Rice price formation in the short and long run: the role of market structure in volatility
Nearly half the world's population consumes rice as a staple food, 90% of which is produced in Asia. Of the 440 million metric tons (mmt) of milled rice equivalent produced globally, most is consumed within a short distance of where it was produced. Only 7-8% crosses an international border. Nevertheless, the world rice market is important - providing essential supplies and sending out global price signals. It is therefore a major concern when such market signals are highly volatile. This paper looks at the unique causes of rice price volatility, addressing questions on price formation trends and the influence of speculation on the rice price spikes.
Causes of rice price volatility
The variation coefficient of world rice prices has often been double that of wheat or maize. Understanding this volatility is difficult due to the nature of the market, where countries stabilize internal rice prices by the disposal of surpluses onto the market or by importing more to meet deficits. Thus, the political decisions of key Asian countries directly influence worldwide supply and demand (Timmer and Falcon, 1975).
Rice price volatility is also driven by rice production structure and by marketing and consumption trends in Asia. Millions of farmers, traders, processors and retailers, along with billions of consumers, all handle a commodity that can be stored for over a year in a consumable form
. The price expectations of these market participants are critical to their decisions on how much to grow, sell, store and consume.
Analyzing food price formation
Understanding causation in food price formation implies an empirically refutable analytical model based on supply and demand mechanisms, with equilibrium prices derived from basic competitive forces
. Gradual changes in supply and demand can cause gradual increases in prices but the food price explosion of 2007 and 2008 clearly requires additional explanation beyond a simple supply and demand model.
The "supply of storage" model provides the basis for interpreting short-run price behaviour for storable commodities (Houthakker, 1987). It stresses the inter-related behaviour of speculators and hedgers as they judge inventory levels in relation to use. To make this model operational in the short run, it is used to gain insight on the formation of price expectations. These expectations seem to be driven by a combination of commodity price behaviours and the specifics of individual commodities. Broad commodity price trends are captured by various indices, such as the IMF commodity price index. Thus, traders operating in any one commodity market will track the broader price movements for all commodities (Sanders and Irwin, 2008). These broad price movements seem to be driven by basic macroeconomic forces such as economic growth rates, the value of international currencies and relative inflation rates (Timmer, 2008a).
Traders also follow their specific commodity closely. Here inventories determine price formation, once a crop's harvest/supply situation has been established. Typically, commodities with reasonably reliable inventory data tend to have short-run prices driven by unexpected supply behaviour, whereas commodities with poor inventory data, especially where significant inventories can be in the hands of millions of small agents (farmers, traders, consumers, etc.), tend to have extremes in price behaviour generated by rapidly changing price expectations, and consequent hoarding or dis-hoarding of the commodity. The short-run price dynamics for rice thus look significantly different to wheat or maize.
The role of market structure in rice prices
The actual production/consumption balance for rice has been relatively favourable since 2005. Short-run substitutions in both production and consumption between rice and other food commodities are limited, and until late 2007 it seemed that the rice market might avoid the price spikes seen in wheat, maize and vegetable oil. The limited futures market for rice also made financial speculation less attractive.
However, with only 7 to 8% of global production traded internationally, the world rice market is very thin, leaving it vulnerable to large price moves. The market is also relatively concentrated, with Thailand, Vietnam, India, the US and Pakistan routinely providing nearly 80% of available supplies.
Does this difference in market structure also account for the difference in long-run price trends between rice and the other two staple food grains? Only to a limited extent. The political economy of high variance in world rice prices leads countries to retreat into autarky, dumping their own instability into a smaller world rice market. One consequence of this self-sufficiency drive among rice importers is larger overall production than expected, which contributes to a long-run decline in world prices. But other factors (faster technological change, slower population growth in rice-consuming countries, changing consumer tastes...) could account for less tension between supply and demand for rice, than for wheat and maize.
The 2007 - 2008 commodity prices spike
The scope for extreme volatility is clear in light of the political and market structure dimensions for rice. Understanding how these factors contribute to the formation of price expectations on the part of market participants is essential to grasp the proximate causes of unstable rice prices. These expectations can drive "destabilizing speculative behaviour" among millions or billions of market participants, such that price formation seems to have a large, destabilizing, speculative component.
In 2007, concerns grew that world food supplies were limited and prices for wheat, maize and vegetable oils were rising. Several Asian countries increased domestic rice stocks to protect against future shortages
. The fear of shortage spread and a cumulative price spiral began.
In India, the 2007 wheat harvest was damaged by drought and disease, so the food authority announced that it would retain more domestically-produced rice rather than import more wheat. An outright ban on non-Basmati exports followed in April 2008 (India is usually the world's second largest rice exporter).
Other rice exporting countries followed suit as rice prices spiked. The commerce minister of Thailand's new government openly discussed export restrictions from Thailand - the world's largest rice exporter (9.5 mmt in 2007). By March 2008, rice prices in Thailand jumped $75 per mt, continuing to skyrocket until April, reaching $1100 per mt.
Poor consumers in many rice importing countries, including Vietnam and the Philippines, were seriously affected by this price volatility; while some countries, including India, China and Indonesia, were able to protect domestic populations from the rice shock through export restrictions. Leaders in these countries argued that domestic stability was more important than international stability
, reflecting Asia's longstanding view that rice price stability is essential for political stability and economic growth (Timmer and Dawe, 2007).
This is a vicious circle: world market instability causes countries to withdraw from importing and exporting rice, thus thinning the market, which increases instability and causes ad hoc policy reactions. Breaking this cycle requires long term binding agreements between rice importers and exporters to rebuild market confidence. National or international pressure for such agreements is, however, virtually non-existent
Hoarding behaviour, a "rational" response to panicky governments, was a critical component of the 2007 rice price rises. Financial speculation, however, seems to have played only a small role (partly because futures markets for rice are limited). Instead, decisions by millions of households, farmers, traders and some governments sparked a sudden surge in rice demand and changed a gradual price increase into an explosion.
Market structure plays an important role in short-run rice price formation, stemming from the highly unusual organization of the world's rice economy, with many small producers, traders, retailers and consumers handling a product that is storable at each stage.
The supply of storage model shows that this highly decentralized storage capacity is subject to changes in price expectations throughout the supply chain. As data on the rice stocks held by market participants is non-existent, their impact on rice price formation is unpredictable.
Most analysts think that long-run linkages exist among the staple grains, based on commodity substitutions in production and consumption (Timmer, Falcon and Pearson, 1983). Commodity markets for rice, wheat and maize are similar in that they are influenced by basic supply and demand factors, although the long-run fall in rice prices has been faster than for wheat and maize. Wheat and maize are more closely linked to each other than to rice markets, while any price impact caused by bio-fuels would have been mediated through the maize market and its financial linkages to other commodity markets.
Financial speculators were almost certainly responsible for some of the increase in wheat and maize prices. Conversely, speculation did not cause the rice price rise, due to the different industrial organization of domestic and international rice markets, along with government intervention.
The unwinding of speculative positions leads to rapid price falls, until perceived surplus stocks are consumed. With the exception of India, this process of short-run de-stocking seemed to end in late 2008, well before world prices began to rise again. The world financial crisis will probably prevent more rapid rises in rice prices, but the supply/demand balance for the next few years suggests export prices will be closer to $500 per ton rather than $300 per ton in the pre-crisis period.
These conclusions are reached mostly by eliminating the other explanations and by logical reasoning. There is no specific quantitative test of the hypothesis on offer, and indeed, one may not be possible.
Rice, a market that is not predominantly global