| Summary: | This dissertation tests the efficiency of selected NYMEX petroleum futures spreads. It is argued that seasonal changes in hedging activity cause seasonal biases in futures price spreads. A multimarket equilibrium model is presented which illustrates the effect of seasonal hedging behavior on spreads. In addition, the potential of nonseasonal hedging and speculator interest to diminish or eliminate seasonal biases is discussed. Two hypotheses are tested. The primary hypothesis is: NYMEX futures markets for No. 2 heating oil, unleaded gasoline, and crude oil are not semi-strong efficient. If futures markets are not semi-strong efficient then all publicly available information is not reflected in contract prices. The secondary hypothesis is: by anticipating seasonal changes in hedger interest in NYMEX No. 2 heating oil, unleaded gasoline, and crude oil futures markets, a speculator can profit from changing NYMEX futures spreads. Nine speculative trading rules are presented to test the two hypotheses. They specify the contracts to spread so the speculator can profit from theorized biases in selected NYMEX futures spreads. The number of contracts to spread and the period the spreads are held is clearly specified. The trading rules are tested over five trading years. An econometric model then analyzes the profitability of the spreads over time to determine if trends exist which would indicate spread biases. The empirical results support the primary and secondary hypotheses of this dissertation. Several of the trading rules profited over the five-year test period, and when losing years occurred, losses were small relative to profitable years. Nonseasonal hedging and speculator interest are argued to cause losses occasionally. Nevertheless, the results suggest that a speculator can profit in the long run by trading selected NYMEX spreads. The long run is defined in this context as five years or more. The profitability of some of the rules over time exhibited non-random behavior. These results suggest that spread biases caused profits to steadily increase during several spread periods. This evidence supports the hypothesis that NYMEX futures markets are not semi-strong efficient; it suggests that biases do exist.
|
|---|
