This volume contains the proceedings of the conference on ‘Primary Commodity Prices: Economic Models and Policy’, organized by the Centre for Economic Policy Research and held in London on 16/17 March 1989. It formed part of the Centre's programme in International Trade and was supported by Rio Tinto Zinc pic, the European Investment Bank, the Foreign and Commonwealth Office and the Overseas Development Administration. We are extremely grateful to our sponsors for both their financial support and their advice and encouragement in drawing up the programme.

We also wish to thank Alasdair Smith and Richard Portes for encouraging us to run the conference, Wendy Thompson for managing the business arrangements, Andreas Köttering for acting as rapporteur and Jacqui Eggo for ensuring that the conference ran smoothly. Paul Compton, the Centre's Publications Officer, guided this volume to press rapidly and efficiently; he and we were pleased to work with John Black of the University of Exeter, who served as Production Editor. They have kept editors, authors and discussants to CEPR's very tight production schedule, so that these timely, policy-relevant analyses by leading researchers would quickly reach a wide audience.

Conferences are cooperative efforts. We trust that our colleagues, sponsors and readers will find the volume as worthwhile as we have. We believe that the research reported here will communicate to readers the latest thinking on commodity price modelling, encourage researchers to explore new and stimulating paths, and inform current policy debates on price and earnings stabilization.

Introduction

The instability of world commodity markets has been a central issue in the North-South dialogue for more than two decades now. The proponents of policy intervention argue that excessive market price fluctuations cause negative micro- and macro-economic consequences which can be avoided by stabilization measures. Price stabilization policies, based on buffer stocks as intended by the Integrated Programme on Commodities (IPC) or export quotas, and compensatory financing schemes were primarily discussed as international measures to stabilize export earnings and to counter the negative consequences of world market instabilities.

Until recently, the analytical literature on price stabilization dealt primarily with commodity initiatives which fulfil their primary objectives perfectly. Analyses of price stabilization assumed that hypothetical buffer stock schemes stabilize prices successfully as intended by the decision-maker (Behrman, 1978; Nguyen, 1980). They concentrated on the consequential effects of successful price stabilization on other goals. On the basis of this approach, an argument for buffer stock schemes is that pure price stabilization can increase the aggregate welfare of all market participants. Moreover, price stabilization is expected to reduce earnings instability in many cases (Nguyen, 1980) and, hence, contribute to economic growth (Lim, 1976). In another branch of the literature, it is argued that compensatory financing systems are superior to buffer stocks and export quotas. Again, those results are based on the assumption of a perfectly stabilizing compensatory financing scheme (Newbery and Stiglitz, 1981, p. 299; Bird, 1987).

The purpose of this study is to discuss some of the factors that are crucial in the determination of long-run international petroleum prices, to explore what impact these factors might have, and to clarify whether the price formation process in petroleum differs significantly from that in other primary commodity markets. The time horizon of the study is the next 2–3 decades. Sections 2, 3, 4 and 5 explore, in turn, the roles of (a) exhaustibility; (b) monopolistic forces; (c) inter-fuel substitution; and (d) future technological shocks, represented here by the possibility of a moratorium on nuclear power, or of severe restrictions on carbon dioxide emissions. However, in order to provide a broader context, and a historical perspective on the subject, Section 1 reviews the prices of petroleum and other primary commodities since the beginning of the present century.

The historical evidence of prices: oil and other primary commodities

Figures 7.1–7.4 depict indices of real prices between 1900 and 1986 for oil, metals and minerals, non-food agricultural commodities and food commodities. An ocular inspection of the oil price developments reveals a gentle downward slope beginning after the first world war, and ending in the early 1970s, when, in the view of a majority of economists, the OPEC cartel drove up prices to historically unprecedented levels. The slow downward slope is interrupted by a much steeper temporary decline in the latter half of the 1940s, caused by the opening up of the rich Middle East oil resources on a large scale.

Introduction

Countries which produce primary commodities face rapidly fluctuating prices and revenues from their output and exports. Many of the producing countries, particularly those middle income or poor countries in the third world, depend on commodity exports as their primary source of foreign exchange, employment and development funds. Moreover they are typically dependent on the exports of just one or two commodities, and are therefore unable to diversify away from those price and earnings risks. For example, World Bank figures for 1986 show that out of 42 countries with GNP per head of less than $1,600, 20 were more than 70% dependent on just two commodities for export earnings. Revenue fluctuations can severely restrict growth by interrupting income levels and the flow of development funds in countries which are highly indebted and unable to borrow from the capital markets in order to smooth their revenues. Yet the indebted countries have increasingly relied on commodity exports for servicing their debts.

This paper examines the statistical distribution of earnings from commodity market sales, with arbitrary distributions for market clearing prices and quantities. We are concerned with limiting the impact of demand and supply disturbances by introducing certain rules governing stockholding, or by committing certain quantities to be bought/sold at a known price on the futures market, or by purchasing options to buy/sell at preannounced prices. The producers' decision is then to select an ‘intervention’ rule which improves the earnings distribution, typically either in terms of raising average earnings or of reducing their variance.

Introduction

Consultants and industrial economists think about the world in a quite different way from academics.

First, they have more information. Much of the data available to consultants comes from proprietary studies, which may not be available to academics. And if data is available, academics are often surprisingly unaware of it. For a good recent example of this, see Fama and French (1988).

Second, consultants are less sophisticated in matters of modelling technique. In part, this is because we are simple fellows. But in addition, consultants are severely constrained by the need to present their models and results to clients, who will often be non-economists. Clients distrust forecasts, and will be very quick to reject any findings which are based on a model which they do not understand. It is essential that models are simple and transparent, and that the clients can see the little wheels going round.

Thus consultants use models that have simple logic but which are knowledge-intensive. This paper describes a forecasting approach of this type, as applied to the aluminium industry; although the detailed specification of a model will not be pursued here. The paper naturally reflects the special features of the aluminium industry; but models using similar methods are in common use by consultants and industrial economists concerned with other metal commodities.

Such models generally work by asking in what way the world would be different if only expectations were formed in a rational and consistent manner, instead of the way that they are.

Introduction

Understanding agricultural prices is a complex and highly evolutionary process. In most countries, agricultural prices are more highly regulated than other prices. Thus, understanding the effects of domestic agricultural policy is crucial. This problem is complicated by frequent revisions not only in the levels of policy instruments but also by changes in the active set of instruments. Moreover, each major new agricultural price swing over the last several decades has drawn attention to an additional set of policy instruments that has important spillover effects on domestic agriculture. These sets of policy instruments include domestic macroeconomic policy, foreign agricultural policy, and foreign macroeconomic policy as well as domestic regulation of other sectors.

This paper discusses the role of alternative sets of policy instruments in determining agricultural commodity prices, evaluates alternative approaches for modelling agricultural commodity prices, and briefly discusses some empirical experience. Some of the points in the paper are substantiated by specific examples while others are offered simply as a summary of intuition and experience. The two main messages of the paper are (1) that increasing volatility in the agricultural economy calls for imposing more structure in estimation in order to capture the global properties of important relationships and (2) that increased volatility has revealed many important international, intersectoral and macroeconomic linkages that necessitate an ever broadening scope in agricultural modelling.

Domestic agricultural policy instability

The most important variable on the supply side that drives agricultural crop prices is acreage planted. Acreage of some of the most important crops depends heavily on government policy.

Introduction

The general question of price determination is complex since it involves issues concerning the structure of markets in which a commodity is traded and whether these conditions can lead to disequilibrium. For commodities like tea and cocoa, and perhaps also other tree crops, there is a widespread presumption that price determination conforms to the basic paradigm of competitive markets, at least as a good first-order approximation. If this position is accepted, ‘price equations’ should correspond to the standard ‘law of supply and demand’ in which the change of price is a function of the gap between market demand and supply.

The elementary static market model consisting of supply and demand equations plus the equilibrium condition, will determine three endogenous variables, viz., quantity demanded, quantity supplied and the price. If an inventory accumulation identity is added to the model, so that equality of production and demand period by period is no longer required, the basic picture does not change if a new equilibrium condition is added which stipulates that the prevailing price must be such that inventories are willingly held. When inventories are included a price equation is not necessarily required to close the model.

The original research objective which instigated the present paper was intended to develop a satisfactory representation of the price formation process in the context of global econometric models for perennial crops like cocoa and tea (Akiyama and Trivedi, 1987).

Primary commodities constitute over 40 per cent of world trade and World Bank figures for 1986 reveal that of the 42 countries with a per capita GNP of less than $1,600, 20 were more than 70 per cent dependent on just two commodities for export earnings. As a result fluctuations in primary export earnings can disrupt LDCs' economic stability and interrupt the flow of finance for development. Primary commodities also play an important role in the behaviour of inflation in the industrial economies. Commodity price volatility has increased almost ninefold during the 1980s. The relationship between commodity prices and world economic activity appears to have broken down – prices were much more depressed than past experience would have suggested – and the relationship with the dollar appears to have shifted as well.

The experience of the 1980s provides an excellent opportunity to test and refine hypotheses about commodity price behaviour and to develop their implications for economic policy and the management of commodity markets. This book reports the proceedings of a major international conference on primary commodity price determination which was held in London during March 1989. The essays and discussants' comments illustrate both the current state of the art in commodity economics and the rich array of alternative approaches which exists. This introduction sets the analytical and policy scenes, places the contributions in their contexts and gives brief non-technical summaries of the principal results.

Introduction

It is well known that primary commodity prices show large fluctuations. Their causes and effects have been a matter of interest for a long time. Agricultural economists have focused upon firm-level aspects (like producer's income, planting decisions and the like) whereas development economists directed their attention towards macroeconomic aspects (e.g., welfare measures, changing terms of trade). Crucial in most studies is the question how to lessen the volatility of prices. Futures markets for commodities are usually found to be a stabilizing factor in these markets. Unfortunately, the investigation of interactions between markets (influencing primary commodities trading) has been neglected. A general equilibrium framework where futures prices themselves become endogenous is difficult to deal with. Most analysis in this direction, like Kawai (1983), is too general to provide insight into the specific effects of volatility on output and trade.

Several empirical studies like Chu and Morrison (1984), suggest a large number of economic indicators showing substantial correlation with commodity price volatility. How can one incorporate all uncertainties influencing the primary commodity market into one model? We chose to link the commodity market with the currency market, following Kawai and Zilcha (1986), as these seem to be the two variables of foremost interest. Since most primary commodities are traded internationally and are usually denominated in a few currencies, the exchange rate is believed to be one of the main factors of importance in production decisions.

Introduction

The traditional textbook paradigm requires that primary commodity prices adjust to clear the market. Since production and consumption are generally inelastic over the short run (certainly periods of one year or less) this suggests that fluctuations in demands (boom or recession in the industrialized countries) or supply (good or bad harvests) may result in considerable price variability. Primary commodity prices are indeed highly volatile, and this has been taken by some as providing a prima facie case for price stabilizing intervention.

On the other hand, it is also widely recognized that storage is at least potentially important in these markets, and that the demand for storage might in principle be highly elastic. Private agents will typically buy low and sell high, and this will provide a degree of automatic stabilization via the market. If that is the case, it is not clear that one should expect the public sector to be any more efficient than the private sector in providing the socially optimal level of storage.

Extensive storage by the private sector also has other implications. Agents undertaking storage activities will be conscious of the opportunity cost of the funds that they use for this purpose, and this implies that storage, and therefore also commodity prices, will be sensitive to the rate of interest. Primary commodities will then behave to some extent like financial assets.

The aims of this paper are:

1. (1) to outline some methods of incorporating rational expectations into disequilibrium models;

2. (2) to discuss the applicability of these methods in the modelling of commodities characterized by market interventions like price supports and buffer stocks; and

3. (3) to review the usefulness of other sources of price expectations like survey data, futures prices and Bayesian vector autoregressions, if the methods suggested here cannot be implemented.

Sources of disequilibrium

The methodology used for incorporating rational expectations into disequilibrium models depends on the particular sources of disequilibrium. In commodity modelling the sources are the different methods used for price stabilization: McNicol (1978, p. 25), for instance, lists the following types of market intervention.

1. (1) Purchase alone, with the material held in storage indefinitely or disposed of in a way that does not affect market price.

2. (2) Purchases coupled with restrictions on supply.

3. (3) Purchases and sales without any restriction on supply.

4. (4) Restrictions on supply alone.

5. (5) Purchases and sales coupled with restrictions on supply.

The US agricultural programs provide examples of the first two. A support price program is one of purchase alone. Soil bank and acreage control come under category (2) since they limit the output.