The Roosevelt Project
Miguel Herce, John E. Parsons and Robert C. Ready, April 2006
Oil prices are very volatile. But much of this volatility seems to reflect short-term,transitory factors that may have little or no influence on the price in the long run. Many major investment decisions should be guided by a model of the long-term price of oil and its dynamics. Data on futures prices can be used to filter out the short-term volatility and recover a time series of the latent, long-term price of oil. We test a leading model known as the 2-factor or short-term, long-term model. While the generated latent price variable is clearly an improvement over the raw spot oil price series, we also find that (1) the generated long-term price series still contains some of the short-term volatility, and (2) a na´ve use of a long-maturity futures price as a proxy for the long-term price successfully filters out a large majority of the short-term volatility and so may be convenient alternative to the more cumbersome model.