ABSTRACT
This article investigates the time-series behavior of the world crude oil prices over the period from January 1997 to May 2011 by taking into account nonlinearity and fractional integration in the same framework. As a first step, the author examines the nonlinear behavior of the data through the estimation of a two-regime threshold autoregressive model and find that world crude oil prices exhibit a nonlinear behavior. After finding nonlinearity, the author also allows disturbances to be fractionally integrated. The results point to a stationary process with long memory.
Acknowledgment
The author thanks Bruce Hansen for making available his Gauss codes for the TAR model and Mehmet Balcılar for kindly providing the Gauss codes for Robinson test.