“Do bank reserves affect interest rates when reserves are abundant?”
Job Market Paper - Latest version [PDF]

An important concern for central banks is how to control short-term interest rates, with the ultimate goal of stabilizing the economy and maintaining price stability. Prior to 2008, the Federal Reserve achieved this by changing the supply of bank reserves in the interbank market. When reserves increased from billions to trillions during and after the global financial crisis, the Federal Reserve changed how it implements monetary policy and now uses administered interest rates, notably the interest paid on reserves. In this new policy environment, is there still an impact on short-term rates from changes in reserve supply? And if so, what is the nature of this relationship, and does the impact change over time?

I estimate the effect of changes in bank reserves on the federal funds rate spread (a measure of the price of liquidity in the banking system) between 2009 to 2020 while also controlling for changes in the Fed’s policy rates. I test whether the impact of reserve changes depends on overall banking sector liquidity by using threshold models that distinguish between periods of high or low reserve balances. I demonstrate that both changes in reserves and changes in policy rates determine short-term interest rates after 2008, with three key findings. First, for reserves there is a nonlinear relationship: changes in reserves only affect the federal funds spread at lower (though still ample) levels of reserves, but not when they are very high, in line with having differently sloped sections on the reserve demand curve. Second, I find the critical threshold to be time-varying. Third, I find that reserve changes have a stronger effect after the Fed started monetary tightening in 2015. My results show that there is feedback between the Fed’s interest rate and reserve policies, which has important implications for the formulation of current monetary policy.

Work in progress

“Is inflation nonlinear or nonstationary? A cross-country comparison”

The post-pandemic global surge in inflation raised concerns that central banks might see inflation become entrenched and difficult to control once it stayed too long above target. This would happen if inflation structurally behaves differently once it increases above a certain level and a country thus enters a ”high” inflation environment. A challenge for testing this hypothesis empirically is to differentiate nonlinearity from nonstationarity, as time series can have both properties simultaneously. Inflation in particular has frequently been found to have a unit root in the literature. I analyze year-on-year inflation in six advanced economies over the past 60 years using the nonlinear unit root testing procedure by Caner and Hansen (2001) and find evidence of nonlinearity with a threshold and two regimes for all countries. The higher inflation regime is nonstationary, whereas the lower inflation regime is stationary. This means that central banks face the risk of persistent shocks to inflation above the threshold. Both the estimated threshold values at which inflation dynamics change and the speed of transition differ between countries. The distribution of regimes also matches notable inflationary shocks such as globally during the 1970s and in Europe in the early 1990s. The results suggest that country-specific monitoring of transition risk to the high inflation regime is necessary.

“Does the impact of monetary policy depend on how that policy is conducted?”
with Michael D. Bradley

We investigate how the impact of monetary policy depends on the nature of policy implementation. We find that when the Fed has been pursuing neutral or expansionary policy, an increase in the federal funds rate has modest effects on real GDP and inflation, with a substantial lag. However, when rate increases occur after previous increases, the impact on real GDP and inflation is faster and more substantial. These findings suggest that the Fed may want to initially be aggressive and persistent with interest rate increases to dampen economic activity and reduce inflation. We also analyze the effects of monetary policy during four distinct monetary policy regimes since 1969. We find that changes in the fed funds rate only have a significant impact on real GDP and inflation when the Fed actively uses it as an instrument of monetary policy, and less so when the funds rate has a less prominent role. These results contribute to an emerging literature on how both monetary policy goals and the way in which they are implemented affect macroeconomic outcomes.