This dissertation is a collection of three essays that revisits the long-standing puzzle of the apparently disproportionate effect of oil prices in the economy by examining output growth predictability with real-time data. Each study of the predictive content of oil shocks is from a different perspective by using newly developed real-time datasets, which allows for replicating the economic environment faced by policymakers in real time.;The first study extends the conventional set of models of output growth determination by investigating predictability of models that incorporate various functional forms of oil prices and real-time data. The results are supportive of the relationship of GDP and oil in the context of Granger causality with real-time data. In the second essay, I use oil shocks to predict the economy is changing direction earlier than would be predicted by solely using initial GDP releases. The model provides compelling evidence of negative GDP growth predictability in response to oil price shocks, which could shorten the "recognition lag" for successful implementation of discretionary counter-cyclical policies.;In the third essay, I evaluate short-horizon output growth predictability using real-time data for different sample periods. I find strong evidence of predictability at the one-quarter and four-quarter horizon for the United States. The major result of the paper is that we reject the null hypothesis of no predictability against an alternative hypothesis of predictability with oil shocks that include yield curves in the forecasting regression. This relationship suggests the combination of monetary policy and oil shocks are important for subsequent GDP growth.
展开▼
