摘要：It is very common in the literature to assume that fossil fuels prices follow a geometric Brownian motion (GBM)
process. However, the GBM is an imperfect process in the sense that it cannot capture correctly the frequent extreme
movements in fossil fuel prices caused by the information generated within domestic and international markets. In this
paper, we use fossil fuel index data to compare the performance of the GBM, which is based on normal density, with
that of the variance-gamma (VG) process, which allows us to capture the skewness and the excess of kurtosis of price
returns. We show that the VG process fits the data better than does the GBM, as indicated by several goodness-of-fit
tests.
