Oil Price Volatility Modeling: a frame for Measuring Uncertainty Index using the ARIMA-GARCH Model

This article attempts to model oil price volatility through use of conditional variance models (GARCH) in the time interval between January 1983 and December 2010. We present an index to measure the uncertainty, using calculated conditional variance parameters. Based on the results obtained, all models used in this study confirm the existence of conditional variance structure for the time series Iranian oil price data.Moreover models such as TARCH & EGARCH were used to assess leverage effects and both models confirmed such leverage effects. The results of GARCH coefficient tests show that in the long run, conditional variance returns to its average level. In addition, the uncertainty index shows a maximal divergence of 8 percent between similar seasons, while all such variances are contained within the domain of between 1 to 8 Percent.