Bootstrapping the autoregressive distributed lag test for cointegration

We propose a bootstrap autoregressive-distributed lag (ARDL) test. By applying the appropriate bootstrap method, some weaknesses underlying the Pesaran, Shin and Smith ARDL bounds test are addressed including size and power properties and the elimination of inconclusive inferences. In addition, inferences based solely on the significance of the F-test and single t-test from the ARDL bounds test are not sufficient to avoid degenerate cases. The bootstrap ARDL test provides an additional test on the significance of coefficients on lagged levels of the regressors, which provides a better insight into the cointegration status of the model.