An equilibrium model of the pricing of interest rate futures contracts : theory and empirical tests

This study develops and tests a model which determines the equilibrium price of interest rate futures contracts. The derivation of the model is based on the assumptions that all participants maximize expected utility of terminal wealth in a mean-variance framework and that expectations are homogeneous. A rationale is provided to divide the participants into two groups, hedgers and speculators. The function of hedgers is to reduce their exposure to interest rate risk and the function of speculators is to take advantage of transitory profit making opportunities that will arise if there is excessive hedging activity on either the long or the short side. The demand for long contracts and the supply of short contracts for all the participants are determined. The total demand is equated to total supply to determine the equilibrium futures price. It is determined that the equilibrium futures price is a function of the expected futures price, the risk premium transfer from hedgers to speculators, and the costly guarantee which is the net cost of the margin maintenance requirements. The equilibrium futures price will include a positive risk premium if there is excessive hedging activity on the long side (hedgers are net long), a negative risk premium if there is excessive hedging activity on the short-side (hedgers are net short), and no risk premium if there is equal hedging activity on both sides. The costly guarantee will have a positive effect on the equilibrium futrues price if hedgers are net long, a negative effect if hedgers are net short and no effect if there is equal hedging activity on both sides. This holds true only if the classification of participants as hedgers and speculators by this study is the same as the classification defined by the CFTC. If the classifications differ, then the effect of the costly guarantee is independent of hedging activity. The purpose of the empirical tests is to determine if (1) there is a significant positive or negative risk premium that tends to zero as the length to maturity decreases and (2) if the costly guarantee has a significant effect on the equilibrium futures price. The tests are performed with four different empirical specifications of the model. The four empirical specifications are due to two methods used to estimate the expected price- (1) the distributed lag approach and (2) the instrumental variable approach and two methods used to estimate the costly guarantee - (1) the absolute price differential approach and (2) the standard deviation approach. The results for the GNMA and T-bill futures contracts indicate that there is no significant risk premium included in the equilibrium futures price and that the costly guarantee has no significant effect. The results for the T-bond futures contracts are sensitive to the empirical specification of the model and therefore it is difficult to provide conclusive evidence.