Loan Loss Provisioning and Differences of Opinion

53 Pages Posted: 19 Oct 2011

See all articles by Lin Cheng

Lin Cheng

The University of Arizona - Eller College of Management

Date Written: October 1, 2011

Abstract

Prior research suggests that differences of opinion of professional forecasters are indicative of risk and uncertainty. This study investigates whether different attributes of loan loss provisions influence the likelihood and magnitude of disagreement between the two major credit rating agencies. Focusing on a sample of bank issues from 1994 to 2009, I find that Moody’s and S&P disagree on the initial ratings of 66% of the issues and that S&P is the more conservative rater who gives a lower rating 89% of the time when there is a disagreement. I also find that S&P is more likely to give a lower rating during periods of higher economic uncertainty. In addition, I hypothesize and find that lower timeliness and validity of loan loss provisions exacerbate the likelihood and magnitude of initial credit rating splits between the two major credit rating agencies. Furthermore, discretionary loan loss provisions are positively associated with the likelihood and degree of differences of opinion. Finally, I find bank bondholders demand a yield premium on split rated bonds. The findings in this paper suggest that accounting information quality influences the perceived risks and uncertainties in banking industry and is useful in facilitating bondholders’ market discipline of banks’ risk exposures.

Keywords: Rating disagreement, Loan loss provisions, Accounting quality, Rating agency, Banking

JEL Classification: G21, G24, G32, M41

Suggested Citation

Cheng, Lin, Loan Loss Provisioning and Differences of Opinion (October 1, 2011). Available at SSRN: https://ssrn.com/abstract=1945929 or http://dx.doi.org/10.2139/ssrn.1945929

Lin Cheng (Contact Author)

The University of Arizona - Eller College of Management ( email )

McClelland Hall, Room 301Q
1130 E. Helen Street
Tucson, AZ 85721
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
248
Abstract Views
3,063
Rank
224,198
PlumX Metrics