Threshold effects in the capital asset pricing model using panel smooth transition regression (PSTR) Evidence from net oil export and import groups

In this study, we used the PSTR (panel smooth transition regression) model to investigate the nonlinear relationship between beta (systematic risk) and returns (world market excess returns) for net oil export and net oil import groups. We set the volatility of world market excess return as the threshold variable and the percentage changes of crude oil price and exchange rate as the control variables. Our results support the use of a nonlinear model to elucidate the behavior both groups. We found that all beta values are positive and higher in the low regime (i.e., volatility of world market excess return is low) and lower in the high regime (i.e., volatility of world market excess return is high). For the net oil export group, the crude oil price change percentage is positive in the high regime, but the exchange rate percentage change is positive in the low regime. For the net oil import group, in both the low and high regimes, changes in crude oil price and exchange rate have equally positive effects on the individual market excess return.