The Impact of Regulation and Economic Conditions on the Dynamics of Financial Markets

by Stefan Kerbl

Institution: Vienna University of Economics and Business
Year: 2012
Keywords: RVK QK 600, QK 950; Österreich / Kreditmarkt / Volatilität / Kreditrisiko / Risikomanagement / Kreditwesen / Regulierung / mathematisches Modell / Aufsatzsammlung
Record ID: 1031793
Full text PDF: http://epub.wu.ac.at/3572/1/Dissertation_Stefan_Kerbl.pdf


This dissertation encompasses four studies on selected topics in financial regulation and financial stability. The first paper asks whether there is empirical evidence of cyclicality in regulatory capital requirements prescribed by Basel regimes. This much debated issue was until then only addressed in theoretical papers, or simulation studies. While we do not find evidence on cyclicality in the Basel I or Basel II Standardized Approach, we find statistically and economically significant evidence concerning Basel II IRB portfolios. The second paper implements an agent based model to simulate an artificial asset market. This setup is then used to assess the impact of (i) a short selling ban, (ii) a Tobin Tax like transaction tax, (iii) mandatory Value-at-Risk limits and (iv) arbitrary combinations of these. I present results that show that while reducing volatility, a short selling ban nurtures market bubbles, and a Tobin Tax increases the variance of the returns. In this model a mandatory risk limit is beneficial from all stability perspectives taken. I examine the robustness of the model regarding its initial parameterization and show that high levels of a Tobin Tax lead to substantial market turbulence. The third paper considers the question which macroeconomic variables are linked to a time series of special interest from a financial stability perspective: firm defaults. Furthermore, we evaluate the empirical evidence of a hidden credit cycle by adding a latent factor to our models. We conclude that there is no empirical support of a hidden credit cycle in Austria once sufficient regressors are included and industry sectors differ in their respective macro drivers. The forth paper extends this work by implementing Bayesian Model Averaging (BMA) - a modern technique to counter model uncertainty. Furthermore we enrich this statistical approach by combining BMA with Bayesian ridge regression. We draw the conclusion that BMA is indeed a powerful tool to counter model uncertainty. Interest rates and components of inflation are distilled as major drivers for firm failures in Austria. (author's abstract)