A new vision for agriculture
momagri, movement for a world agricultural organization, is a think tank chaired by Christian Pèes.
It brings together, managers from the agricultural world and important people from external perspectives,
such as health, development, strategy and defense. Its objective is to promote regulation
of agricultural markets by creating new evaluation tools, such as economic models and indicators,
and by drawing up proposals for an agricultural and international food policy.
Focus on issues

Market Interdependence and Volatility Transmission among Major Crops



Cornelis Gardebroek, Manuel A. Hernandez, Miguel Robles,

IFPRI



Global food prices have experienced high volatility levels with two spikes––in 2007/2008 and in 2010/2011––that are highlighting the structural hyper-volatility of agricultural markets, whose consequences on food security and agriculture are factual. The liberalization process initiated in the 1980s has been a factor in the transmission of price volatility to global markets from local markets. The progressive opening of borders has gone hand in hand with the dismantling of agricultural policies, especially in Europe, resulting in an increased exposure to erratic prices for farmers.

We highly recommend this excerpt from a recent paper published by the International Food Policy Research Institute (IFPRI) that examines the interdependence of some commodities––especially agricultural commodities––and the volatility transmission between these commodities. As a result, the IFPRI’s researchers are addressing the concept of co-movement and are presenting extensive studies on the subject, studies conducted by those who feel the phenomenon is attributed to the financialization of agricultural markets and herd behaviors in commodity markets (endogenous price shocks), or conversely, those who feel it is due to exogenous shock (climate hazards and macroeconomic shocks).

Likewise, the interpretation of futures trends in agricultural markets is by no means meeting a consensus, and the issues of volatility transmission between agricultural sectors are also subject to debates due to their interdependence. Yet, we have no choice but note that food security and agriculture are now more than ever vulnerable, not only due to exogenous shocks, but also to exogenous shocks, which make up the specific nature of current agricultural markets.


momagri Editorial Board






In recent years agricultural commodity prices have taken a roller coaster ride. Three sharp price increases were observed in 2007–2008, 2010, and 2011, respectively, all of which caused major unrest on markets and in the media and have reinforced the attention of researchers and policymakers to further understand the behavior of commodity prices. One stylized fact of commodity prices that remains puzzling is their apparent high degree of co-movement. Although commodity price increases (and fluctuations) often have different causes, the prices of various agricultural commodities regularly move together (Gilbert 2010).

Co-movement of a wide range of commodity prices has been extensively studied in the literature. In their seminal paper, Pindyck and Rotemberg (1990) analyzed co-movement of seven raw commodity prices and found that after controlling for macroeconomic and market conditions, prices still moved together. Pindyck and Rotemberg named this phenomenon as excess co-movement in commodity prices and attributed it to herd behavior in financial markets.

This excess co-movement hypothesis, however, has been challenged by subsequent studies. Deb, Trivedi, and Varangis (1996) argue that most results by Pindyck and Rotemberg are due to misspecification because heteroskedasticity and structural breaks are neglected. To analyze herd behavior in commodity markets, Deb, Trivedi, and Varangis recommend further research using daily prices. Cashin, McDermott, and Scott (1999) used concordance analysis to examine commodity price cycles. They find strong evidence of co-movement within agricultural and metal commodities but not between them. Ai, Chatrath, and Song (2006), in turn, do not find evidence for excess co-movement when analyzing five major agricultural crops in the United States. They concluded that fundamental factors are more important than macroeconomic factors in explaining price co-movement.

Saadi (2010) provides a review of commodity price co-movement in international markets. He discusses several explanations for price co-movements, for example, not only macroeconomic factors such as exchange and interest rates but also common supply and demand factors affecting prices of agricultural commodities. The latter include co-varying harvest levels (for example, drought hitting corn, soybean, and wheat harvests in the United States), joint low stocks, and substitution in supply and demand (for example, wheat replacing corn in animal fodder). According to Gilbert (2010), price shocks for individual commodities are often supply related, whereas joint price movement can be explained from macroeconomic and monetary conditions. Natanelov et al. (2011) use cointegration and Granger causality tests to analyze co-movement of crude oil and agricultural commodity futures prices. Although they present evidence of agricultural futures prices to co-move with crude oil prices, agricultural and biofuel policies also seem to distort these relationships.

Excess co-movement in commodity prices may be problematic for several reasons. First, it casts doubt on the efficiency of commodity markets as long the source of co-movement is beyond fundamental factors. Second, it creates difficulties to balance the portfolios of exporting countries and commodity traders. In agricultural markets, farmers who grow multiple crops may also be subject to strong income fluctuations due to synchronized ups and downs in prices, with important implications for food security. In addition, a synchronized increase in several commodity prices may generate inflation pressures in the short term on highly dependent commodity-import countries.

Most of the literature on price co-movement cited above, however, focuses on price levels or conditional mean prices. Less attention is given to interrelations in conditional volatility. Examining market interactions in terms of the conditional second moment can provide better insight into the dynamic price relationships of markets (Gallagher and Twomey 1998). A period of increased volatility in, for example, wheat prices could also lead to more volatility in corn or soybean prices due to substitution in, for example, wheat prices could also lead to more volatility in corn or soybean prices due to substitution in demand or joint underlying causes of volatility. Moreover, the excess co-movement hypothesis is often motivated by phenomena on financial markets, such as herding and speculation activity, which may also lead to increased volatility interactions between commodities. An ongoing debate in agricultural commodities is whether the apparent higher market integration of agricultural financial markets and the increasing role played by noncommercial actors or index traders in past years may have led to higher cross-market interactions between crops (Irwin, Sanders, and Merrin 2009).

Another important issue that is often neglected is that different data frequencies may lead to different conclusions on the existence of co-movement in price levels and volatility. For example, short-horizon interactions between markets are more likely affected by daily trading in financial (futures) markets as compared with long-horizon interactions, which are more likely driven by structural changes in markets. Hence, the use of different data frequencies can provide a richer picture of the potential underlying factors driving market interdependencies across commodities. Similarly, longer-horizon returns can obscure temporary responses to innovations, which may last for only a few days or weeks (Elyasiani, Perera, and Puri 1998). In contrast, working with different data frequencies might help us overcome the minimum detectable effect problem. If we think of innovations (shocks) as arising on a continuous fashion, then over a month-long period the cumulated innovations result in a higher variance than over a day-long period. Daily cumulated innovations might be too small to make a statistically detectable impact, but on a monthly basis this might be less of a problem. Most of the aforementioned studies use particular data frequencies, for example, quarterly, monthly, or weekly. (…)

This paper intends to contribute to the literature on volatility spillovers between agricultural commodities, which is still limited.


1 http://www.ifpri.org/publication/market-interdependence-and-volatility-transmission-among-major-crops?utm_source=feedburner&utm_medium=email&utm_campaign=Feed%3A+Ifpriupdate+%28IFPRI+Website+Update%29
Page Header
Paris, 18 December 2018