Does the Federal Reserve cause market bubbles by diverging from Taylor rule?

Abstract

In this thesis, I study how the Federal Reserve affects financial market valuations and stock market liquidity. My work relates to previous literature that uses monetary policy tools to explain stock market liquidity and stock market volatility. Previous research confirms that monetary policy affects equity valuations and risk-aversion of investors. Austrian school theory views business cycles as a symptom of central bank interventionism in credit market and monetary field. Contribution of this thesis is to provide explanation on transmission mechanism of monetary policy to financial market valuations. Emphasis is on employing stock market liquidity as a transmission channel of monetary policy.

I calculate how the Federal Reserve's short term rate has diverged from Taylor rule -implied rate, and define divergences from Taylor rule as expansive or restrictive monetary policy. I use Q-ratio to measure financial market valuations, which is a ratio between U.S. corporate equities and U.S. corporate net worth. I define stock market liquidity as price impact to dollar volume trading in NYSE/Amex stocks. Data spans from 1954 to 2007.

My first new finding shows that monetary policy affects stock market liquidity. Standard deviation effect of monetary policy increases stock market liquidity during two quarters, and explains 54.3% of variation when controlling for autocorrelation. Second new finding shows that stock market liquidity affects financial market valuations. Standard deviation increase on stock market liquidity increases level changes in Q-ratio by 4.78%, and decreases by 4.9% in previous quarter. Effects are large compared to sample mean (1.57%), and explains 18.5% of the variation in Q-ratio levels.

Results hold for numerous robustness tests. Parameter stability test shows that the estimated coefficients for my key findings are stable over time. I employ VAR -model for key variables to study joint movement of monetary policy, stock market liquidity, and Q-ratio. Based on this model, Granger causality -test shows that these variables Granger cause each other. Impulse response functions also confirm same effects I find for my new findings.