The Effects of Mandated Stress-Testing on Bank Risk

<p>In the aftermath of the Financial Crisis, The United States Congress passed the Dodd- Frank Wall Street Reform and Consumer Protection Act in 2010 (12 USC 5365(i)(1)) with particular rules on how Bank Holding Companies (BHC’s) are to be supervised. If a BHC has total consolidated assets greater than $50bn then they are required to undergo the Comprehensive Capital Asset Review (CCAR) exercise. If a BHC has total consolidated assets between $10bn and $50, they are required to undergo Dodd-Frank Act Stress-Testing (DFAST). The purpose of stress-testing and classifications are to ensure that BHCs have sufficient capital and planning processes in place to withstand future crises. However, the classifications were seemingly set arbitrarily and there is a need to investigate whether the regulation has the intended purpose of BHC risk reduction. By using a novel quasi-experimental design strategy, this research investigates and estimates whether the regulation had the intended effect of reducing bank risk. First, two unique measures of bank risk (Z-score) are developed and estimated for 220 banks in a sample from The Banker’s Database. These measures are used to develop an Regression Discontinuity Design (RDD) empirical strategy to test the DFAST and CCAR thresholds, $10bn and $50bn, respectively. The thresholds are tested as an overall period from 2012 - 2018 and by year. Additionally, extensive use of a McCrary Test for manipulation of the running variable is used as a robustness check. It is a statistical method that allows for detection of whether BHC’s or a regulatory agency adjusted total assets in order to be categorized differently. There is no evidence for partial or complete manipulation. BHCs that participated in CCAR stress-tests, experienced a near doubling of their Z-score score, interpreted as a major decrease in bank risk. The results indicate a 106% increase in Zscore at the $50bn CCAR threshold for the overall period of 2012 - 2018. So, at the threshold of $50bn, comparing the average bank above the threshold to below, the bank that was CCAR stress-tested had a 106% increase in Z-score compared to the bank that was not CCAR stresstested. This also implies a dramatic decrease in the probability of default. Additionally, the by year testing indicates a positive magnitude of risk reduction. Particularly strong results are found for risk reduction effects for CCAR stress-tested banks in 2015 and 2016 (332% and 208%, respectively compared to a base year of 2012) . The results indicate a similar magnitude for the $10bn DFAST threshold for the overall period of 2012 - 2018, but are not statistically significant. Similarly, the by year testing indicates a positive magnitude of risk reduction for small/medium sized banks.</p>

Name thesis:degree_name

Economics, PhD

Level thesis:degree_level

Restricted to Claremont Colleges Dissertation

Discipline thesis:degree_discipline

School of Social Science, Politics, and Evaluation

Year dc:date.available

2020

Author dc:creator

• Mathur, Nikhil

Contributors dc:contributor

• Hisam Sabouni
• Clas Wihlborg
• Pierangelo De Pace

• Banking
• Financial Crisis
• Financial Intermediation
• Financial Policy
• Government Policy and Regulation
• Risk Management
• Economics
• Finance

Repository record dc:identifier

https://scholarship.claremont.edu/cgu_etd/248

OAI identifier oai:identifier

oai:scholarship.claremont.edu:cgu_etd-1328