Themes > Features

18.04.2000

Global Commodity Prices : High Volatility , Low Income

Primary commodity prices are well known to be highly volatile. Indeed, they are not just subject to short boom and bust cycles driven by demand and supply shocks, but also longer term cycles and secular trends driven by more structural and evolutionary forces.
 
There are many reasons for such volatility. First there is the strong relationship between climatic change and weather conditions. What is often less noted is the fact that even demand conditions are affected by weather, such as the demand for oil. Since the demand for most primary commodities is anyway less price-elastic than for manufactures, this means that any variability in supply gets associated with far greater price variability than in the case of manufactured goods.
 
In the case of perishable commodities, supply is inevitably very inelastic in the short term. For certain other primary commodities such as tree crops, the complete production process takes considerable time so that shifts in demand have a large impact on price in the intervening period.
 
For perishable primary commodities in particular, the costs of holding inventories tend to be high, which further affects price response. Since holding inventories of virtually all primary commodities, even the non-perishable ones, tends to involve significant costs, they are in turn significantly affected by the prevailing rate of interest. Thus movements in the rate of interest, which change the costs of holding these inventories, can be associated with changes in product price as such stocks are released to or withdrawn from the market by private traders. This tendency, which was noticed a century ago, is still an important element in the various forces making for commodity price volatility.
 
Indeed, even producers' organisations which are often intended to stabilise world market prices in particular commodities can have the opposite effect if they contribute to volatility through large shifts in supply. One example of this is OPEC, which has certainly affected international oil prices through its changing decisions on production and supply.
 
However, even by the standards of the past, the volatility of primary commodity prices appears to have increased quite dramatically in the past two decades. As Chart 1 indicates, primary commodity prices have been more volatile than the prices of manufactures in the last two decades, and both oil prices and non-oil commodity prices have fallen relative to the prices of manufactures. The chart provides data from the World Bank for the period from 1984 to late 1999. The unit value index of manufactures in this chart relates to the manufactures exported from the G–5 countries (France, Germany, Japan, United Kingdom, and United States) weighted by the country's exports to developing countries.

This evidence of higher volatility is confirmed by other studies, all of which suggest that the volatility of primary commodity prices increased sharply following the collapse of the Bretton Woods system in the early 1970s and has remained high thereafter, going up further in the 1990s. According to the World Bank publication Global Economic Prospects for Developing Countries 2000, the standard deviation of the absolute value of the year-on-year changes in the non-energy commodity price index during 1970–98 was 11.5, compared with 5.7 for the manufactures unit value index during the same period.
 
Within this general category, it is clear that energy prices have been particularly volatile. According to the same source, measured in US dollars, the coefficient of variation of energy prices was at least twice that of manufactures in each decade. The world oil market has been especially unstable over the 1990s. Peak prices during the Gulf War, price declines related to restructuring and non-OPEC supply subsequently, revival in the mid-1990s, falling prices and stock build-up combining with relatively warmer weather to produce a complete collapse. in 1998. By late 1998, many observers were predicting that energy prices would stay low for the coming decade, and within months this prediction was confounded by the Pheonix-like emergence of oil prices in 1999 to almost regain the level of the pre-collapse period.
 
Of course, oil prices are a special case because of the very different circumstances of global trade in energy, which have once again underlined the crucial role played by OPEC decisions in setting international prices in this sector. But it is worth noting that in general non-oil prices have certainly followed similar long-terms trends, and are now displaying similar (if less marked) volatility.
 
Not only has price volatility been a problem, but commodity prices have also declined relative to manufactures' prices during the past two decades. According to the World Bank, the real price indexes (nominal price indexes divided by the manufactures unit value index) of both agriculture and metals and minerals fell by about 45 percent during 1980- 98. It is now widely accepted that the decline in the ratio of the price of primary products to manufactures is statistically significant.
 
Energy prices have fallen by 76 percent in real terms since 1980 and in 1998 were less than half the level reached after the first oil price rise in 1973–74. Even with the recovery in 1999, the real energy price index likely will remain at only 55 percent of the 1974 level.
 
As is evident from Chart 2 and 3, the past two years have been especially bad for absolute levels of prices as well. After the mini-boom of 1994-96, there has been a fall in all the major commodity groups, which has been substantially related to both supply behaviour and demand changes, led for example by the east Asian crisis from mid-1997.

It is clear that supply conditions have played a role in depressing commodity prices, especially for agricultural and mineral goods. Thus, total world output of agricultural goods increased by 1 per cent per annum in the period 1990-94. This increased to 2.6 per annum over 1994-98. However, world consumption of agricultural commodities is estimated to have grown at a much slower rate.
 
The decline in primary commodity prices since 1997 was in part a response to an unusually large increase in supply. The rate of growth of world production of agricultural commodities rose from 1 percent per year in 1990–94 to 2.6 percent in 1994–98, whereas world production of metals and minerals was flat in 1990–94, but increased by 3.5 percent per year in 1994–98. 6 The acceleration of consumption of primary commodities between the two periods was much less.
 
It could has been argued that increased production was in part a response to the high prices that prevailed during the brief boom of 1993–95. If that were the case, then the expectations of producers must have been sorely disappointed, for agricultural commodities saw prices fall by 4.5 per cent in 1996 then recover slightly in 1997 before dropping by more than 16 per cent in 1998. The first ten months of 1999 saw such prices continue to fall, by an average rate of 14 per cent. Declines have been particularly pronounced for crops such as wheat, maize, coffee, rubber and sugar, and in 1999 the depressive tendency also extended to cocoa prices. However, as Chart 3 shows, the downturn has been fairly general across broad categories of food, beverages and raw materials.

There is a parenthetical point that deserves to be made here. These price movements have been so sharp that they have in turn affected the domestic prices of such commodities in countries where external trade has been increasingly liberalised. This is a point often missed in India for example, where discussion of the effects of trade liberalisation on Indian agriculture are often still based on world prices that prevailed in 1996. Unlike in that period, the current level of international prices of most agricultural goods is such that they are well below Indian prices, and so expectations of big increases in agricultural exports or resistance to import penetration which remain based on the earlier prices are likely to be misplaced.
 
Meanwhile, minerals and metals products prices have been declining quite sharply from 1995 onwards. As is clear from Chart 2, the cumulative decline from then to 1999 amounts to nearly 25 per cent. Once again, supply behaviour has played an important role : world production of metals and minerals did not increase at all over 1990–94, but subsequently increased by 3.5 percent per year in the period 1994–98.
 
The big falls in 1998 were of commodities that had already seen their prices decline substantially in the previous years, such as aluminium, copper, lead and zinc. The supply-demand imbalance is especially evident in the case of these commodities. Aluminium production increased by 5 percent per year during 1995–97, while consumption grew by only 3 percent per year. Copper production grew by 6 percent from 1995 to 1997, while consumption grew by 4 percent. Nickel production increased by more than 5 percent from 1995 to 1997, while consumption fell slightly.
 
The question, of course, is why there was such an increase in production. The standard explanation has been in terms of lagged responses to the relatively high prices of the mid-1990s. Another explanation has been in terms of improved technology, in terms of innovations ranging from improved seeds in agriculture to new solvent extraction techniques in mining, and improvements in production processes like horizontal drilling and increased computerisation.
 
These would certainly have played some role, but they are not likely to provide a complete explanation. For one thing, the increase in production of both minerals and metals as well as agricultural commodities in the later period was much sharper for developing countries. The Food and Agriculture Organisation's index of the volume of agricultural production rose by 3.8 percent per year for all developing countries from 1990 to 1997. The global supplies of metals and minerals also increased rapidly, more so for developing countries.
 
This difference is more likely to be related to other economic forces than technology or a generalised prices response. In fact, many of the largest increases in production came from countries whose economies desperately had to generate foreign exchange to service external debt and meet the bill for much-needed imports.
 
In other words, severe balance of payments pressures and structural adjustment cum liberalisation strategies based on increased openness had created a situation where many developing countries were increasingly forced to rely on increasing volumes of traditional primary commodity exports to generate foreign exchange for financial obligations as well as to meet basic consumption, much of which had earlier been met by domestic consumption. This pressure to generate higher export volumes on the part of most of the developing country producers in turn had a depressing effect on world prices.
 
Along with supply increases, demand conditions have also played a role in the most recent cyclical decline of commodity prices. The proximate cause that is always mentioned in this connection is the East Asian crisis, which dramatically reduced import demand in the crisis countries as well as created recessionary ripples which lowered consumption of primary commodities in other countries as well. The is because the East and Southeast Asian countries together with Japan account for very substantial shares - between one-fourth and one-third - of world consumption of most primary commodities.
 
This augurs ill for development prospects in these countries precisely because of the two processes that have been highlighted so far in terms of primary commodity prices : their volatility and their secular decline. Given the high - and indeed, growing - degree of dependence of many developing countries on primary commodity exports, the functioning of these international markets has not been such as to give either a reasonable possibility of sustained growth or freedom from the continuous need to stabilise economies and smooth the savings and consumption patterns over the relatively short and sharp price cycles.
 
These boom-and-bust cycles naturally impact on developing countries as a group far more adversely than on industrial countries. After all, primary commodities accounted for as much 42 per cent of all developing countries' total merchandise exports in 1997, compared with only 19 per cent for the rich countries.
 
But still, even within developing countries, there are two groups that have been the most affected by the most recent boom-bust cycle. The first group consists of the major oil exporting countries, that is, countries where oil accounts for more than 50 percent of merchandise exports. The other group is that of the non-oil exporting countries of Sub-Saharan Africa, where non-oil primary commodities, on average, make up 80 percent of exports.
 
For some countries, especially the least developed countries of Sub-Saharan Africa, primary commodities account for 90 per cent of exports, and very often more than half to two-thirds of export value is accounted for by one single commodity. This naturally makes both the balance of payments and the domestic real economy extremely vulnerable to fluctuations in output and price of this one good or group of goods.
 
Despite this dismal performance of primary commodities, the overall terms of trade of all developing countries does not show such a negative tendency, as will be clear from Charts 4 and 5. This is actually because of the degree of diversification out of primary into manufactured exports that was achieved by certain developing countries, such as the previously named "dragons" of East and Southeast Asia as well as India and some Latin American industrialisers. However, such trade and production diversification is precisely what is now under threat in some countries where it had proceeded quite substantially, and is further constrained in less developed countries where it had not gone very far.

In regions where no such process of export diversification was evident, such as Sub-Saharan Africa, not only did terms of trade deteriorate very substantially, but this also had very large negative effects on real national income. Chart 8 shows this, in terms of more than 10 per cent decline in terms of trade over just two years 1997 and 1998, and an effect of this alone on income amounting to 2.7 per cent of GDP.

The reason all this is particularly worrying is that primary commodity prices are not expected to recover very much in the medium term. Charts 6 and 7 show one set of projections by the World Bank for oil and non-oil prices until 2001. What is interesting is that despite the huge falls of the previous years, prices will not increase much and are not even likely to regain their levels of the mid-1990s. And this is notwithstanding the much-vaunted "recovery" in East Asia which is supposed to increase world demand for primary commodities once again.

One reason for this is that in the current slump, except in the case of oil, producers/traders of primary commodities have been reluctant to reduce output to limit or reverse the fall in price. rather they have chosen the option of increasing the level of stocks held (aided by relatively low interest rates) and this has meant a fairly large overhang of inventories. Stocks of metals such as copper and aluminium are at record levels; grain stockpiles held by the major exporters have more than doubled over just three years; sugar stocks are also reaching unprecedented levels.
 
Thus low prices now reflect more than the usual cyclical features. They are now characterised by severe overcapacity in a situation of generally sluggish world demand for such goods. It is clear that they are unlikely to bring much joy to exporters in the medium term.
 
This is important, because this means the eradication of the hopes which characterised the Uruguay Round of trade talks, that more liberal trade would lead to higher prices for developing country exporters of primary commodities. Instead, what we witness is low prices and overproduction, in a situation where competitive pressure and the need to push out more exports have made these commodities even greater sources of volatility and vulnerability in domestic economies.