentrates and blister copper differ from producer to producer, depending on the geology of the mines and the type of processing facilities involved. Also, in blister copper, the size of product differs depending on the producers. Thus, product differentiation among producers is stronger in the copper concentrates and blister copper markets than in refined copper markets. Copper concentrates and blister copper coming from different producers are not perfectly interchangeable. Customers tend to have a preference for certain types of copper concentrates or blister over other types. 4. Product Homogeneity in Refined Copper: In contrast to concentrates and blister, refined copper is more standarized in quality and size. Refined copper is almost always easily interchangeable among different brands. The high degree of homogeneity of refined copper provides its customers a greater flexibility in securing desired quantities, while the low degree of homogeneity in copper concentrates and blister gives the customers less flexibility in obtaining required materials. 5. Lack of Terminal Markets for Concentrates and Blister: For refined copper, there are efficient terminal markets such as LME and COMEX which are daily auction markets. There terminal markets can adjust supply and demand - 139 - quickly. In contrast, for copper concentrates and blister copper, there are no such terminal markets and adjustment of any market imbalance tends to take time. 6. Production capacity is difficult to adjust in the short run. The inflexibility in supply and demand of copper concentrates and blister copper tends to drive the producers and consumers in these products to seek long-term contracts for sales-purchase of these products. In contrast, the existence of terminal markets for refined copper gives both the producers and consumers a feeling that they can sell or buy the product in the terminal markets any time they wish. 7. Ore Producers' Preference for Integrated Operation: If not for financial constraints, ore producers would be interested in installing their own smelters and refineries to increase the value-added on their products. Governments in developing countries are interested in promoting local processing to increase local value added and foreign exchange earnings. Downstream processing by ore producers has slowed recently because of increased investment cost for new smelters and refineries, but the long-term interest in downstream processing still exists. This promotion of local processing in ore producing countries is perceived as a constant threat to the security of supply for the existing users of copper concentrates (custom smelters) and, to a lesser extent, for the users of blister copper (refining companies), as well, although blister copper can be substituted by scrap copper to some extent. 8. The primary interest of custom smelting and refining companies is to maximize their capacity utilization at all times in order to minimize the per- unit cost of smelter/refinery operations because these are highly capital- intensive activities with relatively low variable costs. Securing stable, long-term supply of copper concentrates as feed materials for their smelting/refining facilities is of utmost concern to the custom smelters/refineries. Therefore, regardless of the short-term market situation in refined copper, custom smelters are interested in developing new copper deposits and/or in rehabilitating the existing and closed mines in order to secure minimum requirement of copper concentrates for their smelter operation. In order to ensure the security of supply of material, they are often willing to extend financial assistance to suppliers. In this sense, the relationship between suppliers of copper concentrates and custom smelters is interdependent, while the relationship between suppliers of refined copper and its customers is more independent. 9. Implications of the interdependence of non-integrated ore producers and custom smelters for the future supply of copper can be summarized as follows. First, custom smelters tend to promote development of new mining projects through extending financial assistance and/or offering long-term purchase contracts for the output. Japanese participation in the development of the Escondida mine in Chile is an example. A Japanese custom smelters' consortium organized by Mitsubishi Corporation, a general trading firm, has a strong interest in developing the Escondida mine because of their interest in - 140 - securing a long-term supply source of copper concentrates, and has taken an equity interest and is prepared to engage in a "loan-purchase" arrangement. 10. Second, custom smelters are interested in rehabilitating some of the closed mines and in rationalizing the operations by providing technical, managerial and financial assistance. The Sumitomo group's participation in the modernization of the Morenci mine in Arizona is an example of this. Sumitomo Metal Mining and Sumitomo Corporation decided to buy 15% interest in Phelps Dodge's Morenci mine and to extend technical and financial assistance in this project in order to get a share of copper concentrates produced at this mine. The deal offers benefits to both parties. The Sumitomo group secures a long-term source of copper concentrates for their smelting operations without a long gestation period usually associated with new mine development. Phelps Dodge Corp. has obtained necessary financing for modernization of the mine along with efficient technical expertise of Sumotomo Metal Mining. 11. Third, high cost integrated copper producers have been in a difficult situation because fabricators in their purchases of standardized refined copper are indifferent between integrated copper producers and custom smelters/refineries. Table C.1: MARKETING CHARACTERISTICS FOR MAIN PRODUCTS OF COPPER Flexibility Flexibility Degree of Product Supplier Customer in Supply in Demand Homogeneity Copper Concentrates o Mines without smelting or o Custom smelters with Low Low Low refining facilities smelting and refining facilities Blister Copper o Mines with smelting o Custom smelters with Low Low Low facilities and without smelting and refining refining facilities; facilities; o Mines with excess capacity o Custom refineries; in smelting over refining o Fabricators o Semifabricators Refined Copper o Intregrated copper producers; o Fabricators High High High o Custom smelters with smelting o Semifabricators and refining facilities; o Custom refineries Source: The World Bank, Industry Department. - 142 - ANNEX D. OIL COMPANIES' BEHAVIOR IN THE COPPER INDUSTRY 1. The energy-related companies' entry into the copper industry accelerated the changes in the organization of the world copper industry in the 1970s. The main motives for involvement with the copper industry were to diversify their operations and to reinvest their huge cash surpluses earned from the sharp increase in oil prices. Oil companies selected the non-fuel minerals industry as a target industry. They viewed the copper industry as similar to the oil industry. Both industries deal with depletable natural resources, deal with the same types of government agencies, and involve the same types of public issues and exposure. Also, oil companies may have been confident that they could restructure and improve the ailing and backward copper enterprises to make those operations profitable through injections of fresh capital and efficient management skills already proved by the historically sound operations in oil. A growing perception of the increased scarcity of natural resources, caused by intensified nationalism in developing countries and the "Club of Rome" projections, probably accelerated the oil companies' diversification into the copper industry in North America. 2. With these motives, oil companies, with their abundant cash, were expected to expand copper production capacities. Following their initial entry into the copper industry, however, most companies have made few additional investments in the acquired mines. One important exception is EXXON's acquisition of Disputada copper mines in Chile. EXXON acquired the previously nationalized Disputada operation from the Chilean Government in 1978 at a cost of $110 million. EXXON injected additional funds to restructure Disputada operations and achieved a substantial increase in production in 1983. Output of copper in concentrate rose to 59 thousand tons, compared with 26 thousand tons in 1978 when EXXON acquired the mines. Furthermore, EXXON is pursuing an expansion program for Disputada's El Soldad mine, increasing its production capacity by 500% to 30,000 mtpy of contained copper. 3. In contrast to EXXON's Disputada, oil companies found that many acequired US mines were not viable. They have closed them or sold them. In addition, new properties, such as EXXON's Crondon ore-body in Wisconsin, remain undeveloped. Whether the mines closed by oil companies in the United States could have continued to be in operation had they remained under independent copper companies' management is open to question. Also in question is whether the copper mines acquired by oil companies are basically uneconomical. As a result of the closing of many copper mines owned by oil companies during the 1980-84 period, both the capacity and production of copper decreased. Table D.1 shows the major mine closures (including temporary closures) made in the United States by oil companies during the 1980-84 period. 4. As mine closures increased in recent years, a new trend emerged. Oil companies have been trying to sell off their copper units entirely and leave the copper business. Occidental Petroleum quickly sold off the copper operations it acquired when it took over Cities Services. Other oil companies - 143 - Table D.1: COPPER MINE CLOSURES BY OIL COMPANIES IN THE UNITED STATES, 1981-1985 Subsidiary Oil Company Copper Company Closed Mine ARCO Anaconda Carr Fork Berkeley Anamax Twin Buttes Pennzoil Duval Esperanza Mineral Park Sherrita Louisiana Land & Copper Range White Pine Exploration Co. AMOCO Cyprus Pima Bagdad Johnson Sohio Kennecott Chino Ray Bingham Occidental Cities Services Copper Cities Pinto Valley Quintana Quintana Copper Flat Source: The World Bank, Industry Department. - 144 - in search of buyers for their copper units include Pennzoil, Arco and Louisiana Land. Also, Texaco, following its takeover of Getty, expressed its intention to sell off all of its mineral divisions, and has already sold Getty's 50% interest in the Econdida copper property in Chile to BHP. 5. Mine closures and divestitures of copper units by oil companies have taken place primarily in the United States. Oil companies have retained their interests in copper mines located in countries having comparative advantage in copper production. For example, AMOCO maintains its 30% interest in the proposed OK Tedi copper-gold II project located in Papua New Guinea, although it closed uneconomic copper mines located in the United States. Other examples of oil companies retaining interest in copper properties are as follows: o British Petroleum (BP), with 47% interest, has been operating Tentonic Bore copper, zinc and silver deposit in Western Australia. Also, BP has 49.5% interest in Western Mining Corp's Olympic Dam copper, uranium and gold project at Roxby Downs, Australia and has been actively preparing its development. o Royal Dutch Shell has retained through Billiton an 11% stake in the profitable Cuajone copper project in Peru since 1976. o EXXON hs retained 100% interest in the Desputada mines in Chile and plans to expand the mines substantially. 6. Some differences between oil companies involved in the copper industry and the traditional copper mining companies are worth noting. First, Oil companies seem more prepared to "bite the bullet", pulling out of uneconomic operations and projects more quickly than traditional copper mining companies. This difference in behavior may be explained by the nature of the oil business. For oil wells, the cost and risk of exploration and development are greater than in mining of hard minerals such as copper. Therefore, it is essential for the management to decide on withdrawal before substantial expenditures are lost. Also, cutting back and shutting down operations is easier in the oil business than in mining. In the oil business, production from a gas or oil well can be cut simply by turning and shutting a few valves. Decisions on mine closures, however, are usually associated with higher costs and impaired stockholders' assets and must take these matters into account. Thus, management in oil companies is more accustomed to making decisions for production cuts or withdrawal than management in traditional independent copper companies. 7. Second, oil companies are prepared to inject huge amounts of investment funds into promising projects/units, as seen in BP's case. Copper mining projects owned by oil companies, however, must compete for funding with investment proposals for oil and gas projects within the same companies. These factors, along with the fact that copper mining is, after all, not the essential business of the oil companies, may have accelerated the - 145 - closures of economically marginal mines and may have kept oil companies from modernizing the copper industry in the face of deteriorated market prospects for copper. 8. Large oil companies are oriented more toward internation concerns than are many copper producers, especially copper producers in the United States. Before the nationalization of their mining properties in Chile and Peru, US copper producers had operated in international arena. However, since the nationalization of these properties, US copper producers have generally concentrated on domestic operations. In contrast, oil companies, many of which are based in the United States, have sought additional mineral interests in countries such as Chile and Australia where the potential in copper production seemed more attractive. 9. In conclusion, oil companies' participation in the US copper industry seems to have accelerated the weeding-out of economically marginal mines in the United States. However, oil companies do not seem to have contributed substantially to the restructuring and improvement of the copper industry. When oil companies increased their involvement in the copper industry in the 1970s they possessed a financial strength that could have been used for modernizing the copper operations and/or for developing new copper mines. While oil companies still have the financial resources to exercise a strong influence in shaping the copper industry, it is becoming increasingly evident that they do not view copper as an essential business and their past patterns of operation are likely to continue, i.e., (a) to be quick to withdraw from economically marginal projects, (b) to maintain an international orientation and (c) to be ready to invest funds in promising projects wherever and whenever these are identified. - 146 - ANNEX E: IMPACT OF CHANGING VALUE OF US DOLLAR ON COPPER PRICES 1. The annual average price of copper at LME in US$/ton (current $) is hypothesized to be a function of (a) the ratio of the level of commercial stocks in the Western world at the beginning of the year to Western world consumption of refined copper (adjusted for net trade with CPEs and changes in the US Government stockpile and the Japanese Government-financed, industry stockpile) during the year, (b) an index of inflation and (c) the annual average exchange rate of the US dollar vis-a-vis a basket of major currencies. The results of regression analysis using two different inflation indexes and three different measures of relative exchange rates of the US dollar are shown in Table E.1. Several different combinations of inflation variables and exchange rate variables have been tried. The period covered is basically 1968-84, but in some cases, the period 1968-83 was used. All equations are in natural-log-linear form, and no dummy variable is used. The results show that all independent variables used are highly significant, as indicated by the t-ratios. 2. Inflation index: In addition to the US GNP deflator, the weighted average index of the industrial countries' national currency GDP deflators has also been used. Using the latter instead of the former raises the elasticity coefficient to around 0.57-0.59, up from 0.43 for the US GNP deflator. These results suggest that only 50-60% of the inflation in industrial countries is reflected in the US$ price of copper within each year. 3. Exchange rates: In addition to the US$ to SDR rate, two other exchange rate indexes have been used--i.e., the exchange rate of the US dollar to a basket of major currencies (currencies of United States, Canada, United Kingdom, France, Germany, Italy and Japan) weighted by total merchandise imports of these countries in 1979-81 and the exchange rate of the dollar vis- a-vis the basket of currencies excluding US$. What is significant is that the estimated elasticity coefficients for the US$ exchange rates are all above 1.0, the lowest estimate being 1.07-1.09. 4. Theoretically, currency appreciation raises the foreign price and lowers the home price of traded goods, with the extent of this impact depending on the relationships of the home to foreign market size, and of home and foreign elasticities of supply and demand for the goods concerned. Under free market assumptions, the size of the elasticity coefficient for the exchange rate variable should be equal to or less than unity. It is possible that the particular formulation of the regression equations may be causing the exchange rate variables to pick up the effects of some extraneous factor(s); one plausible interpretation is that, in the case of copper during the data period covered, there was a tendency for the copper price to react speculatively to changes in the relative value of the US dollar, i.e., over- shooting. Table E.l: COPPER PRICE EQUATIONS /a Inflation Exchange Rate Stock US GNP National USS to US$ to US$ to Consumption Deflator Currency SDR Major Major Ratio /b GDP Currencies Currencies Price Estimation Deflator /c (excl. US$) /d (inml. US$) /e Data for 1984 (Variable Notation) Constant (CSTXCONN2) (USGNPDFL) (ICGDPFNC) (USSDR) (ERCOM2) (ERCOM1) Period R2 D.W. c/lb. % error /f Equation (1) 4.244 -0.362 0.432 -1.903 1968-83 0.85 2.45 60.0 -3.9% (6.99) (-4.97) (4.00) (-5.35) Equation (2) 4.300 -0.390 0.574 -1.085 1968-83 0.72 2.32 60.5 -3.1% (5.77) (-3.96) (4.05) (-3.38) Equation (3) 4.282 -0.391 0.576 -1.411 1969-83 0.71 2.31 60.2 -3.7% (5.72) (-3.94) (4.047) (-3.35) Equation (4) 4.249 -0.387 0.586 -1.072 1968-84 0.73 2.43 61.5 -1.5% (6.50) (-4.29) (4.82) (-4.445) a- Equation (5) 4.255 -0.388 0.590 -1.386 1968-84 0.73 2.41 61.3 -1.8% (6.42) (-4.268) (4.82) (-4.409) Equation (6) 3.808 -0.362 0.615 -1.094 1968-84 0.72 2.37 61.2 -2.0% (4.95) (-4.065) (4.65) (4.42) Equation (7) 3.780 -0.362 0.619 -1.415 1968-84 0.72 2.35 61.0 -2.4Z (4.88) (-4.05) (4.654) (-4.38) /a All equations estimated in log-linear form. The dependent variable is the price of copper in London in current US$ per ton. Figures in parentheses are t-ratios. 7n Comerical stocks at the beginning of the year, divided by refined copper consumption in the Western World adjusted for net trade with CPEs and for government stockpile releases. /c Industrial countries' national currency GDP deflator, provided by EPDGL. 7d US$ vs. the weighted average exchange rates of Canada, France, P.R. Germany, U.X., Italy and Japan; weighted by total imports of all merchandise in the period 1979- 81. /e US$ vs. the weighted average exchange rates of the countries mentioned in /d above plus the United States, weighted by total imports. 7f Actual 1984 price - 1,377.3 $/MT - 62.5¢/lb. Source: The World Bank, Economic Analysis & Projections Department. - 148 - 5. The correlation between inflation indexes, on the one hand, and exchange-rate indexes, on the other, is surprisingly low. The correlation matrix for the independent variables used is presented in Table E.2. 6. Interest rates: Interest rates were also considered, but their effects on the copper price have been found insignificant. Furthermore, the Eurodollar loan interest rate was found to be highly correlated with inflation indexes. Table E.2: CORRELATION MATRIX FOR INFLATION, EXCHANGE-RATE AND INTEREST-RATE VARIABLES /a ' RC USGNPDFL.I CGDPDFNC.USSDR JERCONI ERCON2,EURLON.CSTKCONN2 CORRELATION "ATRIX USCNPDFL ICGDPDFNC USSDR ERCONI ERCON2 USGNPOFL 1.60SE+BS ICGOPOFNC 9.990E-01 1.00SE+0S USSOR 2.759E-O1 2.989E-01 1.090E+09 ERCONI 1.072E-01 8.676E-02 -9.036E-01 1.000E+00 ERCOK2 1.060E-01 8.553E-02 -9.041E-01 1.B0SE+00 1.000E+00 EURLON 6.215E-01 6.038E-01 3.623E-01 -1.675E-01 -1.682E-01 CSTKCONN2 5.366E-01 5.704E-01 4.742E-01 -2.553E-01 -2.561E-01 EURLON CSTKCONN2 EURLON 1.SSSE+00 CSTKCONN2 -1.031E-01 1.00SE+90 /a Data period -- 1970-1984. Source: The World Bank, Economic Analysis & Projections Department. - 150 - ANNEX F. SUMMARY OF THE COPPER MODEL STRUCTURE 1. The econometric model from which some of the findings in this report and the long-term price projections have been derived concentrates essentially on the supply and demand aspects of the market. 1/ The model is disaggregated by major consuming and producing countries to permit analysis of behavioral patterns, mainly in response to prices. Since substitution for copper has been substantial over the historical period, and will no doubt persist in the future period, an analysis of copper consumption by end-uses for the six major industrial countries (including Japan) is attempted, albeit with imperfect data. The model also attempts to reflect the changing pattern of comparative advantage between the United States and the remaining producing countries; this is done by distinguishing the US producers' price from the LME price, and analysing US mine production in conjunction with the US copper trade. 2. A primary reason for constructing the world copper market model was to provide a framework for forecasting long-run prices under alternative scenarios. However, with the variance in perceptions that exist at this juncture about the structure of the copper market and its likely evolution in the forecast period, it was felt that copper investment functions estimated from historical data would not be relevant to the forecast period. Therefore, to allow for maximum flexibility in exploring alternative investment behavior in the producing countries, mine capacities for all countries are treated exogenously. Mine production, or supply, is estimated as a function of mine capacity and copper price. As the flowchart in Figure F.1 shows, separate supply equations are estimated for Australia, Canada, Chile, Mexico, Papua New Guinea, Peru, Philippines, South Africa, USA, Zaire, Zambia and the rest-of- the-Western-world. 3. The period for which the model is estimated is 1968-83. While this sample period is shorter than required for sound econometric analysis, it was used nevertheless for various reasons. For many of the major developing producing countries, political independence and nationalization of the copper industries had occurred in the mid-1960s. These developments affected the copper market structure. Up to the 1960s, the aluminum industry was also penetrating the copper market very rapidly. Analysis of copper demand by end- uses is therefore necessary to understand the differential rates of substitution between copper and aluminum in different industries in the historical period, as well as for exploring potential substitution by new products such as fiber optics in the future period. Such end-use copper consumption data are available only from 1968. 1/ For details of the model, see "A Study of the World Copper Market, 1964- 1995: An Econometric Analysis" prepared by C. Suan Tan, EPDCS, The World Bank, 1986. - 151 - Figure F.1: STRUCTURE OF THE WORLD COPPER MARKET MODEL - FLOWCHART MINE CAPACITY Autralia |Canda Chile |Mexico IPapua New Guinea Philippines Peru S Atrica USA I Zaire |Zambia Rest of Western World MINE PRODUCTION Australia Canada Chile I Mexico Papua New Guinea Philippines Peru S. Africa USA Zaire Zambia I Rest of Western World US Producers Exchange Rates; Stocks National Detlators Prmar Secondory rCopper Supply cp Rene USA | Non-USA . Prc - Prce I Western World ReineNet US Imports C tndtreProduction I b d CopE e ReIned CoPPer a i liabeliuy V al Governmen Wor Reie Cope Cosmto Ke, EPigeeous Variables W E il s Wol 9S6ke30477 Ao 191 EI ICS - 152 - 4. Since the reference price for copper used in the international market is for refined copper, mine production must be translated into refined copper supply for use in the refined copper price formation function. This raises the question of the extent to which producing countries are involved in the refined copper trade. The United States stands apart from the other producing countries in that it is a major producing as well as consuming country, and refines the bulk of its mine output domestically. Consequently, the conversion of US mine output into primary refined copper is treated independently. The non-US mine production is aggregated before conversion to non-US primary refined copper output. Similarly, secondary scrap recovery is differentiated into US and non-US scrap production. The total Western world refined copper supply is then the sum of primary refined supply and scrap supply. Since the LME price is influenced by the Western world availability of refined copper, the total supply in each period is adjusted for changes in the government stockpiles of refined copper and for the net refined copper imports from the Comecon countries. 5. In its treatment of the demand for refined copper, the model focuses on Western world demand only and treats copper consumption in the following countries separately: Brazil, Canada, China, Federal Republic of Germany, France, India, Italy, Republic of Korea, the United Kingdom and the United States. A rest-of-the-Western-world consumption equation captures the balance of consumption in the world (excluding CPEs). For each separately identified country, refined copper consumption is estimated as a function of the industrial production index, the copper price, the price of aluminum (which has been a major substitute) and a trend variable to capture material substitution towards/away from copper due to technological reasons. 6. To examine the actual and potential impact of technology and relative price changes on substitution, as between copper and fiber optics for example, it is necessary to analyse copper consumption by end-uses. This requires disaggregating consumption data by end-uses and the formation of the corresponding end-use industrial production indices. Data availability allowed end-use consumption analysis for only six countries: the Federal Republic of Germany, France, Italy, Japan, the United Kingdom and the United States. For the remaining countries, only total consumption is analysed. For each country, prices of copper and aluminum are first converted into domestic currencies before deflation by the national consumer price indices. 7. In any period the difference between refined production and consumption leads to a change in the total stocks level, which is an explanatory variable in the LME price formation process. Other determinants of the LME price are the ratio of consumption to world (excluding Comecon) copper availability and the real exchange rate between the Pound Sterling and the US dollar. 8. Once the LME price is determined, the domestic US producers' price and the scrap price are determined via price linkage equations. While the linkage of scrap price to the LME price is direct, that between the US producers' price and LME price is less direct. This is because in the period 1968-83 under study, the US copper industry has been gradually losing ground - 153 - as the leading producer. The changing comparative advantage between the United States and the major foreign producers is reflected in the changing composition of net US copper imports and manifested in the 1978 shift from independent determination of the US producers' price to determination on the basis of a premium over the New York Commodity Exchange's quotation. 9. Modelling the copper market for the purpose of price forecasting is difficult because of, inter alia, the changing position of the US copper industry vis-a-vis that of the rest of the world. The model is therefore structured so that a diversity of scenarios may be analysed. For example, on the question of US producers' price, the model allows price projections under alternative scenarios regarding future productivitt and production positions of US copper industry. Under a scenario where the US copper industry undertakes modernization to increase productivity, a return to the two-price system where the US producers' price differs significantly from the world price may be envisaged in a protectionist environment. Alternatively, should the US copper industry choose to move in the direction of further integration with the world market, a convergence of the US producers' price and LME price may be envisaged. - 154 - ANNEX C. CHINA'S COPPER DEMAND PROSPECTS Background 1. Since 1979, China has undertaken a major program of adjustment and reform aimed at improving the efficiency of the economic system. The program so far appears to have been successful as measured by the rapid productivity increases in various sectors. The World Bank staff believe that these changes will lay the foundation for more rapid economic growth in the period beyond the mid-1980s. 2. If this view of China's long-term economic future is accepted, a strong implication is that copper consumption will grow significantly. China's domestic production of copper is believed to have grown steadily over the last two decades. It is also believed to have a sizeable volume of reserves of copper deposits. Nevertheless, there are reasons to expect that China may opt to develop new capacity for copper production only slowly, at least in the next few years. 3. Despite its growing domestic supply of copper, China has long been a net importer of copper. Imports of copper have tended to rise over time, and indeed, increased sharply in 1983. Are China's imports of copper likely to continue to rise in the next decade or so? Or, was the sharp increase in China's copper imports in 1983 an aberration? In view of the fact that data on China's consumption, production and trade of copper are scarce and that whenever statistics are available they tend to be of doubtful reliability, special difficulties arise in arriving at a plausible range of projections for China's imports of copper. 1/ Historical Data 4. China's mine production, smelter production, refined production and refined consumption of copper in the period 1967-83 as estimated by Metallgeselschaft AG (MG) are shown in Table G.l. Estimates for refined production and consumption for 1983 have been added on the basis of data from World Bureau of Metal Statistics (WBMS). Imports of ore, blister and refined copper during the 1967-83 period, as estimated by WBMS, are also shown. Exports of refined copper can safely be ignored as they are less than 2,000 tons for most of the years concerned. Presumably, the difference between 1/ According to Marian Radetzki, personal communication, the trade practioners involved in non-ferrous metals trade with China seem to believe that what happend in 1983 was an exception. During his recent field trip to China, Radetzki was told by the officials of the China National Metals Imports & Exports Corporation, and the Central Planning Committee, that his statistics on non-ferrous metals (including copper) based on Metallgeselschaft AG. and World Bureau of Metal Statistics were significantly off the mark. Table G.1: CHINA: COPPER PRODUCTION, CONSUMPTION AND TRADE DATA, INDUSTRIAL PRODUCTION, ELECTRICITY GENERATION, 1967-1984 Mine Smelter Refined Refined Industrial Electric Ore Blister Refined Total Production Output Output Consumption Production Generation Imports Imports Imports Imports Year (MG) (MG) (MG) (MG) (GHYERBK) (WRLDBNK) (WBMS) (WBMS) (WBMS) (WBMS) '000 ton -000 ton 000 ton -000 ton 1975-100 000 Gwh '000 ton -000 ton -000 ton 000 ton 1967 90 90 100 150 NA NA 0.0 0.5 21.3 21.8 1968 90 90 90 160 NA NA 0.0 0.0 18.0 18.0 1969 85 85 110 150 NA NA 0.0 1.5 42.3 43.8 1970 100 100 120 180 NA NA 0.0 2.0 68.9 70.9 1971 110 110 150 210 74.3 138.4 0.0 24.2 58.0 82.2 1972 115 115 180 240 79.2 152.4 0.0 42.2 43.4 85.6 1973 120 120 200 270 86.7 166.8 0.0 45.4 90.3 135.7 1974 130 150 220 280 86.9 168.9 0.0 37.1 112.1 149.2 1975 140 150 230 300 100.0 195.8 8.6 36.6 36.9 82.1 1976 150 150 240 320 101.3 203.1 14.2 38.5 32.1 84.8 1977 155 150 260 330 115.8 229.4 15.0 40.0 37.4 92.4 1978 160 170 290 350 131.4 256.6 15.0 36.6 54.4 106.0 1979 160 170 290 360 142.6 281.9 14.6 53.8 134.2 202.6 1980 170 175 295 370 155.1 300.6 16.3 50.4 128.2 194.9 1981 175 190 300 370 161.5 309.3 20.4 30.0 107.4 157.8 1982 175 205 300 400 173.9 327.7 37.2 29.1 110.9 177.2 1983 185 210 314 400 NA NA NA NA 430.0 430.0 1984 NA NA NA NA NA NA NA NA 250.0 NA Sources: MG - Metallgesellschaft AG., World Metal Statistics, various issues. WBMS - World Bureau of Metal Statistics, World Metal Statistics, various issues. CHYERBK - Economic Information Agency, Statistical Yearbook of China, various issues. WBank - The World Bank. NA - Not available. - 156 - refined production and consumption should be roughly equal to imports of refined copper, but significant discrepancies are prevalent. The main reason could be that significant trade exists between China and the Eastern European Non-market Economies. Because of the doubts about these statistics, they should be regarded only as representing the broad orders of magnitude and trends. 5. Data for mining, smelting, refining and fabricating capacities in China are not available. Plans for future capacity increases at various stages of production are also not available. Projecting China's Copper Imports 6. Currently, refined copper consumption per capita is low. For 1982, it is estimated to be around 0.87 pounds. Per capita consumption in 1982 for a range of selected countries is shown in Table G.2 for comparison. The low per capita consumption of copper in China is particularly striking since China has maintained a relatively high share of heavy industry (which is relatively more copper-intensive) in GNP. 7. From the results of simple regression analysis, refined copper consumption in China appears to have been highly correlated with industrial production and electricity generation. The latter is considered to be a good proxy for electricity consumption. Data on these two variables for the period 1971-82 are shown in Table G.I. The regression results are shown below: LN(CUCONCH) = 3.5582 + 0.4473 LN (IPCH) (1) (20.56) (12.65) Adj. R2 = 0.93 DW = 0.53 LN(CUCONCH) = 2.336 + 0.6323 LN (EGCH) (2) (8.06) (11.77) Adj. R2 = 0.93 DW = 0.60 where: CUCONCH = Refined copper consumption in China IPCH = Industrial production index for China EGCH = Electricity generation in China data period = 1971-1982 (annual). - 157 - Table G.2: PER CAPITA CONSUNPTION OF COPPER, SELECTED COUNTRIES, 1982 Refined Per Capita GNP per Copper Copper Country Population Capita Consumption Consumption --Million-- --US$-- --000 tons- ---lbs.-- China 1,008.2 310 400.0 0.87 India 717.0 260 83.2 0.26 Philippines 50.7 820 3.6 0.16 Korea, Rep. 39.3 1,910 131.9 7.40 Brazil 126.3 2,240 249.3 4.33 Chile 11.3 2,210 32.8 6.29 Peru 17.4 1,310 21.0 2.66 Mexico 73.1 2,270 87.5 2.64 Portugal 10.1 2,450 14.2 3.10 France 54.4 11,680 419.0 16.98 Germany, F.R. 61.6 12,460 730.8 26.13 Japan 118.4 10,080 1,243.0 23.14 United States 231.5 13,160 1,664.1 15.85 United Kingdom 231.5 13,160 1,664.1 15.85 Sources: Population and GNP per Capita - The World Bank Atlas, 1985. Refined Copper Consumption - World Bureau of Metal Statistics. - 158 - 8. Using the "elasticities" thus estimated 1/ and projections of industrial production and electricity generation for China, a range of projections for China's copper consumption for the benchmark year of 1990 have been made and are shown in Table G.3. Assuming that the historical "elasticity" relationships hold over the projection period, 2/ and that the forecasts of industrial production growth and electricity generation used are in the reliabile ranges, the results indicate that China's copper consumption could grow at between 3.1 and 5.1 % per annum. These growth rates in turn imply growth from 400,000 tons in 1982 to 510,000-595,000 tons by 1990. 9. Is China's domestic copper supply likely to grow at a sufficiently rapid pace to keep up with this projected consumption growth? The historical data presented in Table F.1 show that in the period 1971-82, China's copper mine production and refined production grew at 4.3% and 6.5% per annum, respectively, while consumption grew at 6.0% per annum. If these past growth rates for mine and refined production could be contained, the answer to the above question would be positive. However, there are some indications that production capacity in China may not grow as fast as in the past. First, the growth of mine production appears to have slowed down markedly in recent years, to 2.6% per annum in the period 1976-82. Second, the government currently places high-priority on the development of iron/steel and energy sectors; the development of non-ferrous metal production capacity seems to receive lower priority. Third, available copper reserves in China are believed to be of relatively low quality. Finally, world market prices for copper have been very low for the last few years and are likely to remain so in the near future. Therefore, considering the relatively capital-intensive nature of copper production, the government may have decided that it would be more cost effective to import copper, either in refined form or in various fabricated forms. If these speculations are broadly realistic, China's imports of copper should grow significantly, although all imports will not necessarily be in refined form. 3/ 1/ In view of the very low values of Durbin-Watson statistics obtained, the estimated elasticity values are highly unreliable. Because of the unknown reliability of the data used, however, no attempt is made to further refine these estimates. 2/ For most developing countries, the intensity of copper use per unit of GNP could be assumed to increase as per capita GNP increases, up to a point. The elasticity of copper demand with respect to industrial production could therefore, be expected to increase for a low-income country such as China. 3/ Marian Radetzki argues that "...the growing deficit of China's mineral trade was a result of the country's comparative advantage in labor intensive activities, of the mediocre quality of its mineral resources, and of the fast growth of domestic mineral consumption." (memo, "China's Foreign Trade in Metal Minerals--Performance and Prospects," September, 1983). Table G.3: CHINA: ALTERNATIVE PROJECTIONS FOR INDUSTRIAL PRODUCTION, ELECTRICITY GENERATION AND COPPER CONSUMPTION, 1990 Alternative Scenarios for Industrial Production/Electricity Generation Projected Copper Consumption Growth Rate per Growth Rate per Level as of Annum for Annum for 1990 (in 1982 Actual 1982-1990 1982-1990 '000 tons) Industrial Production Index 174 (a) 5.0 3.1 510 (1975 = 100) (b) 8.2 5.1 595 Electricity Generation 328 (a) 5.8 3.7 535 ('000 Gwh) (b) 6.2 3.9 543 In Source: See Table D.1. - 160 - ANNEX H. IMPACT OF FIBER OPTICS ON THE COPPER MARKET, 1986-95 1. For many decades, copper wire has been the favored transmission medium for telephone communications systems. Although new technologies such as microwave, satellite and coaxial cables have been developed, copper wire's importance in telecommunications has not diminished. In 1983, copper consumption in telecommunication cables in Japan, the United States and Western and Southern Europe amounted to about 650,000 tons, representing 19.4% of the copper consumption in wires and cables and 9.5% of all copper consumed in these regions. 1/ 2. Optical fiber technology now offers a viable alternative to copper wire systems in telecommunications. Indeed, the optic fiber technology is claimed to be superior to copper-wire based technology: "...it offers a means of communications that is more powerful, more efficient, more accurate, and, in many applications, more economical than transmission over copper wire." 2/ Commercial application of fiber optics technology to telecommunications has taken place only in the last decade, but the applications of the technology have been developing rapidly. 3. British physicist John Tyndall is considered the founder of fiber- optic electronics; in 1870, he found that light will follow water into a container and flow out of it. In the 1950s major corporations began to study fiber optics. In 1968, Standard Telecom Labs in England developed silica glass with an extremely low transmission loss. The Corning Glass Works and Bell Telephone Labs in the United States and Nippon Sheet Glass Co. in Japan have since developed glass fibers with typical loss factors of only 1 to 2 decibels per kilometer. Since 1976, "...advancements have come so quickly that it is difficult to determine who developed what." 3/ In 1979, fiber optics was introduced into inter-office trunks. In 1983, fiber optics became the medium of choice for long-haul transmission. With the latest advances in both manufacturing and application techniques, optical systems are now considered far superior in performance and capacity to copper-wire-based systems in long- haul transmission and in high volume data transmission. A telecommunications 1/ Data referred to are based on a Brook Hunt report, in which end-use copper consumption data include direct use of scrap, which entails double counting to the extent that "new" scrap is recycled. 2/ Harvey Blustain, The Impact of Fiber Optics on the Copper Wire and Cable Industry, IGI Consulting, Inc., Boston, Mass., 1985, a multi-client study available to subscribers only, p. 1. 3/ Robert G. Seippel, Fiber Optics, Reston Publishing Co., Reston, Va., 1984, p. 18. - 161 - expert predicts that "...improvements in lasers, the use of longer wavelengths, and multiplexing techniques will increase still further the capabilities of optical fibers." 1/ 4. Economics of fiber optics applications in telecommunications has also improved dramatically, and as of 1984, fiber optics cables became more cost effective than copper wires on the basis of equivalent transmission capability. "All of the evidence indicates that optical fibers are a potentially cheaper transmission medium than copper wires. The advantages are particularly clear at longer distances or when there is a need for larger bandwidth... From its present 41C a fiber-meter, it is expected that fiber cable produced in the United States will go to 17C within two years and to 10C by 1990." 2/ Taking into account the fact that in long-haul transmission, optical fiber systems require less repeaters/boosters than copper-wire systems, it is certain that optical fiber systems will replace the copper installations in this application. 5. Fiber optic applications in the telephone network are not likely to proceed at an equal rate of penetration in all parts of the system. Telephone networks can conveniently be disaggregated into three general segments: the long-distance network, inter-office exchanges and the local feeder and distribution loop. In the long-distance network area, in view of the clear advantages of fiber optics technology the majority of new routes is likely to be in optical fiber cables. According to one recent estimate, if all new systems that are currently planned in the United States are completed, the total capacity of the long-distance network will experience a six-fold increase in the United States. 3/ However, this will have little impact on the copper wire/cable market as the long-haul network accounts for less than 3% of all copper in the telephone system. 6. Similarly, large portions of the inter-office trunking systems to be constructed in the future will be using optical fiber technology. "Beyond 1986, [copper] will be used only in short-distance, perhaps just intra- building, applications." 4/ 7. It is in the local loop where the predominant proportion of copper- based telecommunication cables is used. A Local loop consists of two sub- segments, i.e., the feeder loop, which extends from central exchanges to concentration and branching points and the distribution loop, which connects branching points and subscribers' homes or business offices. In the feeder loop, in which 70% of all copper-based telecommunication wire and cable is 1/ Harvey Blustain, op. cit., p. 13. 2/ Ibid. p. 26. 3/ Ibid., p. 76. 4/ Ibid, p. 81. - 162 - used, new additions are likely to be increasingly in optical fibers because, despite the estimated 10% higher cost, optical fiber cables have the advantages of unlimited bandwidth and much greater reliability. "The near- total migration to fiber in the feeder loop is a reality." 1/ Thus, the loss by copper in the feeder loop could be substantial. In the distribution loop segment of the local loop, in which 24% of all communication wire/cable is installed, there are a number of obstacles which militate against provision of optical fibers to residential customers, economics being the main problem at present. One educated guess is that in the United States only 1% of residential loops will use optical fibers by 1990 and at best 20% by 1995, and for business customers a 50% penetration rate on new lines may be envisaged by 1995. 2/ Thus, copper losses due to fiber optics in the distribution loop is not likely to be substantial, at least up to 1995. 8. The global copper model was used to simulate the potential impact of fiber optics on the copper market over the next decade. The base-case assumptions for the exogenous variables are as described in Chapter V of this report. The assumptions adopted about the penetration by fiber optics in the telecommunication cable market are the forecasts made by Brook Hunt and Associates (BHA) and Hugh Douglas (HD). 3/ 9. BHA expect that fiber optics will have displaced 40% of the copper consumed in the telecommunications industry by 1990. This concurs with the rates of copper displacement by fiber optics presented by Hugh Douglas (1982) for the telecommunications industry of the United States, the United Kingdom, France, Federal Republic of Germany, Italy and Japan. The year-by-year rates of displacement considered most likely in these six major industrial countries are as follows: 1/ Ibid, p. 90. 2/ Ibid., p. 93. 3/ Brook Hunt Associates, Copper Quarterly Service, XV, July 1984; and Hugh Douglas & Co., The Impact of Fiber Optics on Copper Wire Markets, a multi- client report, San Francisco, 1982. - 163 - Year Rate of FO Adoption (Z p.a.) 1983 3.2 1984 6.2 1985 10.0 1986 15.0 1987 20.1 1988 25.2 1989 32.0 1990 40.0 1991 46.0 1992 51.0 1993 58.0 1994 64.0 1995 70.0 10. In the end-use equations of the model, copper consumption in the telecommunications industry is included in the electrical sector. The rate of copper displacement by fiber optics should therefore be introduced as having an impact directly on copper consumption in the electrical sector. But this approach is not feasible because of the double-counting problem in the sectoral consumption data. To overcome this problem, total consumption is derived first and then adjusted for the displacement by fiber optics. 11. To estimate the volume of copper that will be displaced by fiber optics at the rates of displacement given in Hugh Douglas (1982) it is necessary to derive coefficients for the downward adjustment of copper consumption in the electrical sector. This is done by taking into account the copper consumption in the telecommunciations industry as a share of consumption in the entire electrical sector, and the electrical sector consumption as a share of total copper consumption. The adjustment coefficients used to calculate the reduced copper consumption due to fiber optics penetration are summarised in Table H.1 for the six major industrial countries where fiber optics are expected to make major inroads. 12. Under the assumptions of the base case GDP growth rates and the rates of fiber optics penetration given above, the model simulation indicates a loss for the copper market of about 230,000 tons by 1990 and a little over 300,000 tons by 1995. 1/ These estimates do not include any market losses in other industrial countries or in developing countries; technological "leap-frogging" is likely to occur in some of the faster-growing developing countries, as the 1/ These are equivalent to 4.3% and 5.6% of total copper consumption that would have been projected for the six countries for 1990 and 1995, respectively, under the assumption of no effects from fiber optics technology. Table H.1: COEFFICIENTS USED TO REDUCE TOTAL COPPER CONSUMPTION TO ACCOUNT FOR FIBER OPTICS PENETRATION United United Year States Kingdom Germany, F.R. France Italy Japan 1983 100.00 100.00 100.00 100.00 100.00 100.00 1984 99.91 99.50 99.48 99.12 98.89 99.36 1985 99.72 98.49 98.50 97.81 98.28 98.42 1986 99.45 97.19 97.04 96.04 95.91 97.16 1987 99.11 95.35 95.11 94.09 93.30 95.76 1988 98.71 93.29 92.97 92.28 90.74 94.37 1989 98.33 91.29 90.88 90.52 88.38 93.06 1990 97.96 89.43 88.93 89.55 86.43 91.93 1991 97.66 87.86 87.31 88.58 85.11 91.33 1992 97.45 86.82 86.21 88.20 84.33 90.88 1993 97.34 86.18 85.57 87.89 83.90 90.63 1994 97.27 85.82 82.10 87.73 83.69 90.52 1995 97.23 85.66 85.03 87.67 83.61 90.47 Source: Estimates by the World Bank, EPD, based on Hugh Douglas, The Impact of Fiber Optics on Copper Wire Markets, San Francisco, October (a multiclient study); see paras. 8-11 of the text of ANNEX H in this report. - 165 - penetration of optical fibers in these markets may be greater than otherwise estimated. Additional displacement of copper from these sources may be guesstimated to be in the range of up to 50 thousand tons in 1990 and 100 thousand tons in 1995. 13. The above-mentioned forecasts of likely losses of copper market directly due to applications of fiber optics technology to telecommunications are considerably lower than the estimates made earlier by outside organizations. Brook Hunt Associates quarterly copper service has made detailed references to the impact of this technology but has not made explicit estimates for likely losses of copper market. The report by Hugh Douglas & Co. (dated 1982) estimates the loss to be over 400,000 tons by 1990 and over one million tons by 1995. The report by IGI Consulting cited earlier suggests a loss of about 90,000 tons of copper market in the United States by 1990, plus substantial additional scrap copper that might be created as a result of the displacement of installed copper cables. - 166 - ANNEX I. LONG RUN MARGINAL COSTS IN COPPER PRODUCTION 1. As discussed in the main report, long-run marginal costs (LRMCs) are only one of the factors to be taken into account in forecasting mine capacity expansion on a country-by-country basis. Other key factors include investment climate, financial and institutional constraints, and expected copper prices. Also, LRMC can only be estimated within rather broad ranges for a given capacity expansion. 2. The purpose of this note is to explain the methodology for deriving the relevant LRMCs for the copper industry and how estimated LRMCs are used for mine capacity projections. A methodological discussion below will be followed by a description of the key factors to be considered in LRMC calculations. Next, how crude estimates of LRMCs can be used to project capacity expansion is illustrated in the context of two major copper producing countries. Estimation Methodology 3. LRMCs are estimated using the Average Incremental Cost (AIC) method. The LRMC for a potential project is estimated by first discounting (at the social rate of interest) all incremental costs (capital and operating) that will be incurred in the future for providing the estimated additional copper produced over a specified period (usually 20 years or more). These costs are then divided by the discounted value of incremental output over the period, i.e.: AIC = Present Value of Capital Plus Operating Costs Streams Present Value of Production Stream 4. A numerical example for an application of the AIC method is presented below. For a hypothetical greenfield copper mining project to produce concentrates, it is assumed that: (i) capital requirement per annual ton of capacity is $4,000; (ii) its construction takes 4 years, following 3 years of preparation (costing $150 per ton capacity) spread over time as shown in column (1) below; (iii) operating costs, at full capacity, including a provision for payment of taxes, are $800 per ton; (iv) full capacity is reached in three years after completion of mine construction; (v) replacement costs totalling $75 per ton per year starts three years after physical completion; and (vi) operations are expected to last for 23 years (from year 8). - 167 - AIC calculations would be as follows: (1) (2) (3) Capital Cost Operating incl. Replacement Costs /a Production Year (in US$ per ton) (in US$ per ton) (in tons) 1 40 2 50 3 60 4 600 5 1,200 6 1,400 100 7 800 300 8 - 500 0.5 9 600 0.7 10 - 700 0.9 11 75 800 1.0 I (through) I I l 30 75 800 1.0 Net Present Value 2,997 3,953 4.58 (at 10% discount) AIC = $2,997 + $3,953 = $1,517/ton 4.58 tons /a Includes preoperating costs. 5. If in addition to copper, $200 worth of byproducts are recovered, and sale of concentrates is at a $320 discount from refined copper (treatment and refining charges), the total AIC net of byproducts on a refined copper basis is $1,517 + 320 -$200 = $1,637/ton. This translates into C74.3/lb. Considerations for Application 6. There are other methods of calculating LRMCs, but for discrete investment projects such as those encountered in copper mining, the use of the AIC provides a convenient way of ironing out the "lumpiness" of investments through averaging out the investment and operating costs of the projects over the life of the project. 7. The LRMC approach is particularly useful for comparing investment proposals involving projects with widely different characteristics, such as the lengths of life of the mines, construction time required and learning curves (cases ranging from almost immediate achievement of full capacity - 168 - production to mines requiring three or four years before reaching full capacity production. In these circumstances, the mere comparison of capital costs per ton of copper capacity or of operating costs between projects can be misleading. 8. Major factors affecting LRMC include: (i) structural factors, such as geological and mining conditions (ore grade, stripping ratios, etc.), topography, availability of water and infrastructure; (ii) policy environment, such as exchange rate policy, taxation, import duties on inputs, labor legislation, mining and foreign investment codes and environmental regulations; (iii) institutional, such as experience of copper companies and relevant agencies, management independence and flexibility, availability of trained labor and labor union control and practices. 9. In practice, all of these factors cannot be considered separately and explicitly, and yet, they must be incorporated into the assessment of projected cost structures, production capacities and learning curves. This assessment in turn is based in good part on the operating and production cost history of copper production or mining in general in the countries concerned and, more specifically, of the individual companies expected to contribute to the contemplated production increases. Indeed, while some of the above- mentioned factors affect any project or producer in a country (such as the policy environment), there are a number of location- or company-specific considerations to be made in developing AIC and in assessing the mines' relative position in the world cost curves. Taking Account of Exchange Rates 10. Focus on historical cost and production performance requires an effort to identify deeply-rooted trends. One of the main problems in establishing such trends is the influence of substantial changes in exchange rates on cost data. The analysis in Chapter V, Section D, shows this clearly for the 1975-84 period. It was therefore decided to isolate the direct impact of the currency value fluctuations by adjusting cost data through the use of a 1984 real exchange rates (between the US dollar and the national currencies of copper-producing countries) for the entire period of analysis. 11. The choice of 1984 real exchange rates for the purpose of establishing the historical trends in production costs is admittedly arbitrary. It is generally accepted, however, that by 1984 actual exchange rates of many non-US copper producing countries had come closer to sustainable (from macro-economic point of view) levels compared with the levels prevailing a few years earlier when their national currencies were clearly overvalued in relation to the US currency. The 1984 exchange rates, therefore, may be taken as relatively reasonable yardsticks for establishing the relative cost position of most non-US producers. Although it might be argued that the 1984 exchange rates represent undervaluation of the currencies of some copper- producing countries against the US currency, further real devaluation might be necessary for many copper-exporting developing countries in view of the debt overhang and the poor copper price prospects. The developments in 1985 showed dramatic real devaluations for such important copper exporting developing - 169 - countries as Chile, Peru, Zaire and Zambia. Such devaluations should be viewed as a macro-economic imperative rather than as an effort primarily to reduce copper production costs (measured in US dollars) as is sometimes claimed. 12. LRMCs are higher than the net production costs shown in Chapter V, Section D as LRMCs should include depreciation charges. Nevertheless, the net costs presented in Table 5.1 in Chapter V may be assumed to reflect the relative positions of LRMCs of various copper-producing countries. Specific Applications of AIC 13. The high-cost countries such as the United States and Zambia are not expected to undertake any significant expansion of capacity. Moreover, the United States shows relatively high variable costs, which explains in part the projected fall in US production capacity in the period to 1990. The United States is expected to experience both closures of mines and development of new facilities at the same time. 14. While no large greenfield project is considered viable in the United States given the copper price prospects, a number of leaching operations (e.g., at San Manuel, Pinos Altos, Tyrone and Black Bird) can be brought on stream at low AIC (below US$0.50 per pound). Therefore, the United States does have a low LRMC, but the potential for further capacity expansion is limited, and the projected expansion is expected to be no more than the tonnage lost by mine closures resulting from ore reserve depletion and cutbacks due to high variable costs. Total capacity of the United States is likely to fall by 350,000 tpy to 1,100,000 tpy by the 1990s. 15. In contrast, some countries that lie at the low end of the 1984 net costs curves may be able to expand capacity at low cost. Chile, for example, shows relatively low net production costs and also faces relatively low LRMC, and is likely to increase its capacity substantially. 16. Potential for low cost expansion in Chile is believed to be substantial. The Chuquicamata mine of CODELCO is in the process of a major expansion (which could amount to as much as 250,000 tpy to bring its total capacity to 0.8 mtpy by the early 1990s). The AIC for expansion projects in Chile are expected to be well below the projected copper prices. However, the operating costs at some of the ongoing operations are likely to increase as the ore grades are expected to fall. In the case of Chuquicamata, since the very efforts aimed at arresting such cost increases are associated with the investments in expanding the total capacity, the expansion program for this mine can be considered firm despite perceived financial constraints (the Chuquicamata expansion program would require an investment of almost one billion dollars over the next five years for its full implementation). Additional low-cost expansions within the CODELCO group are expected at Andina and possibly El Teniente. - 170 - 17. In the Chilean private sector, expansions are expected to be carried by Disputada (at both the Los Bronces and El Soldado mines) and Mantos Blancos. AIC in these cases, however, would be just below the projected prices, making these expansions contingent on the investors' own perception of long-term market prospects. 18. Furthermore, two greenfield projects might come on stream in the private sector in Chile within the projection period; namely, Escondida and Cerro Colorado (with the capacity of about 200,000 and 65,000 tpy, contained copper, respectively). However, since their AICs are projected to be higher than the prices projected for the remainder of this decade and only moderately lower than the projected prices for 1995, the go-ahead here will depend, among other things, again on the investors' confidence in prospects for an improvement in the copper market. 19. Given the above considerations, it has been assumed for the base case that Chile would expand capacity by 300,000 tpy by 1990 and by an additional 200,000 tpy by 1995. Aside from considerations of financial requirements, Chile's capacity expansion is expected to be moderated by concern for overexpansion and for the possibility of driving international copper prices down to its LRMC level, thus reducing economic rents. These broader considerations and the need for harmonization of Chilean copper development policies with macroeconomic objectives should weigh particularly heavily in CODELCO's investment decisions. 20. Zaire's net production costs are low, but it is not expected to expand its capacity very much. This is because (a) there is little scope for expansion of existing capacity and (b) the market prospects for cobalt (a very important byproduct) are not good. - 171 - ANNEX J. THE SURVIVAL STRATEGY OF JAPANESE COPPER SMELTERS 1. This note reviews how major copper smelting companies in Japan have survived the prolonged stagnation in world copper market and succeeded in re- positioning themselves from a vulnerable to stable position in the last decade or so. 2. For two decades until the early 1970s, with demand for copper increasing greatly and with domestic mine output stagnating at less than 10% of domestic demand, demand for copper in Japan had increasingly been met by imports. In the 1960s and early 1970s, large scale coastal smelters were built in succession and the Japanese copper industry transformed itself from a mining industry to a custom smelting industry. As a result, by the early 1970s copper concentrates became the dominant form of copper imports in Japan. 3. Since the mid-1960s, Japanese copper smelters have dominated the world trade of copper concentrates, accounting for over 60% of world trade. An alternative to the massive expansion of smelting/refining capacity at home could have been to increase imports of refined copper, but this course was not taken. Table J.1 shows the trends in copper smelting and refining capacity and the self-sufficiency ratio of copper in Japan from 1955 to 1984. 4. Japanese copper smelters have been an important influence in the market for copper concentrates (and blister copper) as well as on the supply- side of the refined copper market. Under the Japanese Government's natural resource security policy, the copper smelting companies often provided long- term loans for development of new mines overseas. At the same time, they offered long-term purchase contracts for the output of these mines. The objective was to secure stable and adequate supply of copper concentrates as feed for their smelters at home, while minimizing their involvement in the risks associated with exploration and development of mines. 1/ These loan- purchase arrangements were instrumental in opening such well-known large-scale mines as the Bongainville mine in PNG, the Ertzberg (or Gunung Bijih) mine in Indonesia and the Lornex mine in Canada. Sogo Shosha (general trading companies) often played a key role in organizing consortia of smelting/refining companies and arranged financing for these agreements. (Generally speaking, they were successful in securing long-term supply sources of copper concentrates to feed their smelters. Similar developments also occurred for other non-ferrous metals (especially lead, zinc and nickel). Major copper smelting companies in Japan have also been heavily involved in smelting of other non-ferrous metals. 5. These non-ferrous metal smelting companies, however, faced several difficult problems in the 1970s; volatility of copper prices, the revaluation of the yen against the US dollar and sharp increases in energy costs. Because 1/ T.E. Kolenda, "Japan's Develop-For-Import Policy," Resources Policy, December 1985, pp. 257-266. Table J.1: COPPER SMELTING AND REFINING CAPACITY, PRODUCTION OF COPPER ORE, BLISTER AND REFINED COPPER AND REFINED COPPER CONSUMPTION IN JAPAN, 1955-1984 Consumption Ratio of Ore Capacity Production of Refined Production to Refined Year Smelting Refining Ore Blister Refined Copper Copper Consumption -- 000 tpy----- - ----- ------------------000 tons- ---------------- ---------%----------- 1955 73.0 82.0 113.3 105.4 69.3 1960 89.2 187.2 248.1 304.0 29.3 1965 107.1 337.4 365.7 427.5 25.1 1970 727.9 885.6 119.5 606.1 705.3 820.6 14.6 1973 979.9 1,107.6 91.3 1,000.5 905.8 1,201.8 7.6 1980 1,231.0 1,234.8 52.6 929.3 1,014.3 1,158.3 4.5 1984 1,444.6 1,204.8 43.3 929.0 935.2 1,368.3 3.2 Source: Smelting and Refining Capacity: Japan, Ministery of International Trade and Industry. Production and Consumption: World Bureau of Metal Statistics. - 173 - their profits are based mainly on the difference between contractual treatment charges and actual costs, they, as custom smelters, are not in the position to reap windfall profits even when copper prices rise. When copper prices fall, however, they suffer from decreased volume of business resulting from sluggish demand for their output. As the value added in smelting and refining process is relatively thin, these companies do not have capability to absorb large down-side changes in demand. 6. Because, in contracts, treatment and refining charges were generally fixed in US dollars, Japanese copper smelters suffered from sharp deterioration in their purchasing terms of copper concentrates (measured in yen) in the 1970s when the yen appreciated consistently against the dollar. The sharp increases in energy costs in the 1970s also put additional cost pressures on the Japanese copper smelters, threatening their survival. Indeed, the abrupt increases in energy prices had the effect of a double blow to them, through the consequent slowdown in demand for copper and the increased costs resulting from higher energy costs. 7. These adverse factors, added to the intrinsic weakness of being essentially custom smelters with thin value added, forced Japanese copper smelters to decide, on the one hand, to diversify away from custom smelting business into manufacturing activities, related and unrelated to non-ferrous metal smelting, seeking greater value added and greater stability in their operation, and, on the other hand, to step up their efforts to rationalize their mining and smelting operations. 8. Much of their diversification (even in areas seemingly unrelated to nonferrous smelting) has been achieved on the basis of the expertise and technologies acquired through their long experience in mining and smelting. Their new areas of activities have included production of high performance materials (such as hard metals, metallurgical powder products and precision cast/forged products), electronics materials (such as silicon, compound semiconductors, single-crystal oxides and semiconductor materials), and nuclear fuel related activities (production, transportation and waste management of nuclear fuel). The share of non-ferrous metal sales in total sales of these copper smelting companies has declined steadily. 1/ The declining trend in the share in total profits has been even more emphatic than that of the share in total sales. 9. Despite energetic efforts to diversify the scope of business, Japanese copper smelting companies have not given up on their copper mining and smelting operations. They consider these operations essential as the 1/ For instance, in the case of Mitsubishi Metal Mining Co., the share of metal smelting (copper, lead, zinc, tin, gold and silver) in the total sales declined from 72% to 64% in the 1976-81 period, while the share of copper smelting declined from 49% to 38%. During the same period the share of non-smelting business sales (metalworking, electronics components, nuclear energy equipment, etc.) increased from 28% to 36%. - 174 - basis for development technology useful for their diversification. They closed some uneconomic domestic mines, retaining a small number of mines for training and technological development benefits. In non-ferrous smelting operation, they increased productivity by introduction of automation and energy-efficient technologies, such as oxygen enrichment systems. Oxygen enrichment systems alone increased productivity in smelting by an estimated 20%. 10. The current strategy of Japanese copper smelting companies appears to be to try to maintain the volume of copper smelting (and mining) business at the present levels. To secure minimum quantities of copper concentrates to feed their existing smelting/refining facilities, they are likely to provide loan financing for and/or equity participation in the development of new mines and the rehabilitation of existing or closed mines abroad. In fact, several Japanese copper smelting companies and Mitsubishi Corporation recently made commitments to the possible development of the Escondida mine in Chile. Sumitomo Metal Mining Co. also decided to participate in the modernization of Phelps Dodge's Morenci mine in the United States, primarily for the purpose of securing sources of copper concentrates for their existing smelting capacity. - 175 - REFERENCES "1967 Study by Andean Region of Economic Affairs Department of Pan American Union", Engineering and Mining Journal, February 1967. Aluminum Association, The, Aluminum Statistical Review: Historical Supplement, 1982. , Aluminum Statistical Review, 1981, 1982, and 1983. American Metal Market, Daily Newspaper, various issues. Auty, Richard M., "Material Intensity of GDP," Resources Policy, December 1985, pp. 275-284. Barber, Charles, F., (ASARCO Inc.), "Economics of Copper in the Mid-80's, A Role for the Multilateral Financial Institutions", The 1983 Hamilton Lecture, Canadian Institute of Mining and Metallurgy, Toronto, Canada, December 8, 1983. Berk, Rhea, Howard Lax, William Prast, and Jack Scott of Atlantis, Inc., Aluminum: Profile of the Industry, McGraw-Hill, Inc., 1982. "Big Growth Seen for Crosslinked PE and PB in Hot-Water Pipe," Modern Plastics International, December 1981. Blustain, Harvey, The Impact of Fiber Optics on the Copper Wire and Cable Industry, a multi-client study by IGI Consulting, Inc., Boston, Mass., 1985. Bothworth, Barry P., and Robert Z. Lawrence, Commodity Prices and the New Inflation, Bookings Institution, 1982. Brook Hunt & Associates Ltd., Quarterly Service, Copper, Issue XV, July 1984. Brook Hunt & Associates Ltd., Quarterly Service, Copper, Issue XII, December 1982. Brook Hunt & Associates Ltd., Western World Copper Costs and Production Strategy 1975-85, 5 vols., London, December 1984 (a multiclient study for subscribers only). CIPEC, "Zambia ZCCM Rationalization," Quarterly Review, October-December 1985. - 176 - Chu, K.Y. and T.K. Morrison, "The 1981-82 Recession and Non-Oil Primary Commodity Prices," IMF Staff Papers, vol. 31, no. 1, March 1984, pp. 93- 140. CRU, Copper Studies, October 1984. Data Resources, Inc., U.S. Long-Term Review, Winter 1984-85. Economic Information Agency, Statistical Yearbook of China, Hong Kong, various issues. Enoch, C.A., and M. Panic, "Commodity Prices in the 1970s, Bank of England Quarterly Bulletin, March 1981, pp. 42-53. Fitzgerald, M.D. and Gerald Pollio, Copper: The Next Twenty Years, Chemical Bank, New York, November 1981. GATT, "Problems of Trade in Certain Natural Resource Products: Background Study on Copper and Copper Products," MDF/W/10, October 1984. Gluschke, Wolfgang, Joseph Shaw and Bension Varon, Copper: The Next Fifteen Years, A United Nations Study, D. Reidel Publishing Co., 1979. Gonzalez-Vigil, Fernando, "New Technologies, Industrial Restructuring and Changing Patterns of Metal Consumption", Raw Materials, vol. 3, no. 3. Grilli, E.R. and M.C. Yang, "Real and Monetary Determinants of Non-Oil Primary Commodity Price Movements," in J.A. Kregel (ed.), Distribution, Effective Demand and International Economic Relations, MacMillan, 1983. "Growing Challenge to Metal Pipe: CPVC and PB Pipe", Plastics World, September 1981. Herfindahl, Orris C., Copper Costs and Prices: 1980-1957, The Johns Hopkins Press, Baltimore, 1959. Hugh Douglas, & Co., The Impact of Fiber Optics on Copper Wire Markets, a multiclient report, San Francisco, 1982. International Monetary Fund, International Financial Statistics, various issues. Kolenda, T.E., "Japan's Develop-For-Import Policy," Resources Policy, December 1985, pp. 257-266. Lesemann, Robert H., (Commodities Research United Ltd.), "Copper -- Hugh Stock Buildup Will Slow Market Recovery," Engineering and Mining Journal, March 1976, pp. 88-91. - 177 - Lowell, J. David, "Copper Resources in 1970," the 1970 Jackling Award Lecture, The Transactions of the Society of Mining Engineers, AIME, June 1970, pp. 97-103. Malenbaum, Wilfred, World Demand for Raw Materials in 1985 and 2000, New York, McCraw-Hill, 1978. Mardones, Jose Luis, Enrique Silva and Cristian Martinez, "The Copper and Aluminum Industries," Resources Policy, vol. 11, no. 1, March 1985, pp. 3- 16. McMahon, A.D., Copper: A Materials Survey, USBM IC 8225, US Bureau of Mines, 1986. Metal Bulletin, various issues. Metallgesellschaft AG., World Metal Statistics, various issues. Mitsubishi Metal Mining Co., (Kenji should this be cited? if so, I need more of a reference. see page 166. National Research Council, National Materials Advisory Board, Committee on Technical Aspects of Critical and Strategic Materials, Mutual Substitutability of Aluminum and Copper, April 1972. Navin, Thomas R, Copper Mining and Management, Univerity of Arizona Press, 1978. OECD, Economics Outlook, December 1985. OECD, Economic Outlook Historical Statistics, 1960-1983, Paris, 1985. Phelps Dodge Corporation, "Free World Copper Mine Capacity," company internal document, various issues. Prain, Ronald (Sir), Copper: The Anatomy of an Industry, Mining Journal Books Ltd., London, 1975. Radetzki, Marian and Carl Van Duyne, "The Demand for Scrap and Primary Metal Ores After a Decline in Secular Growth," Canadian Journal of Economics, XVIII, no. 2, May 1985, pp. 435-449. Radetzki, Marian, China's Foreign Trade in Metal Minerals--Performance and Prospects, September 1983 (memo). Radetzki, Marian, "Effects of a Dollar Appreciation on Dollar Prices in International Commodity Markets," Resources Policy, September 1985, pp. 158-159. Radetzki, Marian, "Fluctuations in Invisible Stocks: A Problem for Copper Market Forecasting," Commodity Paper No. 27, The World Bank, October 1977. - 178 - Radetzki, Marian, State Mineral Enterprises: An Investigation into Their Impact on International Mineral Market, Resources for the Future, Washington, D.C., 1985. 1 Ridler, D. and C.A. Yandle, "A Simplified Method for Analyzing the Effects of Exchange Rate Changes on Exports of a Primary Commodity," IMF Staff Papers, 19, 1972, pp. 559-578. Tan, Suan C., A Study of the World Copper Market, 1964-1995: An Econometric Analysis, Commodity Working Paper No. 10, The World Bank. UNCTAD, Processing and Marketin of Copper: Areas for International Cooperation, TD/B/C.1/PSC/30, United Nations, 1982. United Nations, Monthly Bulletin of Statistics. United States Bureau of Census. United States Bureau of Mines, "Copper", Mineral Facts and Problems 1985, Washington, D.C., 1986. United States Council of Economic Advisers, The Annual Report, Feburary 1985 United States Trade Commission, Report on the Copper Industry, Part I, "The Copper Industry of the United States and International Copper Cartels," U.S. Government Printing Office, 1947. Vernon, Raymond, Two Hungry Giants, The United States and Japan in the Quest for Oil and Ores, Harvard University Press, 1983. Wagenhals, Gerhard, The World Copper Market, Structure and Econometric Model, Springer-Verlag, 1984. World Bank, The World Bank Atlas, 1985, Washington D.C. World Bank, World Development Report, 1985, Oxford University Press, New York, 1985. World Bureau of Metal Statistics, World Metal Statistics, various issues. "Zambia Bites the Bullet," Mining Journal, vol. 306, no. 7851, February 7, 1986, pp. 89-91. Zorn, Stephen A., "Producers' Associates and Commodity Markets: The Case of CIPEC," in F. Gerard Adams and Sonia A. Klein (eds.), Stabilizing World Commodity Markets, Lexington Books, 1978, pp. 215-234. - 179 - Distributors of World Bank Publicafions Algeria: Office des Publications Universitaires, 1, place Maldives: (See Sri lAnka) centrale de Ben-Aknoun, Algiers Mexico: INFoTEc, San Lorenzo 153-11. Col. del Valle, ArgeNtdia: Carlos Hirsch, SRL, Galeria Guemes, Florida Deleg. Benito Juarez, 03100 Mexico City 165, 4th Floor-Ofc. 453/465, 1333 Buenos Aires Morocco: Societe d'Etudes Marketing Marocaine, 2 Rue Anstrlia, Papua New Guinea, FQi, Solomon Islds, od Moliere, Bd. d'Anfa, Casablanca Vanuatu: Info-Line, Overseas Document Delivery, Box Netherlands: Medical Books Europe, BV (MBE), Noorderwal 506, GPO, Sydney, NSW 2001 38, 7241 BL Lochem Austria: Gerold and Co., A-101l Wien, Graben 31 New Zealad: R. Hill and Son, Ltd., Private Bag, New Babrain: MEMRB, P.O. Box 2750, Manama Town 317 Market, Auckland Bangladesh: Micro Industries Development Assistance Nigeria: University Press Limited, Three Crowns Building Society (MIDAS), G.P.O. Box 800, Dhaka Jericho, Private Mail Bag 5095, Ibadan Belium: Publications des Nations Unies, Av. du Roi 202, Norway: Tanum Karl Johan, A.S., P. 0. Box 1177 Sentrum, 1060 Brussels Oslo I Brazil: Publicacoes Tecnicas Internacionais Ltda., Rua Pdaan: Mirza Book Agency, 65, Shahrah-e-Quaid-e- Peixoto Gonide, 209, 01409 Sao Paulo, SP Azam, P.O. Box No. 729, Lahore 3 Burma: (See Singapore) Panaa: Ediciones Libreria Cultural, Panamena, S.A., Av. Canada: Le Diffuseur, C.P. 85, 1501 Ampere Street, 7, Espana 16, Panama Zone I Boucherville, Quebec, J4B 5E6 Pama New Guinea: (See Austraia) Chile: Editorial Renacimiento, Miraflores 354, Santiago Peru: Editorial Desarrollo SA, Apartado 3824, Lima Colombia: Enlace Ltda., Apartado Aero 34270, Bogota Pbilppines: National Book Store, 701 Rizal Avenue, Metro D.E.; Apartado Aereo 4430, Cali, Valle Manila Costa Rica: Libreria Trejos, Calle 11-13, Av. Fernandez Portugal: Livraria Portugal, Rua Do Carmo 70-74, 1200 Guell, San Jose Lisbon Cote d'lvoire: Centre d'Edition et de Diffusion Africaines Saudi Arbia: Jarir Book Store, P. 0. Box 3196, Riyadh (CEDA), 04 B.P. 541, Abidjan 04 Plateau 11471 Cyprus: MEMRB, P.O. Box 2098, Nicosia Sigapore, Taiwan, Burma: Information Publications, Denmark: SamfundsLitteratur, Rosenoerns Alle 11, Private, Ltd., 02-06 Ist Fl., Pei-Fu Industrial, Bldg., 24 DK-1970 Frederiksberg C. New Industrial Road, Singapore Egypt, Arb Republic of: Al Ahram, Al Galaa Street, Cairo Solomon Islads: (See AustrLia) Fiji: (See Australia) South Africa: Oxford University Press Southern Africa, Fmnlaod: Akateeminen Kirakauppa, P.O. Box 128, P.O. Box 1141, Cape Town 8000 SF-00101, Helsinki 10 France: World Bank Publications, 66, avenue d'lena, 75116 Spain: Mundi-Prensa Libros, S.A., Castello 37, 28001 Pads Madrid Germany, Federal Republic of: uNo-Verlag, D-5300 Bonn 1, Sri Links and the Madives: Lake House Bookshop, P.O. Simrockstrasse 23 Box 244, 100, Sir Chittampalam A. Gardiner Mawatha, Greece: KEME, 24, lppodamou Street, Athens-I 1635 Colombo 2 Hong Kong, Macau: Asia 2000 Ltd., 6 Fl., 146 Prince Sweden: For single tities: ABCE Fritzes Kungl. Edward Road, W, Kowloon, Hong Kong Hovbokhandel, Attn: Mr. Eide Segerback, Regeringsgatan Hungary: Kultura, P.O. Box 139, 1389 Budapest 62 12, Box 16356, S-103 27 Stockholm. For subscription lndia: uBs Publishers' Distributors Ltd., Post Box 7015, orders: Wennergten-Wilams AB, Box 30004, S-104 25 New Delhi 110002; 10 First Main Road, Gandhi Nagar Soko Bangalore 560009; Apeejay Chambers, P.O. Box 736, 5 Switterbad: Librairie Payot, 6, rue Grenus, Case postal Wallace Street, Bombay 400001, 8/1-B, Chowringhee Lane, 381, CH 1211 Geneva II Calcutta 700016; 7/188, I(CA), Swarup Nagar, Kampu TaIwan: (See Sigapore) 208002; Sivaganga Road, Nungambakkam, Madras 600034 Thzaais: Oxford University Press, P. 0. Box 5299, Dar es Indonesia: Pt. Indira Limited, JI. Sam Ratulangi 37, Jakarta Salaam Pusat, P.O. Box 181 Thailad: Central Department Store, 306 Silom Road, Ireland: TDC Publishers, 12 North Frederick Street, Bangkok Dublin I Trnidad and Tobao: Systematics Studies Unit, 55 Eastern Isrel: Jerusalem Post, Jerusalem Post Building, P.O. Box Main Road, Curepe, Trinidad, West Indies 81, Romena Jerusalem 91000 Tunisia: Societe Tunisienne de Diffusion, 5 Avenue de Italy: Licosa Commissionaria Sansoni SPA, Via Lamarmora Carthage, Tunis 45, Casella Postale 552, 50121 FRorence 4' . Turkey: Haset Kitapevi A.S., 469, Istiklal Caddesi, Japan: Eastern Book Service. 37-3. Hongo 3-Chome, Beyoglu-Istanbul Bunkyo-ku 113, Tokyo Bunky-ku 13, okyoUganda: Uganda Bookshop, P.O. Box 7145, Kampala Jordan: Jordan Center for Marketing Research. P.O. Box a B P 3143, Jabal, Amman United Arab Emirates: MEMRB Gulf Co., P. 0. Box 6097, Kenay: Africa Book Service (E.A.) Ltd., P. 0. Box 45245, Sharjah Nairobi United Kingdom: Microinfo Ltd., P. 0. Box 3. Alton, Korea, Republic of: Pan Korea Book Corporation, P. 0. Hampshire GU 34 2PG, England Box 101, Kwangwhamun, Seoul Vanuatu: (See Austalia) Kuwait, MEMRB, P.O. Box 5465 Venezuela: Libreria del Este, Aptdo. 60.337, Caracas Macau:MSee Hong Kong) 16- Maca: (Se Hng Kng)Western Samoa: Wesley Bookshop, P.O. Box 207, Apia Malaysia: University of Malaya, Cooperative Bookshop, i abwe: T e sles PvoksLtd, Box 37, Aa Limited. P. 0. Box 1127, Jalan Pantai Bari, Kuala Zimbabwe: Textbook Sales Pvt. Ltd., Box 3799. Harare Lumpur I I The World Bank Headquarters European Office Tokyo Office 1818 H Street, NW. 66, avenue d'lena Kokusai Building Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marunouchi 3-chome U Telephone: (202) 477-1234 Telephone: (1) 47.23.54.21 Chiyoda-ku, Tokyo 100, Japan Telex: WUI 64145 WORLDBANK Telex: 842-620628 Telephone: (03) 214-5001 RCA 248423 WORLDBK Telex: 781-26838 Cable Address: INTBAFRAD WASHINGTONDC [-ui-J