Non-Linear ARDL Model of Price Index Dynamics of Transportation Sector

In this article, using the autoregressive model with non-linear distribution breaks, we investigate the dynamics of the price index of the transportation sector during the years 2010-2019. In order to investigate the shocks of the variables affecting the price index of the transportation sector, at first, using Hodrick-Prescott filter, predicted shocks and unexpected positive and negative shocks have been extracted. The results of the estimation of the model show that both in the short term and in the long term, the positive shocks of the added value of the transportation sector, interest rate, crude oil price and exchange rate have a positive effect and the negative shocks of these variables have a negative effect on the price index of the transportation sector. Also, the Wald test was used to check the asymmetry of the positive and negative shocks of the investigated variables. The results show that the unexpected positive and negative shocks of the investigated variables have asymmetric effects on the price index of the transportation sector. In the following, CUSUM and CUSUM SQ tests were used for the residual sentences of the short-term model to investigate the stability of the estimated parameters in the long-term model. The results of this test show that the coefficients of the variables are stable during the investigated period. In other words, there is no structural break in the model, and the average of residual sentences is zero, and the first classical assumption is maintained.