In the first essay of this dissertation, “Monetary Policy and the Housing Cycle,” I investigate the role of monetary policy in a housing boom that precipitated the U.S. financial crisis of 2007. I find expansionary policy between 2002 and 2005 accounts for about 50% of the peak deviation of real residential investment from its long-run trend, which occurred in the second quarter of 2005. To determine if monetary policy was a contributor to the housing boom I estimate a large dynamic stochastic general equilibrium model (DSGE) to fit the economy in several different time periods. I mathematically isolate a series of changes in the Fed Funds rate that are statistically unrelated to changes in the macroeconomy and classify these deviations as a measure of monetary policy. The magnitude of the monetary policy series is relatively small during the housing boom but explains half of the of the 2005 peak in residential investment because of inertia in the Fed Funds rate.
My second essay follows up on a paper I wrote with Donald Haurin that was published in the Journal of Housing Economics in December 2009. The 2009 paper look at the usefulness of a housing-specific consumer sentiment index as a proxy for housing demand. The index is derived from the Survey of Consumers question: “do you think it is a good time or a bad time to buy a house?” We showed the consumer sentiment series is more predictive of turning points in the housing market than the more commonly used Housing Market Index. A detailed examination of the sentiment data reveals that interest rates play a significant role in determining consumer sentiment. In “Predicting Reversals in New House Construction” I investigate 20 economic, financial, and sentiment time series for evidence of usefulness in predicting turning points in the housing market and find several that rival the one identified in Croce and Haurin at predicting reversals. However, all of the best-performing indicators are too early to signal a housing market peak in the period before the crisis of 2007 and individual indicators performed very differently at different turning points.