Structural Shifts in Credit Rating Standards

• Date: 01 December 2013
• Published date: 01 December 2013
• DOI: http://doi.org/10.1111/jofi.12070
• Author: AYSUN ALP

THE JOURNAL OF FINANCE •VOL. LXVIII, NO. 6 •DECEMBER 2013

Structural Shifts in Credit Rating Standards

AYS UN A LP ∗

ABSTRACT

I examine the time-series variation in corporate credit rating standards from 1985 to

2007. A divergent pattern exists between investment-grade and speculative-grade

rating standards from 1985 to 2002 as investment-grade standards tighten and

speculative-grade loosen. In 2002, a structural shift occurs toward more stringent

ratings. Holding characteristics constant, firms experience a drop of 1.5 notches in

ratings due to tightened standards from 2002 to 2007. Credit spread tests suggest

that the variation in standards is not completely due to changes in the economic

climate. Rating standards affect credit spreads. Loose ratings are associated with

higher default rates.

CREDIT RATING AGENCIES face widespread criticism regarding the quality of their

ratings. The agencies came under fire after the accounting scandals of Enron

and WorldCom, but that scrutiny was never as severe as the scrutiny following

the 2008 financial crisis. During the crisis, mortgage-backed securities, origi-

nally rated AAA, were rapidly downgraded or experienced substantial losses,

suggesting that these securities had inflated ratings. Influential economists,

commentators in the popular press, and hearings in the congressional com-

mittees suggest that the rating standards were lax, perhaps due to conflicts of

interest arising out of revenue models in which rating companies derive fees

from the issuers whose securities they rate.1The issuer-pays revenue model

creates incentives for the rating agencies to assign inflated ratings, which con-

tributed to the 2008 financial crisis by masking the true risk of underlying

securities.

∗Aysun Alp is at Sabanci School of Management, Sabanci University, Istanbul, Turkey. I am

indebted to an anonymous referee, Cam Harvey (the Editor), and an Associate Editor for detailed

comments that greatly improved the paper. I am also grateful to Nagpurnanand R. Prabhala,

Haluk Unal, Albert S. “Pete” Kyle, Michael Faulkender,Gerard Hoberg, Mark Loewenstein, Dalida

Kadyrzhanova, and the seminar participants at Maryland, Sabanci, Koc, and Rowan Universities

for valuable comments and suggestions. All errors are my own.

1“Triple-A-Failure” by Roger Lowenstein, New York Times Magazine, April 27, 2008. “Bring-

ing Down Wall Street as Ratings Let Loose Subprime Scourge” by Elliot Blair Smith,

www.bloomberg.com, Sept 24, 2008. “Berating the Raters” by Paul Krugman, New York Times

Magazine, April 25, 2010. “Downgrade the Rating Agencies” by Kathleen Casey And Frank Part-

noy, New York Times Magazine, June 04, 2010. “Rating the Raters: Enron and the Credit Rating

Agencies,” Congressional Hearing, March 20, 2002. “Credit Rating Agencies and the Financial Cri-

sis,” Congressional Hearing, October 22, 2008. “Reforming Credit Rating Agencies,” Congressional

Hearing, September 30, 2009.

DOI: 10.1111/jofi.12070

2435

2436 The Journal of Finance R

While much of the focus is on the ratings of mortgage-backed securities, rel-

atively less attention has been paid to corporate bond ratings. In the United

States, the corporate bond market comprises 22% of bonds outstanding.2Raters

have had a much longer presence and command substantial influence in mar-

kets for corporate bonds. It is interesting to ask whether the systematic pat-

terns of slackening ratings in the mortgage-backed securities markets are also

present in the corporate bond markets. A finding of this nature would be sup-

portive of systematic misconduct by agencies and would lend less credence to

the argument that the lax ratings in the mortgage-backed securities markets

were perhaps attributable to the unfamiliar nature of mortgage securities.

In this paper, I study whether the rating agencies tightened or loosened

their corporate rating standards during the 1985 to 2007 period. Taking the

year 1985 as a reference point for standards, I ask whether a firm receives a

higher or lower rating in later periods while displaying the same risk character-

istics. Higher (lower) ratings imply loosening (tightening) in ratings. I report

two main patterns in standards. First, I find that between 1985 and 2002

there is a divergent pattern between investment-grade and speculative-grade

rating standards. Specifically, the rating agencies tighten the standards for

investment-grade ratings, while loosening the standards for speculative-grade

ratings. Second, I find that around 2002 there is a shift toward more stringent

ratings in both investment-grade and speculative-grade rating categories.

My first major finding is the puzzling divergent pattern between investment-

grade and speculative-grade standards during the 1985 to 2002 period. As

in Blume, Lim, and Mackinlay (1998)—henceforth BLM—I find that the

investment-grade ratings tighten during this period. Holding firm character-

istics constant, the tightening amounts to an average of 1.1 notches. How-

ever, in contrast to investment-grade ratings, speculative-grade ratings loosen

0.6 notches during the same period. The loosening is consistent with the

widespread criticism of agencies during the Dot-com crash. Before 2002, the

tightening trend in standards shown by BLM pertains only to the investment-

grade category, where default rates are rare. However, the rating agencies did

indeed get looser in their speculative-grade rating assignments, the region in

which the majority of defaults occur.

My findings raise the question of why rating agencies loosened the standards

for speculative-grade bonds while tightening those for investment-grade bonds

prior to 2002. I offer a partial answer to this question. I show that, over the 1985

to 2002 period, most of the growth in the universe of rated firms comes from

speculative-grade, first-time issuers. Rating agencies appear to assign more

issuer-friendly ratings to this asset class, in which there is substantial growth

for first-time entrants. The issuer-pays rating model can induce agencies to

assign more lenient ratings to attract further business. Indeed, such behavior

mirrors the failure of the rating agencies in structured financial products dur-

ing the 2008 financial crisis. Similar to the expansion in speculative credits

2As of Q2 2011, the Securities Industry and Financial Markets Association (SIFMA) (http://

www.sifma.org/research/statistics.aspx).

Structural Shifts in Credit Rating Standards 2437

prior to 2002, the mortgage-backed securities of subprime loans were newly

rated instruments and experienced substantial growth during the post-2002

period. My evidence suggests that this phenomenon might have occurred in

the corporate bond market as well.

My analysis also shows that there is a structural break toward “more

stringent” standards in both investment-grade and speculative-grade ratings

around 2002. The structural break is both statistically and economically sig-

nificant. From 2002 to 2007, the investment-grade ratings tighten 1.3 notches,

and the rate at which these ratings tighten is faster than in the 1985 to 2002

period. The speculative-grade standards also tighten by an average of 1 notch

after 2002. For the entire sample, the post-2002 drop in ratings amounts to

1.5 notches. The timing of the 2002 break coincides with the increased regula-

tory scrutiny and investor criticism that began with the high-profile accounting

scandals. These scandals and the subsequent passage of the Sarbanes-Oxley

Act (SOX) accelerated discussions regarding the quality of ratings and the role

of rating agencies in capital markets. It appears that widespread criticism, the

regulatory threat to open up the rating industry to competition, and the repu-

tational concerns of the rating agencies led them to make structural changes

to their rating standards.

The shift toward more stringent ratings after 2002 is parallel to the behavior

of rating agencies after the 2008 financial crisis. In 2009, Standard & Poor’s

(S&P) announced “big changes” to collateralized debt obligations (CDO) and

residential mortgage-backed securities (RMBS) rating criteria that would lead

to stricter rating assignments (Standard and Poor’s (2009)). More recently, fol-

lowing the sovereign debt crisis, S&P announced more major changes in rating

criteria that included government ratings (Standard and Poor’s (2011b)). Stan-

dard & Poor’s states that many of its “recent criteria changes were prompted by

experience gained from the 2007 to 2009 financial crisis.” My evidence from the

corporate bond market indicates that agencies take significant steps to correct

their ratings only after high profile failures to restore confidence in ratings.

Next, I analyze whether the changes in rating standards can be explained by

changes in the economic climate. It is possible that today’s economic environ-

ment is riskier and firms need to maintain better characteristics to achieve the

same creditworthiness compared to earlier years. If so, then tighter standards

are justified. However, if bond markets perceive the changes in standards to be

unjustified, then credit spreads should reflect variation in the standards. Ce-

teris paribus, the average spreads for a fixed rating category should decrease

(increase) when ratings tighten (loosen). I find evidence to support this predic-

tion. In time-series tests, I show that bond yield indices are positively related to

the measure of rating standards, that is, looser standards have an association

with higher spreads. These patterns can be detected in firm-level tests as well.

As in Baghai, Servaes, and Tamayo (2011), I find that the difference between

the actual rating and the predicted rating helps explain credit spreads in the

cross-section. Bond markets appear to offset roughly one-third of the effect of

tightening when pricing debt. These findings suggest that the bond markets do

not view the changes in standards as completely justified.