Analyzing 50 futures markets over a 20-year period, the author explores what appears to be an inverse relationship between the empirically based Samuelson hypothesis and the theoretical construct known as “carry arbitrage.”
The author investigates what appears to be an inverse relationship between the Samuelson hypothesis and the strategy of carry arbitrage. The degree to which carry arbitrage is validated through observation may be linked to the extent to which the Samuelson hypothesis is not. A better understanding of futures price volatility can have a beneficial effect on investors’ use of proper hedging and on margin requirements, thereby improving risk management in these markets.
How Is This Research Useful to Practitioners?
Looking at 50 diverse futures markets from 1993 to 2012, the author explores the link between the Samuelson hypothesis and the carry arbitrage strategy. The degree to which the inverse relationship between these two conditions may hold is a function of the futures market in question.
The Samuelson hypothesis states that futures market price volatility increases as the contract expiration approaches. The carry arbitrage strategy entails purchasing the underlying instrument or commodity with borrowed funds and selling short a forward or futures market position to hedge price risk. The related reverse carry arbitrage strategy entails selling short the underlying instrument, investing the proceeds of the sale, and hedging price risk with a long forward or futures position. It is important to note that carry arbitrage is a theoretical model that empirical (observable) data may or may not support.
Carry arbitrage is more easily accomplished in some markets than in others, with many markets falling into a “semi-arbitragable” category. The presence of market participants willing to arbitrage, carrying costs, the ability to sell short, the ability to lend and borrow at the risk-free rate, and minimal margin requirements are important factors in how readily this strategy may be effectively executed. Markets bereft of some or all of these attributes may be subject to greater volatility as the contract expiration approaches because no release mechanism exists to control or moderate volatility.
Traders, futures market economists, regulators, and policymakers would benefit from this research because it would inform them of market and structural forces that drive commodity and financial markets. In addition, it carries important implications for the practice of risk management.
How Did the Author Conduct This Research?
The author provides a robust review of the relevant literature on both the Samuelson hypothesis and the carry arbitrage strategy. He then presents alternative representations of futures markets to establish to what degree percentage price changes in a long-dated futures contract derive from changes in the nearby contract. Futures markets are some combination of markets where carry arbitrage may be conducted (e.g., minimal market frictions) and markets where arbitrage cannot be carried out because of the presence of such market frictions as an inability to sell short. The author terms this condition “quasi-arbitraged.”
He then develops four hypotheses that he tests on a variety of futures markets. These hypotheses explore futures variance by decomposing the futures contract into its long- and short-dated components to determine degrees of correlation. The degree to which the inverse correlation between the Samuelson hypothesis and the carry arbitrage strategy is validated is a function of the markets observed because of the differing structures of futures markets for specific products. The extent to which the Samuelson hypothesis is validated is a function of the degree of covariance (what the author calls “beta”) between the longer- and shorter-dated futures contracts. For example, grains and oilseeds, livestock and meats, energy, and food and fiber exhibit marked differences in variance between short- and long-dated futures contracts, whereas interest rate markets display traits consistent with neither the Samuelson hypothesis nor a flat volatility structure.
The author concludes that futures markets are unique, and the observation that the Samuelson hypothesis is supported when carry arbitrage is not depends on the markets under observation, where volatility patterns vary as a function of supply and demand and of market structure.