| Summary: | 博士 === 國立臺灣大學 === 經濟學研究所 === 105 === This dissertation consists of three essays on money policy and macro-prudential policy. The first chapter investigates the implications of different policies using a two-friction model and welfare analysis. The second part studies the asymmetric effect of monetary policy by constructing a model with occasionally binding constraint. The third chapter revisits the treatment effect of the loan-to-value policy in Taiwan by conducting mediation analysis.
Chapter 1 develops a dynamic stochastic general equilibrium (DSGE) model in which firms and intermediaries are both subject to borrowing constraints. We then study the welfare implications of the macroprudential policies involving a countercyclical loan-to-value (LTV) ratio and credit market interventions. Impulse responses show that the stabilization effect of a policy rule largely depends on the source of shock. The welfare analysis indicates that the LTV ratio rule is Pareto improving, whereas the credit intervention is not.
In particular, the LTV rule responding to output enhances the welfare of individual groups and social welfare most significantly. On the other hand, although the credit intervention policy raises the output level and households'' consumption, it fails to improve agents'' well-being because it leads to higher volatility of major economic aggregates.
In Chapter 2, we first use a structural vector autoregression model to examine whether the US economy responds asymmetrically to expansionary and contractionary monetary policies. The empirical results show that monetary policy has significant asymmetric effects on output and investment. To provide an explanation of such asymmetries, we consider a nonlinear dynamic stochastic general equilibrium (DSGE) model in which collateral constraints are occasionally binding over the business cycle. The nonlinear DSGE model is able to match the empirical findings that macroeconomic aggregates react asymmetrically to positive and negative monetary policies.
Chapter 3 studies whether and how loan-to-value (LTV) ratio policy affects house prices, by conducting mediation analysis in the context of endogenous treatment status. Taking the LTV policies as a treatment, our analysis focuses on whether the policy can tame house prices via a credit contraction channel, which is treated as a mediator. We identify the credit contraction channel as an indirect effect, while impacts that do not operate through the mediation channel are regarded as direct effects. Using Taiwanese quarterly data from 2008:Q2 to 2014:Q3, we find that the LTV policies have negative effects on house prices and the impact mainly comes from direct effects rather than through the mediator. However, neither the direct nor the indirect effects are statistically significant.
|
|---|
