Monetary shocks, oil shocks, and macroeconomic activity: A cointegration analysis

The research problem is: Using the net increase in the relative oil price as a proxy for oil prices, does the econometric technique of Crowder, et al. (1994) provide any evidence that oil and/or monetary shocks influence U.S. industrial production? Derived from the research problem, four research questions are stated regarding (1) the shifts in monetary policy in early 1980s, (2) the shifts in energy policy of the Texas Railroad Commission, (3) the repercussions of such shifts on the persistence of the eventual impact of these aforementioned variables on the U.S. macroeconomic activities, and (4) the stability of the long-run relationships.;The study has a quantitative form. The approach of vector autoregression takes into account the interdependence of the variables. The variables are the industrial production, the inflation rate, the real money balances (rm1), the 6-month commercial paper rates, and the net increase in the relative oil price. The data on the first four variables are from Citibase whereas those of the net increase of the relative oil price were constructed following Hamilton's method (1996). I employ cointegration analysis to estimate and test the long-run relationships. I conduct basic structural break tests (Chow tests on the first-differenced data) to isolate whether shifts occurred in the long-run relationships in line with Hooker and Hamilton's debate. I conduct stability tests following the procedure of the recursive estimation in cointegrated VAR-models (Hansen and Johansen, 1992) to evaluate the parameter constancy.;The study reveals that the long-run behavior of all 5 variables is determined by two distinct shocks that appear to be real and nominal in origin: the technology and inflation shocks. The remaining transitory shocks were those of money supply, IS and oil. The study revealed no indication of a monocausal explanation of cyclical fluctuations. The impulse response functions of output show no significant effect of an initial shock to oil prices on industrial production, a significant impact of technology, inflation (for 12 months), real money balances (up to 12 months) and IS shocks (for up to 12 months). These findings are consistent across the analyses of forecast-error variances and of historical variances. Historically, technology shocks have a distinct impact on economic performance at all times. The period 1959 to 1973 is largely characterized by the contributions of IS and monetary shocks. Monetary disturbances seem to have played a large role in the 1969-1973 period. The 1973-1979 period is dominated by inflation, monetary and IS shocks. The magnitude of the oil impact on industrial production is low. Oil disturbances contribute to explaining the industrial production's fluctuations during the 1990's only. (Abstract shortened by UMI.).