On the unconditional and conditional cross section of expected futures returns
While most of the literature on asset pricing examines the cross section of stock and bond returns, little attention has been devoted to the analyse of the trade-off between risk and return in futures markets. Since, alike stocks and bonds, futures contracts qualify as investments on their own, the purpose of this thesis is to remedy this problem by addressing three issues related to the unconditional and conditional cross section of expected futures returns. Chapter II investigates the presence of a futures risk premium and, ultimately, the validity of the normal backwardation theory in the context of constant expected return asset pricing models and argues that the inconsistency in the literature stems from methodology problems that might result in incorrect inferences regarding the applicability of the normal backwardation theory. With methodologies free from these problems, we show that, while producers and processors of agricultural commodities transfer their risk to one another at no cost, hedgers are willing to pay a premium to induce speculators to enter financial and metal futures markets. Chapter III looks at the integration between the futures and underlying asset markets. While we fail to reject the hypothesis that the prices of systematic risk in futures markets are equal to those in the underlying currency and equity markets, we present new results that the futures and commodity spot markets are segmented. Such results are of primary importance to investors who use constant expected return asset pricing models to adjust the risk-return trade-off of their portfolio and evaluate portfolio performance. The remainder of the thesis investigates the degree of efficiency with which futures contracts are priced. Since futures returns are predictable using information available at time t-1, the purpose of chapters IV, V. and VI is to analyse whether the variation in expected futures returns reflects rational pricing in an efficient market or is the result of weak-form market inefficiency. In this respect, chapter IV investigates the profitability of a trading rule based on available information and concludes that the implemented investment strategy does not generate any abnormal return on a risk and transaction cost adjusted basis. Chapter V looks at the link between the time-varying futures risk premia and the economy and demonstrates that the information variables predict futures returns because of their ability to proxy for change in the business cycle. Finally, chapter VI makes use of time-varying asset pricing models to analyse the relationship between the predictable movements in futures returns and the conditional cross section of expected futures returns. The results indicate that conditional versions of asset pricing models capture most of the predictable movements in futures returns. Hence the predictability of futures returns seem to mirror the change in the consumption-investment opportunity set over time. Chapters V and VI also raise some interesting observations that have not been evidenced to date in the literature on predictability. First, the time-variation in the expected returns of currency and agricultural commodity futures is not consistent with the evidence from the stock and bond markets and with traditional theoretical explanations of the trade-off between risk and expected return. Second, chapter VI demonstrates that shift in the sensitivities of futures returns to the constant prices of covariance risk accounts for most of the predictable movements in futures returns. This result is somehow surprising since the change in the prices of risk is the main source of predictability in the stock and bond markets.