Multiple Equilibrium Of Brent Crude Oil Future Prices

Scientifically understanding of the fluctuation behavior of oil price is a major issue that contains a number of important management science problems. This thesis used Brent crude oil futures settlement price as the research object and applied stochastic process model, price-volume relationship model, maximum entropy spectral analysis, wavelet analysis and econometric model to research Brent oil price’s behavior, time features of the price as well as the interaction relationship between oil prices and other economic indicators. These results could help us to understand current oil fluctuation and predict future oil price.Analyzing the relationship between price and volume, we found out that oil prices have multiple equilibria, the consumption’s change in trend would lead oil prices to transfer its equilibria. Phase-portrait and probability density function analysis can prove the existence of multiple equilibria in oil prices. Oil prices’fluctuation behavior near equilibrium positions could be described by Ornstein-Uhlenbeck mean reversion process and the equilibria transfer behavior of oil price could be described by a jump-diffusion process. Then, we built a volume-price model to describe the mechanism which controls the behavior of oil prices. We used Maximum Entropy Spectral Analysis technique to analyze oil prices main time cycle and found out that oil prices has four main cycle components:10.7months,25.6months,5.3months and3.6months. These cycles’ length and the durations that oil prices stay at equilibrium positions present a multiple relationships. We decomposed the crude oil prices between Jul.1988to Oct.2013into9-layer fluctuation signals and trend signals by using db4wavelet. We found out that the behavior information of oil prices’fluctuations near the equilibrium position mainly contained in1-3layer fluctuation signals whose time scales are relatively short and the behavior information of oil prices’equilibrium transfersis are mainly contained in4or above layers fluctuation signals whose time scales are relatives long. Finally, we wavelet decomposed crude oil prices, the U.S. total capacity utilization, German DAX30index and the dollar index in Feb.1991to Sep.2013and tested the4and above layers fluctuation signals conintegration relationship these economic variables by using co-integration theory. We found that the equilibrium transfersis of oil prices is affected by the real economy, the stock market and the dollar index.