Essays on Monetary and Macro-prudential Policy in a DSGE Model with Banking Sector

The thesis is composed of three chapters which analyze the monetary and macro-prudential policy using a New Keynesian DSGE model with the banking sector.

The first chapter evaluates the effectiveness of time-varying macro-prudential tools and their interaction with monetary policy using the model framework of Gertler and

Karadi (2011) which emphasizes the role of bank capital and endogenous leverage constraint due to the moral hazard problem. We consider the following three types of countercyclical macro-prudential tools: (i) capital requirements (CAR), (ii) deposit-based reserve requirements (DBRR), and (iii) asset-based reserve requirements (ABRR). The main findings are as follows. First, all three types of macro-prudential policy tools reacting to the financial condition (credit to output gap) are effective to mitigate the fluctuation of credit and business cycle by weakening the pro-cyclicality of banks' net worth and the counter-cyclicality of the endogenous leverage and external finance premium. Second, the effectiveness of each instrument differs by the source of shocks to the economy. In response to a TFP shock, the counter-cyclical CAR is the most effective to mitigate the fluctuation of credit and output and the ABRR is relatively more effective in response to capital quality shocks (credit supply shock). Third, all types of macro-prudential tools are welfare improving in combination with monetary policy, but the use of macro-prudential policy is relevant in response to the credit supply shock (financial shock). Finally, when the economy is hit by a financial shock, a relatively direct tool like asset-based reserve requirement (ABRR) can be the most welfare-improving and the relationship between monetary and macro-prudential policy can be complementary each other.

The second chapter examines the role of banks' endogenous reserves and capital and the usefulness of financial condition (credit aggregates and external finance premium) as an indicator for setting up a relevant monetary policy rule. For the task, I construct an extended version of Goodfriend and McCallum (2007) model which is characterized by banks' loan production function with inputs (collaterals, monitoring works and bank capital buffer) under the households' deposit in advance (DIA) constraint. Main findings are as follows. First, banks' endogenous reserves can mitigate the variation of external finance premium by substituting costly monitoring works for lending. However, the effect of endogenous reserves on output and credits depends on types of shocks (credit-demand vs credit-supply shock), which reflects the role of external finance premium (EFP) as a banking attenuator (pro-cyclical EFP) or financial accelerator (counter-cyclical EFP) respectively. Second, the effect of banks' endogenous reserves can be limited when banks can adjust their capital flexibly compared to when banks' keep their capital to asset ratio (leverage) stable. Third, considering the role of money and the frictions on the credit-supply side, monetary policy reacting actively to the financial condition is welfare improving. These results imply that quantitative easing (QE) policy during the crisis was effective but the effectiveness of QE can be limited when banks can lower their leverage in a pro-cyclical manner. And monetary policy should pay more attention to the financial condition and prevailing interest rates as well as inflation and output gap in the presence of substantive financial frictions on the credit supply side in particular when the banking shocks dominate the economy.

In the last chapter (co-authored with Professor Jagjit S. Chadha), we investigate the question facing many emerging economies: the extent to which a workhorse advanced economy model can yield important insights for monetary policy-making. We note that the standard sticky-price, monopolistically competitive model does not allow analysis of money and credit dynamics and led to a concentration of research on simple interest rate reaction functions. However, time-varying financial frictions tend to act as a tax on intermediation activities and so can vary output in a significant manner. In this chapter,

we consider the implications of financial frictions for baseline monetary policy using a model calibrated on Indian data and find that a simple interest rate reaction function may not be welfare maximizing when banking shocks are dominant in the economy.