ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

More on Futures Trading and Commodity Prices

This article critiques 'Impact of Futures Trading on Commodity Prices' by Golaka C Nath and Tulsi Lingareddy, pointing out theoretical and econometric flaws in addition to fallacies about volatility.

DISCUSSION

More on Futures Trading and Commodity Prices

Madhoo Pavaskar, Nilanjan Ghosh

on spot price behaviour. The origin of such a surreal surmise may be traced to the stock market system, about which most researchers, with their academic background in finance rather than in economics, are quite familiar. The futures and op-

This article critiques ‘Impact of Futures Trading on Commodity Prices’ by Golaka C Nath and Tulsi Lingareddy, pointing out theoretical and econometric flaws in addition to fallacies about volatility.

What is expressed here are the personal views of the authors.

Madhoo Pavaskar (madhoo.pavaskar@ maerindia.com) and Nilanjan Ghosh (nilanjan. ghosh@maerindia.com) are with MCX Academia of Economic Research.

I
n their article, ‘Impact of Futures Trading on Commodity Prices’ published in the January 19, 2008 issue of Economic & Political Weekly, Golaka C Nath and Tulsi Lingareddy have sought to assess the impact of futures trading in gram and ‘urad’ on their physical market prices and price volatility. The authors have, however, applied econometric techniques with little regard for economic logic and analytical reasoning. They conclude that volatility in urad and pulse prices was higher during the period of futures trading than prior to its introduction and after the ban on urad futures.

The article actually falls in the category that is dominating recent research in commodity derivative markets – extensive reliance on econometric tools, disregarding not only basic economic theory of futures trading but also the ground realities of such trading in terms of changing supply and demand conditions of the underlying commodities, and the pattern and magnitude of actual trading in those markets. Consequently, the econometric analysis is not infrequently based on prices solely, as if prices, especially of commodity derivative contracts, are determined independently. Not surprisingly, the results are patently absurd, and their policy implications not just misleading but worse still, even mischievous. To put it bluntly, most such articles rely on theoretical myths relating to commodity futures and abound in econometric flaws and fallacies.

Theoretical Myths

The problem lies in the tacit belief amongst most researchers that futures prices in commodities are arrived at by intense speculative activity on both the long and short sides of the market. Based on this myth, they proceed to statistically or econometrically ascertain, depending on their fondness for the one or the other methodology, the impact of futures prices tions segment of the stock market is for all intents and purposes speculative in nature, dominated by speculators of all types, day traders, scalpers, arbitrageurs, etc, while its cash segment comprises mostly investors of all hues.

But what is perhaps lost sight of by these researchers is

tt
hh
at cat c
oo
mm
mm
oo
dd
ii
tt
y ey e
xx
--
changes are not stock exchanges [Pavaskar 2004]. Commodity exchanges are basically institutions that are adjunct to physical markets and provide auxiliary services to assist physical market functionaries in various nodes of the commodity value chain in their diverse marketing and processing activities. Unlike cash prices in stock markets, which are determined at centralised locations – such as the headquarters of stock exchanges – commodity prices vary across locations, by quality characteristics and end-usage patterns, and over seasons (ibid). Moreover, stock market players are primarily those who earn regular income by way of dividend and/or interest or through profit from speculation. They have little interest in price risk management and are not hedgers in the sense commodity players are. Hence, the hypotheses that may be valid for research in stock exchanges are unlikely to be applicable for studies on commodity markets.

In fact, the same underlying forces of supply and demand determine the physical and futures market prices in commodities. It is erroneous to believe, as many often do, that physical market traders are mainly concerned with current supply and demand, while futures market players look largely at the prospective supply and demand. To be sure, physical market functionaries are as anxious about impending supply and demand as those operating in futures markets. Moreover, quite a few of those trading in commodity futures are in no way different from those trading in physical markets. True, commodity futures markets discover prices but these

march 8, 2008

EPW
Economic & Political Weekly

DISCUSSION

are discovered in the sense that they are drawn by mutual consensus from a bewildering number of heterogeneous prices of different varieties of the same commodity for ready and forward deliveries in different periods and locations. Hence, such prices serve as effective benchmark prices for all trades in physical markets for either domestic purchase and sale or import and export. But alas, Nath and Lingareddy’s article betrays a lack of this basic theoretical understanding of the futures price mechanism. Not surprisingly, this lacuna has crept into their empirical econometrical analysis too, which is replete with even more grievous flaws.

Econometric Flaws

No doubt, the authors have used the weekly wholesale price index (WPI) series as compiled and published by the Central Statistical Organisation for their empirical analysis. They prefer the WPI since reliable data on spot prices were not available readily. The first differentials of the price indices have been utilised for the study.

The problems, however, start with the regression framework. The authors have considered a regression equation where prices of gram, pulses, and foodgrains, as also those of urad with a one week lag, besides prices of all commodities (along with a time dummy to capture periods before and after futures trading) have all been taken as independent variables to explain the price of urad. Similarly, in order to analyse gram prices, they have considered as explanatory variables gram prices with a one period lag, prices of urad, pulses, and foodgrains, as also those of all commodities, as well as a dummy to represent the futures and non-futures trading periods. It requires no sixth sense to establish that the regression equations with such mutually dependent explanatory variables suffer from the Achilles heel of multicollinearity, as both gram and urad are pulses, and the WPI of pulses falls in the same league too. Even a student of elementary econometrics knows that a model collapses in the presence of multicollinearity and emits wrong results. The situation is not different here.

There are at least two pieces of evidence in the regression equations presented by

Economic & Political Weekly

EPW
march 8, 2008

the authors that reveal the anomalous economic behaviour of variables, which is difficult to explain. In the equation explaining urad prices, the regression coefficient of overall pulses prices is positive and significant but that of gram prices is negative and significant. If we are to take this relationship seriously, we need to believe that gram is a substitute for urad, and that overall pulses serve as supplementary foods in regular meals. That goes beyond logic and rationality. In the regression equation drawn for gram the same absurd phenomenon is noticed. The explanatory coefficient for pulses is positive and significant, while that of urad is negative and significant as well. Common sense repudiates such a regression equation. Is it then really possible trust such spurious regression results, which have emerged from faulty methodology and flawed economic logic? Hence, will it be too much to assert that the dummy variable that has turned out to be somewhat significant, suggesting that futures prices in urad (though not likewise in gram) may have aggravated the spot price volatility, is more impish than evocative?

More shockingly, the authors have even failed to consider the demand and supply conditions in gram, urad, and pulses to explain price movements in them, prior and subsequent to futures trading. Even the elementary economic theory taught at the undergraduate level relies on supply and demand to explain price behaviour in any commodity. The authors’ absolute neglect of these primary price determinants vividly lays threadbare their faulty econometric analysis, which is lacking in simple economic logic.

Volatility Fallacies

Even more absurdly authors have used standard deviation as a measure of price volatility, not realising perhaps that standard deviation is a scale-biased measure, unsuitable for gauging price volatility over long time periods with varying means at different times. A more accurate way of determining volatility in such cases would have been the coefficient of variation. The authors have failed to reckon this reality.

Volatility is the raison d’etre for futures trading in commodities. Price volatility is not so much a matter of concern for those who hedge in futures to avert risks arising from it as for those who fail to do so. The fuss about price volatility in the presence of futures trading makes little sense in any price research on commodity markets.

As it is, it is wrong to attribute price volatility to speculation [Kabra 2007: 1165]. Such a position has been taken as a premise in many research papers like the one by Nath and Lingareddy, and the commodity futures market is then needlessly vilified and treated as a “whipping boy”. Such anti-market statements have traditionally been based not so much on concrete empiricism underlying long-run rationality as on irrational emotional sentiments based on myopic observations.

The role of speculators in the futures market can never be overemphasised. They take up hedgers’ risks and provide liquidity to the market. Without liquidity, hedgers will desert the futures market. Low liquidity unreasonably increases the entry and exit costs of hedging in a futures market, thereby creating entry and exit barriers. Liquidity, on the other hand, reduces the cost of hedging by minimising the difference between the bid and ask prices [Pavaskar et al 2008].

It is not so much an excess of speculation as the lack of it that has been increasing price volatility in recent years in most agricultural commodities in short supply, following the demand pressure on them, thanks to galloping GDP growth, both at the aggregate and per capita levels. In point of fact, it is rather difficult to differentiate between the “hedger” and “speculator”, and the dichotomy between the two is more academic than pragmatic (ibid). In the course of trading, a hedger can become a speculator, for, depending on market perceptions and expectations only, physical market players decide on whether to hedge or speculate.

References

Kabra, Kamal Nayan (2007): ‘Commodity Futures in India’, Economic & Political Weekly, Vol 42, No 13, pp 1163-70.

Pavaskar, Madhoo (2004): ‘Commodity Exchanges Are Not Stock Exchanges’, Economic & Political Weekly, Vol 39, No 48, pp 5082-85.

Pavaskar, Madhoo, Nilanjan Ghosh and Danish Hashim (2008): ‘RBI Report on Currency Futures: Good, But Not Good Enough’, Commodity Vision, Vol 1, No 3, pp 25-28.

Dear reader,

To continue reading, become a subscriber.

Explore our attractive subscription offers.

Click here

Comments

(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top