The Effect Of Monetary Policy On Determination Of Coal Prices

1013 Words5 Pages
To examine the effect of monetary policy on determination of coal prices in the U.S., we rely on a dynamic framework of cointegrated vector autoregression (CVAR). The prerequisite requirement for the application of the CVAR approach is that the selected variables must be nonstationary (i.e., I(1) series). The presence of a unit root in the variables is thus tested using the Dickey Fuller generalized least squares (DF-GLS) test (Elliott et al., 1996). Panel A of Table 1 reports the results of the DF-GLS test. Since the null hypothesis of a unit root cannot (can) be rejected for any of the levels (first differences) of the three variables at the 5% level, all the series are found to be nonstationary I(1) processes.
It should be emphasized, however, that because of inability of the DF-GLS to capture the possibility of a structural break, the power of the test is likely to decrease with an undetected structural break in the series, thereby providing misleading results. For completeness, therefore, we investigate unit roots in the presence of a structural break using the Perron-Vogelsang (PV) test. The results of the PV test are summarized in Panel B of Table 1. As seen in the DF-GLS test, the null cannot (can) be rejected for all the levels (first differences) of the variables, confirming that all the three variables are I(1) series. Hence, it is certain that the underlying series are apparently all I(1) processes even after taking into account a structural break in the series,
Get Access