Piffer, Michele
(2014)
An analysis of leverage ratios and default probabilities.
PhD thesis, London School of Economics and Political Science.
Abstract
The thesis consists of three independent chapters.
In Chapter 1 (page 7) - Counter-cyclical defaults in “costly state verification” models - I argue that a pro-cyclical risk-free rate can solve the problem of pro-cyclical defaults in “costly state verification” models. Using a partial equilibrium framework, I compute numerically the coefficient of a Taylor rule that delivers pro-cyclical output, pro-cyclical capital and counter-cyclical defaults. This parametrization is consistent with the empirical evidence on Taylor rules.
In Chapter 2 (page 67) - Monetary Policy, Leverage, and Default - I use the Bernanke, Gertler and Gilchrist (1999) model to study the effect of monetary policy on the probability that firms default on loans. I argue that a monetary expansion affects defaults through two opposing partial equilibrium effects. It increases defaults because it leads firms to take on more debt and leverage up net worth, and it decreases defaults because the cost of borrowing decreases and aggregate demand
shifts out, increasing firms’ profits and net worth. I argue that the leverage effect could explain the empirical partial equilibrium finding by Jimenez et al. (2008) that defaults on new loans increase after a monetary expansion. I then argue that this effect does not hold in general equilibrium due to an increase in firms’ profits. In the full model, defaults decrease after a monetary expansion, although the effect
equals only few basis points.
In Chapter 3 (page 131) - Monetary Policy and Defaults in the US - I study empirically the effect of an unexpected monetary expansion on the delinquency rate on US business loans, residential mortgages and consumer credit. I consider several identification strategies and use Granger-causality tests to assess the exogeneity of
the time series of monetary shocks to the Fed’s forecast of defaults on loans. I then compute impulse responses using a variant of the Local Projection method by Jorda (2005). I find that the delinquency rates do not respond to monetary shocks in a statistically significant way. The paper suggests that the risk-taking incentives
analyzed in partial equilibrium by several existing contributions might not be strong enough to prevail and increase aggregate defaults, although they could explain why defaults do not significantly decrease.
Actions (login required)
 |
Record administration - authorised staff only |
Download statistics
Download statistics