UC Santa Cruz

This dissertation focuses broadly on central bank intervention: under what conditions, and when, can stabilization policy be effective? As growth lags across the globe, central banks have become lead actors in spurring recovery from the Global Financial Crisis. In response, a large academic literature has developed in order to theoretically predict and measure to what extent central bank policies have succeeded in stimulating growth.

The first two chapters of this dissertation focus on US monetary policy. I develop two new approaches for investigating the effectiveness of monetary policy at the “zero lower bound”, that is, when interest rates are near zero. A novelty of these approaches is the usage of firm- and industry-level data in order to investigate effects of monetary policy.

The zero bound on interest rates is the problem that- given low growth since 2008- the Federal Reserve would likely have lowered the federal funds rate further in to maximize employment and increase inflation. Previous behavior suggests that the Fed would have reduced the federal funds rate to a trough of 5 percentage points below zero. According to accepted empirics, a contractionary 5% monetary shock should lower GDP growth by a peak of 2.5%.

This dissertation leverages the fact that the industry effects of detrimental monetary policy would be heterogeneous: a substantial portion of this reduction in growth would occur in industries such as construction, which typically would respond with a 10% reduction in revenues to a 5% contractionary monetary shock. In contrast, health care and other industries do not tend to respond as much to interest rates changes, and would thus likely be unaffected

The first approach of this dissertation shows that growth rates of industries historically sensitive to interest rate changes, and therefore to monetary policy, do not seem to have fared poorly at the zero lower bound. This suggests that monetary policy was sufficiently accommodative since 2009 in spite of the zero bound constraint.

The second approach quantifies the federal funds rate that would be implied by observed industry growth rates and macroeconomic variables following 2008. The results suggest that unconventional policy had an effect on the economy such that growth occurred as if the federal funds rate were in fact negative. Unlike previous literature, this estimate is based on growth rates of real rather than financial market variables. The technique extracts the path of the federal funds rate at the zero lower bound consistent with pre-zero lower bound macroeconomic dynamics. Given the negative interest rate estimate from this second approach, and the relative growth rates of interest rate sensitive industries, the first two chapters suggest that the combination of zero interest rates and unconventional policy succeeded in spurring economic growth in a comparable fashion to previous US recoveries.

The third chapter of this dissertation focuses on an international aspect of central banking. Due to a combination of desires for precautionary savings and exchange rate undervaluation, central banks often accumulate reserves. This chapter examines the relationship between foreign reserve accumulation and domestic investment. A central bank's purchases of reserves implies an outflow of domestic savings. If capital markets are imperfect, foreign capital cannot be expected to compensate for this savings outflow. The third chapter thus measures to what extent reserve accumulation is correlated with decreases in domestic investment. In contrast to Reinhart et al., 2016, I find that reserve accumulation does not seem to hamper domestic investment. I conclude with a brief discussion of other possible explanations for the finding and directions for future research.

The layout of this dissertation differs slightly from that of the three separate chapters. The first chapter presents the two approaches created for measuring monetary policy at the zero bound. The second chapter expands upon these two approaches, and shows that the findings are highly robust to specification choices, and unlikely to be caused by alternative explanations. Finally, the third chapter presents the portion of this dissertation examining open economy central bank interventions.