Skip to content
All library documents

Testing Whether Regulation Changed Credit Rating Event Effects

Article Quant Q&A · Author: Lickt0rn
Read the original

Summary

The document asks how to test whether stock-price reactions to Moody’s credit rating changes became smaller after a financial regulation. The author has already split downgrade events into pre- and post-regulation groups and run market-model event studies, finding day-zero effects significant under a simple t-test, Patell Z, and Corrado’s rank test.

Those separate significance results do not establish whether the two groups differ. The core methodological issue is to test the difference between the estimated effects directly, potentially through a multivariate regression model or another comparison that accounts for event-study data. The document points to Binder’s multivariate regression approach as a possible reference, but does not provide a worked method, estimates, or a conclusion. Its evidence is therefore limited to the author’s reported within-group significance; sample sizes, effect estimates, dependence across events, and regulatory timing details are not supplied.

Key ideas

  • Significant effects in two separate samples do not show that the effects differ significantly from each other.
  • The research question is whether downgrade-related stock returns changed after a regulation.
  • The author reports day-zero significance using a market model, a simple t-test, Patell Z, and Corrado’s rank test.
  • A direct test of the difference between groups is needed, with multivariate regression raised as one possible approach.

Tags

EquitiesEvent-drivenStatistics
Full text
# Test for difference in security returns before and after financial regulation


# Test for difference in security returns before and after financial regulation












I'm going to study the effect of corporate credit rating changes (Moody's) on stock prices before and after a specific financial regulation. So far i have used an event study where i have divided the events in two different groups (0: Before, 1: After), and studied the effects individually.

Downgrades (Day 0) - I have used the market model, and different tests like the simple t-test (MacKinlay - 1997), Patell Z (Patell - 1976) and Corrado's rank test (Corrado - 1992) are all statistically significant at the 0.05 or 0.1 level.







But, i'm not sure how to compare these results. In other words, i'm not sure how to test if the effect of credit rating changes after the regulation is smaller than the effect before the regulation.

I have looked at this; StackExchange - Event study topic but didn't quite get it. I'm not sure how you can say that "a" is bigger than "b" without some kind of test. I have also looked at a couple of papers about multivariate regression models (with and without bootstrapping) but they tend to be too complicated for a non-statistican.

I would really appreciate it if someone could help me with how i can test if there is a significant difference between the two groups or help me get how to do this with multivariate regression models (eg. Binder, J. J. (1985). On the use of the multivariate regression model in event studies.). Thanks in advance!

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.