Notes: The null hypothesis of Dickey-Fuller (1979), PP (1988) and DF-GLS (1996) is established as δ = 0; there is unit root. In the DF-GLS test, $ytd$ the detrending series of y_t; The variance of the residual obtained from the r_t = r_t−1 + u_t equation is tested. The null hypothesis of KPSS (1992) is established as $σu2$ = 0, there is no unit root; MacKinnon (1991) table critical values were used for the Dickey-Fuller (1979) and PP (1988) tests. In the ADF and PP tests, the table critical values for the level data are −3.99, −3.43 and −3.13, respectively, for 1%, 5% and 10% and −3.46, −2.88 and −2.57, respectively, for the difference series. The critical values for KPSS test statistics are taken from Kwiatkowski et al. (1992). In the KPSS test, the table critical values for level data are 0.21, 0.14 and 0.11, respectively, for 1%, 5% and 10% and 0.73, 0.46 and 0.34, respectively, for the difference series. The critical values for the test statistics of DF-GLS are taken from Elliott et al. (1996). In the DF-GLS test, the table critical values for level data are −3.46, −2.93 and −2.64, respectively, for 1%, 5% and 10% and −2.57, −1.94 and −1.61, respectively, for the difference series;

*** and * imply rejection of the unit root at the 1%, and 10% levels, respectively, for ADF, PP and DF-GLS; *** and * imply rejection of the no unit root at the 1% and 10% levels, respectively, for the KPSS test