Abstract:
<p>This dissertation consists of three essays on macroeconomics in the frequency domain. In the first essay, I show that whereas the High-Frequency volatility of the majority of the macroeconomic series declined after the early 1980s, their Medium-Frequency volatility did not. Moreover, the Medium-Frequencies capture a large fraction of the volatility of these variables. In order to formally test whether a set of time-series is characterized by a break in their variance at any frequency, I construct a frequency domain structural break test. After deriving its asymptotic and small sample properties, I apply the test to the main U.S. real macroeconomic variables and conclude that the Great Moderation is just a High-Frequency phenomenon.</p>
<p>In the second essay I compute the welfare cost of the Great Moderation, using a consumption based asset pricing model. The Great Moderation is modeled according to the data properties of the stationary component of consumption, which displays a reduction of the volatility at high frequencies, and an unchanged volatility at medium frequencies. The theoretical model, calibrated to match the average asset pricing variables in the data, relies on the evolution of the habit stock, which depends on the lower frequencies of consumption. These two features generate a modest welfare gain of the Great Moderation (0.6 percent). I show that this result depends mainly on the medium frequency properties of consumption.</p>
<p>The third essay, which is joint work with Marija Vukotic, evaluates the effects of a change in monetary policy on the decline of the volatility of real macroeconomic variables, and on its redistribution from high to medium frequencies during the post-1983 period. By using a dynamic stochastic general equilibrium (DSGE hereafter) model, we find that the</p>
<p>monetary policy alone cannot account for the observed changes in the</p>
<p>spectral density of output, investment, and consumption. However, when we also consider a change in the exogenous processes, a different monetary policy accounts for 40 percent of the decline in the high-frequency volatilities and partially accounts for the redistribution of the variance toward lower frequencies.</p>
