Bongaerts Tiebreaker

一决胜负！
Tiebreaker: Certification and Multiple Credit Ratings

Author(s): DION BONGAERTS, K. J. MARTIJN CREMERS and WILLIAM N. GOETZMANN
Source: The Journal of Finance , FEBRUARY 2012, Vol. 67, No. 1 (FEBRUARY 2012), pp.
113-152
Published by: Wiley for the American Finance Association
Stable URL: https://www.jstor.org/stable/41419673
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
and Wiley are collaborating with JSTOR to digitize, preserve and extend access to The Journal
of Finance
This content downloaded from

202.115.120.41 on Wed, 07ff on Thu, 01 Jan 1976 12:34:56 UTC
All use subject to https://about.jstor.org/terms
THE JOURNAL OF FINANCE • VOL. LXVII, NO. 1 • FEBRUARY 2012
Tiebreaker: Certification and Multiple Credit

Ratings
DION BONGAERTS, K. J. MARTIJN CREMERS, and WILLIAM N. GOETZMANN*
结果仅仅支持regulatory ABSTRACT
certification
This paper explores the economic role credit rating agencies play in the corpor
bond market. We consider three existing theories about multiple ratings: informat
production, rating shopping, and regulatory certification. Using differences in rat
composition, default prediction, and credit spread changes, our evidence only suppo
regulatory certification. Marginal, additional credit ratings are more likely to oc
because of, and seem to matter primarily for, regulatory purposes. They do not se
to provide significant additional information related to credit quality.
Credit rating agencies (CRAs) report information about the credit risk o
income securities. The various ways the information is used by financi
and regulatory entities may potentially influence the nature of the in
production process. Bond ratings are used not only for risk assessm
also for regulatory certification, that is, for classification of securitie
vestment grade (IG) and high yield (HY, or junk) status. These classifica
turn influence institutional demand and serve as bright-line triggers i
rate credit arrangements and regulatory oversight. Regulations may
that insurance companies and banks keep much higher reserve capit
issues than for IG corporate bonds. Other institutions such as pens
and mutual funds are often restricted by their charters with respe
amount of HY debt they can hold. Taken together, more than half of
rate bonds are held by institutions that are subject to rating-based res
on their holdings of risky credit assets (Campbell and Taksler (2003
*Bongaerts is with Finance Group, RSM Erasmus University Rotterdam, and Cr

Goetzmann are with the International Center for Finance, Yale University. We would lik
Patrick Behr; Michael Brennan; Mark Carlson; Erwin Charlier; Long Chen; Frank de J
Driessen; Frank Fabozzi; Rik Frehen; Gary Gorton; Jean Heiwege; Mark Huson; Ron Jo
Klaassen; David Lesmond; Hamid Mehran; Catherine Nolan; Frank Packer; Ludovic
Paolo Porchia; Jörg Rocholl; Joao Santos; Joel Shapiro; Chester Spatt; Walter Stor
Swanson; Anjan Thakor; Laura Veldkamp; Evert de Vries; Jacqueline Yen; Weina Zhan
as conference participants at the Financial Crisis conference at Pompeu Fabra Unive
European Finance Association annual meetings in Bergen (Norway, 2010), the Te
Festival at UT Austin, the RMI conference at National University of Singapore, the NBE
on Credit Ratings in Cambridge, the Conference on Credit Rating Agencies at Humboldt
the American Finance Association annual meetings in Denver (2011); and seminar pa
the University of Amsterdam, Rotterdam School of Management, and the Dutch Na
for helpful comments and information. We especially thank Campbell Harvey (the
anonymous associate editor, and an anonymous referee for many helpful comments a
113

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
114 The Journal of Finance®
demand for HY bonds can significantly increase the cost of

现有体系是否有
issuers and thus affects capital structure decisions 效激励了发行 (see
Kisgen and Strahan (2009), Ellul, Jotikasthira, 人，让发行人能 and Lund
stitutional and regulatory importance of credit ratings 够有动机无保留 to
地披露有关公司
has therefore raised questions about whether the 价值的信息是否 current s
proper incentives for issuers to fully disclose value-relevan 能够让投资者认
for investors to invest in research about credit risk. 识到信用风险从
而开展投资活动
Using a sample from 2000 to 2008, we document that almost all large, liquid
U.S. corporate bond issues are rated by both S&P and Moody's. Fitch typically
plays the role of a "third opinion" for large bond issues.1 During the sample
period, the most prevalent institutional rule for classifying rated bonds was as
美国认证规则： follows: for issues with two ratings, only the lower rating is used to classify the
二选低，三选中
issue (e.g., into IG or HY); for issues with three ratings, the middle rating should
be used (see, e.g., the National Association of Insurance Commissioners (NAIC)
guidelines or the Basel II Accord).2 Therefore, if S&P and Moody's ratings are
on opposite sides of the HY-IG boundary, the Fitch rating (assuming it 2021那篇并没 is the
有区分2重评级
marginal, third rating) is the "tiebreaker" that decides into which class the
和3重评级在机
第三个评级加入 issue falls. Notice that this rule directly implies that adding a third rating 制上有什么区
不会使得结果更坏,cannot worsen the regulatory rating classification, but may potentially 别 lead
可能使得评级更好 --有可能中国
to a higher rating. Consistent with this idea, we find that, in about 25% of
多数只有2重评
Fitch rating additions, the addition leads to a regulatory rating improvement, 级？
that is, the resulting middle rating is higher than the lowest rating before the
也没有区分监
管要求2重和自
Fitch rating addition. Ex ante, such an improvement is likely to be particularly
important when S&P and Moody's ratings are on opposite sides of the HY-IG 愿2重
boundary, as absent an improving third rating, the split between S&P and
甚至没有谈到
评级认证的规
Moody's would result in an HY classification. Thus, the value of the Fitch
则？？？
rating is that it can push the issue up into the IG category, but it cannot pull it
down into the HY category.3 中国2重评级
都不多，样本
In this paper, we explore the nature of this tiebreaking role in the context
中只有6.3%
of the broader question of why corporate bonds generally have multiple credit
ratings. We consider three hypotheses that could lead to demand for multiple
credit ratings, namely, an "information production" hypothesis, a "rating shop-
ping" hypothesis, and a "regulatory certification" hypothesis. These hypotheses
are not mutually exclusive but they have different empirical implications that
we exploit to shed light on their relative importance.
不是互斥的，但是在实证上的表现是不同的
1 For smaller corporate bond issues, Fitch is occasionally one of two raters. However, almost all
bonds in our sample are rated by both Moody's and S&P (see also Figure 1). Specifically, about 95%
of all bonds in our database with at least two ratings are rated by both S&P and Moody's. This lack
of cross-sectional variation in having an S&P or Moody's rating means that we can only 样本中删去了没
study the
implications for having Fitch as a third rating. We remove from our sample all bond issues 有同时获得标普
that do
not have ratings from both S&P and Moody's. For this sample of bond issues rated by both和穆迪评级的债
Moody's
and S&P and using quarterly observations for 2000-2008, about 60% of observations have 券 a Fitch
rating. As a result, the main focus of our paper is to consider the "marginal" role of Fitch考察惠誉fitch的
ratings,
while controlling for S&P and Moody's ratings. Throughout the paper, we only consider the three
增量作用
major CRAs, ignoring all others, as they are much smaller at this point.
¿ NAIC is the organization of state insurance regulators.
For firms with split Moody's and S&P ratings, 13% of Fitch additions are such boundary cases.
刚好pull down to HY category了

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 115
Under the information prod

tainty, which is reduced by
barth (2010)). Under the rati
ditional rating in the hope o
认证效应只有评级结果
(2005), Sangiorgi, Sokobin,
不好的发行人，会试图
And, under the regulatory
通过增加第三个评级来 c
and Liu (1994)), market and
提升评级—逆向选择
to credibly separate bond iss
non-informationally sensit
Thakor (1993)). These corre
regulatory certification role
在HY-IG线上的发行也可
need a third rating. This lea
能存在评级购买效应
rating leads to adverse select
ses are not mutually exclus
important and thus more p
We find the strongest eviden
esis. First, we consider what
the HY-IG boundary, when F
bond issue into the IG class.
there is no change following
tween an IG and HY classific
that the certification effect
Second, Fitch rating additio
HY-IG boundary do not seem
not only for Fitch rating ad
both Moody's and S&P, but a
panel regressions of credit s
S&P, and Fitch. In contras
made by both Moody's and S
changes across the whole rat
Third, Cox proportional h
Fitch rating is positively ass
Moody's and S&P ratings, b
and finally, comparing def
for corporate bonds rated
Moody's ratings perform be
Ratings by Moody's and S&
Fitch, whereas the reverse i
providing limited additiona
in Moody's and S&P ratings
In sum, the credit spread
information production hy
provides significant informa
some evidence in support of
boundary for possible regula
additional Fitch rating tend

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
and S&P ratings, investors do not appear to incorporate the

by lowering credit spreads. This would seem to undermine
gage in rating shopping. However, we find that the addi
"extra" optimistic for issues rated just below IG or for issu
tiebreaker around the HY-IG boundary, that is, more so th
rating spectrum. These issues are exactly the issues for wh
rating shopping incentives to matter most. Specifically, if
ings are on opposite sides of the HY-IG boundary, the ad
is more likely to lead to an improvement in the regulatory
which, in this particular case, means an improvement from
cation (using "the worst of two and median of three rating
evidence is suggestive of rating shopping around the HY
the marginal rating being used for regulatory arbitrage.4
Endogeneity is a significant concern in our study, as we r
confounding variables for identification. We seek to mit
issues by focusing on credit spread changes and Fitch ratin
directly estimate selection effects using the Cox proportio
explain the time to add the third rating.
Taken together, our results suggest that a major funct
could be to avoid adverse selection for intermediate qua
(Gorton and Pennacchi (1990) and Boot and Thakor (1993
formed investors may be reluctant to trade bond issues
be at a considerable information disadvantage, that is, H
vestors specialized at producing information might find it
for medium quality issues, unless they can generate a pr
an informational advantage with uninformed investors
no-trade region for these intermediate quality bonds. An a
gives a clear signal about whether research will yield releva
whether relatively uninformed investors may be at a disad
igate the existence of such a no-trade region. The gener
ratings may also be requested as a precautionary measu
sues rated by Fitch ratings are more likely to have sub
S&P rating transitions, suggesting that these issues have
ratings.
In the long run, a necessary condition for any CRA's HY-IG classification to
be viewed as credible is to use and produce value-relevant information about
the firm. A rubber-stamp rating without research will not serve a certification
4 There is some evidence that regulators are concerned about such "ratings arbitrage." See, for
example, proposals for the new Basel II Accord made in July 2008: "If [an issue] has multiple
ratings, the applicable rating would be the lowest rating. This approach for determining the
applicable rating differs from the New Accord. In the New Accord, if an exposure has two ratings,
a banking organization would apply the lower rating to the exposure to determine the risk weight.
If an exposure has three or more ratings, the banking organization would use the second lowest
rating to risk weight the exposure. The agencies believe that the proposed approach, which is
designed to mitigate the potential for ratings arbitrage, more reliably promotes safe and sound
banking practices." Source: //www.occ.treas.gov/fr/fedregister/73fr43982.pdf.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 117
role in the long run. If th

or if regulations with resp
the aftermath of the recen
this paper may become less
fewer firms may opt for
vince the market that its r
quirements but also provide
ticularly about separating i
securities). Indeed, less re
competition among CRAs t
around the HY-IG boundar
The remainder of the pap
motivation for our empiric
of the existing literature
methodology. Section IV pr
Section V concludes.
I. Motivation
A. Credit Rating Agencies and Regulation

There are currently three major CRAs in the U.S. market: S&P, Moody's,
Fitch. In addition to these big three, seven smaller CRAs issue credit r
that qualify for meeting regulatory standards. While the purpose of a
rating is to reflect the creditworthiness of an issue or issuer, the rating ag
have some discretion in the philosophy underlying their rating system an
not required to make their rating methodology public.5
CRAs are licensed as Nationally Recognized Statistical Rating Organiz
(NRSRO) by the Securities and Exchange Commission (SEC). This offici
ignation has a number of effects. First, CRAs are exempt from Regu
FD, allowing corporations to share value-relevant information with the ra
agency without disclosing it publicly. Second, the SEC designation allows c
ratings to be used to meet regulatory requirements that call for a minim
an average rating value. For example, the SEC requires that money m
mutual funds hold instruments with a credit rating in one of the two sho
term higher credit rating categories.6 This effectively provides a "safe ha
for money market mutual funds with respect to litigation over fund fail
Kisgen (2006) discusses the strong link between short- and long-term debt
ings and access to the commercial paper market. He concludes that, in
to have access to the commercial paper market, a rating of BBB is ty
required.
5 Indeed, some ratings have a point-in-time perspective, whereas others (including the three
major CRAs) employ a through-the-cycle perspective. Similarly, while some rating agencies aim
to reflect cross-sectional variation in default probabilities (like S&P and Fitch), others aim to also
incorporate loss given default and reflect dispersion in expected loss (like Moody's).
5 This rule is likely to be revised in the future.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
U.S. insurance companies explicitly rely on NRSRO rating

risk-based capital. In particular, bonds held by insurance
signed capital charges based upon their credit ratings. Fo
life insurance company needs to hold over three times as muc
for a BB-rated bond compared to a BBB-rated bond. At th
European insurance companies will soon be subject to comp
with the implementation of "Solvency II." Banking regulat
the so-called Basel II Accords impose very similar risk-bas
ments.7 Many pension funds and mutual funds are restricted
HY corporate bonds by their charter. Although there is much
treating bank and insurance assets in the context of their tot
ing into account covariance rather than security-specific risk
large portion of U.S. institutional portfolios are still subject t
ulations tied to ratings by a relatively small number of N
of such rules and regulations on the functioning of the co
ket, and in particular in determining supply and demand,
nontrivial since a vast majority of this market is dominat
that are subject to rating-related restrictions, either through
regulations or through restrictions stated in their charters.8
In June 2008, the SEC proposed eliminating language in
pertaining to NRSROs, and instead allowing an alternative
function, perhaps recognizing that reliance on credit ratings
potential to distort the information-gathering and investmen
process. In addition, several other regulations were implemen
rating agencies more accountable and to increase the transpar
process. The motivation for these (proposed) changes stemm
prime mortgage crisis that began in 2007, and from concerns
CRAs may represent an oligopoly enabled by government
other things, critics argue that this oligopoly might not be t
anism for revealing information related to the risk of fixed
and instead might be used as an artificial safe-harbor to e
managers from exercising business judgment. As such, it coul
extract rents from corporations by virtue of serving as "gate
IG rating, especially as the CRAs are paid by the corporat
are rated. Moreover, competition among CRAs could lead
to the bottom," with CRAs decreasing standards to attract
This concern is raised about the structured finance market in the context of
the (subprime) mortgage crisis.
While all of the aforementioned issues are of a regulatory nature, the wider
financial industry has also grown increasingly dependent on CRAs. Financial
7 See http://www.occ.treas.gov/law/basel.htm for an overview of legal and regulatory news per-

taining to the Basel Accords from the Office of the Comptroller of the Currency (OCC).
0 Campbell and Taksler (2003) report that about one third of all corporate bonds are held by
insurance companies, about 15% by pension and retirement funds, 5% to 10% by mutual funds,
and 5% by commercial banks; thus, approximately 60% of this market is held by institutions that
qualify for ratings-based constraints.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 119
institutions center self-reg

tual funds state in their ch
income securities. Further
ten take credit ratings as
corporate credit arrangem
also driven by credit rating
termining whether a bond
indices like the Barclays C
IG Index.9 Inclusion in such
issue, since, for example, in
Several papers show that hi
(see, e.g., Chen, Lesmond, a
rate all (multiple) rating in
effects well beyond those r
B. Why Multiple Ratings M

In this subsection, we cons
would solicit and pay for
pirical evidence provided
subsection. Specifically, the
tion production" hypothesi
"regulatory certification" (
a short description of eac
these hypotheses are gener
related as well as differen
guish which hypothesis ma
are summarized in Table I.
First, a firm might apply for multiple ratings due to a need for increased
information production. More ratings can reduce uncertainty about the credit
quality of the rated bonds. In a setting in which each CRA relies on different
kinds of information to rate bonds, multiple perspectives reduce uncertainty
评级机构的信用
about default probability. CRAs may specialize in evaluating particular drivers
评级方法不同，
事前不确定of default and thus each may have some advantage that justifies its continued
传递不同信息
existence in the marketplace. Thus, one would expect issuers with greater
不确定性越大，
ex ante uncertainty to be more likely to apply for extra ratings,越有可能多重评
since the
级
potential reduction of uncertainty is largest for these issuers. Moreover, un-
der the information production hypothesis, an extra rating in agreement with
existing ratings would reduce credit quality uncertainty and thereby lower
credit spreads. 不同评级的consistency会降低信用质量不确定性，从而降低信用利差
评级选购 Second, the rating shopping effect can arise in a setting in which CRAs do not
的setting
perfectly agree or there is considerable uncertainty about credit quality, while
issuers may have better information about their own credit quality. In this
信息不对称
9 The Barclays U.S. Corporate IG Index Factsheet is available at https://ecommerce.
barcap.com/indices/index.dxml.
1.评级机构not perfectly agree 2.存在相
当大的信用质量不确定性3.发行方关于自己
的信用质量的信息更充分

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Table I
Empirical Predictions
The various empirical predictions of the three hypotheses we consider are summarized in the table
below, where " indicates that the implication is not supported and "+" means that it is supported.
这里无法支持，但是中
Reason for Multiple Information Rating Regulatory国研究又说支持？？
Ratings Production Shopping Certification
新增意见一致的评级
(i) Additional agreeing rating lowers + - -
credit spreads
(ii) Additional relatively optimistic rating + + Only at HY-IG boundary
lowers credit spreads (also away from
HY-IG boundary)
(iii) Uncertainty uniformly increases # of + + possible
ratings
(iv) Additional rating more likely if that possible possible +
could push the issue into IG
classification
(v) Additional rating more optimistic - + Only for strategic CRA
(especially around HY-IG boundary)
(vi) Additional rating associated with + - +
higher expected time variation in
1.solicit multiple bids
ratings 四种方式 2.chooses the rating equals/exceeds the real credit quality
3.publicize private ratings only if favourable
4.choose CRA based on investment banks advice
case, issuers may seek to maximize their average rating by soliciting multiple
bids or following a stopping rule that chooses the first rating agency whose
rating equals or exceeds the firm's own assessment of quality. Applying for
private ratings and making these public only if favorable, or deciding which
CRA to use based on advice from an investment bank that has knowledge
about each CRA's precise rating algorithms (gaming), would lead to similar 只有当额外评级
patterns. The rating shopping hypothesis thus predicts that issuers will apply 会提高信用评级
for an additional rating only if they think it will be an improvement. Therefore, 时，发行人才会
使用额外评级结
additional ratings are on better average. Further, if the issuer applies for an 果
additional rating and this additional rating is an improvement, credit spreads 从而会降低信用
should go down. This can be because the additional rating is actually closer to 利差
the firm's true credit quality or because it is not, but the market mistakenly
takes the new rating at face value. In the latter case, if the market is not fooled,
如果市场不会被骗，那就不存在评级选
there would be no incentive to engage in rating shopping. 购的动机
The third explanation for multiple ratings is regulatory certification. Finan- 也就不满足信息
cial regulation has traditionally relied heavily on credit ratings to determine 不对称setting
the suitability and riskiness of fixed income investments. For instance, bond
ratings are used to score the quality of bonds in the investment portfolios of
insurance companies and banks, with regulatory capital reserve requirements
determined by this score. Ratings are also important in the structured finance
market, the commercial paper market, and the overnight repo market. They
are further used to determine "haircuts" at the discount window of the cen-
tral bank and to determine whether projects qualify for government assistance
(see, e.g., the Basel Committee on Banking Supervision (2000)). They may also
be the basis for financial contracting between private parties, as the world

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 121
认证的影响、作用、重要性
witnessed in the case of AI

creased collateral in its co
the enormous potential imp
The most prominent distin
credit ratings is whether a
HY. In particular, the most
bonds stipulates that, if an
to classify the issue into IG
middle rating is used (see, e
Therefore, if S&P and Mo
boundary, the Fitch ratin
class the issue falls. This
trying to achieve an IG r
a clear discontinuity in in
demand curve, the lower de
of borrowing for those is
(2009)).12
Under the regulatory certification hypothesis, the principal value of a CRA
that systematically gives better ratings (i.e., in our data, Fitch) than the other
CRAs (i.e., Moody's and S&P) is simply that it helps satisfy the bright-line re-
quirements of financial regulation. A rating from this CRA could be requested
10 Quoting the NAIC report: "A security rated and monitored by two NRSROs is assigned the
lowest of the two ratings. A security rated by three or more NRSROs is ordered according to their
NAIC equivalents and the rating falling second lowest is selected, even if that rating is equal to
that of the first lowest." This report can be found at http://www.naic.org/documents/committees_e_
rating_agency_comdoc_naic_staff_report_use_of_ratings.doc. See also Basel Committee on Banking
Supervision (2000). If an issue has only one rating, that rating will be used. However, several
regulations prohibit institutional investors from investing in issues with only one rating.
11 There have been some time series changes in NAIC regulations, but these changes do not
significantly affect the validity of our tiebreaking assumption at any point in time, that is, that
the worst of two ratings or the medium of three ratings is used for NAIC classifications. First,
the NAIC issues its own ratings. From 1994 to 2001, the Securities Valuation Office (SVO) of the
NAIC assigned an NAIC rating to each security. Anecdotal evidence suggests that the ratings from
CRAs were critical, but that the final decision was at the NAIC analyst's discretion. In 2001, a
Provisional Exemption rule was introduced under which bonds with standard features would be
assigned an NAIC 1 or NAIC 2 rating (i.e., allowing smaller capital charges than HY) automati-
cally if at least one CRA rated it A- or higher, or if at least two CRAs rated it BBB- or higher,
without the interference of an SVO analyst. Effectively, this came down to a middle rating rule
(see http://www.naic.org/documents/svo_research_SVO_jan01cc.pdf). Second, on January 1, 2004,
the NAIC implemented a Filing Exemption rule, stating that any issue rated by one or more CRAs
would be assigned an NAIC rating based on the CRA-equivalent rating. In the case of split rat-
ings, the "second best" rating would be taken (see http://www.naic.org/documents/svo_FE_FAQ.pdf).
Third, this second best rule was changed to a "second worst" rule in 2007. However, both the second
best and the second worst rule effectively boil down to a "worst of two if only two and medium of
three ratings" rule in view of the low market share of the other CRAs besides the big three. Our
contact within the NAIC SVO argued that these guidelines were generally well followed by the
individual state regulators.
12 See also Chernenko and Sunderam (2010) on the effects of the market segmentation due to
credit ratings on bond issuance and investments.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
多重评级债券频率分布图
在IG-HY boundary上的发行方有动机寻求额外的
评级
by issuers for which the extra rating might make them qual
sification. In addition, issuers that consider themselves lik
future downgrade from IG to HY could seek an extra rating d
ary motives. This could lead to adverse selection effects, a
firms with higher credit spreads would then be more likely to
rating. Therefore, under the regulatory certification hypothes
the HY-IG boundary by Moody's and S&P should give an iss
centives to get an additional rating from Fitch. Moreover, an
may provide a hedge against the regulatory and rule-base
ble future rating downgrades, while also increasing the proba
regulatory benefits from upgrades. This effect should be mo
issuers expecting to have more volatile 发行方rating的波动较大，寻求多重评级的动机越大
ratings over time.
信息效应和监管认证效应 Gorton and Pennacchi (1990) and Boot and Thakor (1993) show that the
是相关联的
information and regulatory certification hypotheses can be inherently related
in a setting with two types of investors, in which issues with a lower credit
quality carry more uncertainty. Type I investors have a time-varying natural
demand for bonds and high research costs, and type II investors are without the
natural demand but have low research costs.13 Since type I investors are at an
不太明白 informational disadvantage relative to type II investors, they will only invest
in high credit quality securities for which the informational gain of type II
investors is small, that is, in informationally insensitive assets, to avoid losses
due to trading with informed investors (see Gorton and Pennacchi (1990)).
Typically, type II investors provide liquidity to this market to accommodate
aggregate demand shocks. On the other end of the credit quality spectrum, it
is worthwhile for type II investors to generate the information needed.14 The
region in the middle could suffer from a market breakdown if type II investors
only make money if they profit from informed trading with type I investors (as
in Boot and Thakor (1993)). 15
防范信用等级中等的债券的市场崩溃
The importance of regulatory certification could be in preventing a market
breakdown for intermediate quality bonds. In this setting, credit ratings can
restore trading by reducing the uncertainty about the value of information.
Ratings will yield information not only about credit quality, but also about the
profitability of research. If the conclusion is "no substantial information bene-
fit," then type I investors would invest and type II would not bother to research.
If the conclusion is "significant information benefit," then type I investors would
not invest and type II investors would invest to hold the security. The HY-IG
boundary is the prime candidate for the location on the credit quality spectrum
13 For simplicity, one could think about type I investors as commercial banks, insurance com-
panies, and pension funds, where the natural demand for bonds stems from the random flow of
deposits and claims, and type II investors as hedge funds and proprietary trading desks.
14 ТУре II investors do not suffer from the negative effect to utility due to uncertainty; if they
need to trade due to liquidity shocks, they trade among themselves on an equally informed basis.
For type I investors, the losses due to informed trading prevent them from investing in this
region; they realize that they are at an information disadvantage and thus do not enter this market,
while the limited gains for type II investors do not make it worthwhile for them to produce costly
information in this intermediate region.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 123
where the unconditionally e

costs for acquiring informat
effect could arise in equilibr
The regulatory certificatio
similar features. In particula
across the entire rating sca
值得讨论：
around the HY-IG boundary.
在HY-IG边界上评级购
买动机更强
potheses merits discussion.
additional ratings are, on av
if rating shopping is most i
bias of the marginal rating
certification would give no
positive at this boundary as
ically, certification predict
sides of the HY-IG boundary
the (assumed marginal) Fitch
ping hypothesis, regulatory
引入第三评级的结果 would be relatively more po
Fitch在①：L
Fitch在②：Fitch Fitch ratings 认证效应，并未隐含Fitch评级会相对更高
at other parts
Fitch在③：H
结论：认证效应下， Second, the expectation of f
引入第三评级结果
会使得结果不低于L ing shopping but increases t
shopping is more worthwhil
tively stable, as, in that case
undone or become redunda
(future expected) rating vola
tivating issuers to get an a
downgrade below IG.16 For
with adverse selection, as iss
more likely to apply for suc
Each of the three explana
tion, rating shopping, and
总结 cal predictions, though diff
are potential differences in
of agreeing ratings, (ii) cre
across the rating spectrum
the number of ratings, (iv
push an issue into the IG
rating around the HY-IG b
tional ratings and the like
summary).
Under information produ
with the prior ratings will r
16 An alternative way to hedge is

is costly since, in this region of the
upward sloping.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
while more uncertainty will make an additional rating m

therefore lead to more ratings.
Under rating shopping, more uncertainty will again le
since initial ratings will err more often. Additional rat
better, and better ratings should lower the credit spread.1
ation in ratings makes shopping less worthwhile since the
will be less stable.
Under regulatory certification, a better extra rating will only lead to a lower
spread at the boundary but unconditionally an additional rating could reflect
adverse selection (only weaker issuers take an extra rating) and consequently
lead to higher credit spreads. Higher time variation in ratings will give rise
to a rating-hedging incentive and hence increase the probability of having an
extra rating even for issues that are not at the boundary.
C. Related Research
从宏观的信息提供层面开始
As asset pricing relies fundamentally on the production and disseminati

of information, and this process is endogenously determined, the related liter
ature is vast. CRAs are only one type of research and information研究信息提供者：
provider to
很多学术论文关注
the securities markets. Much of the academic literature on the role of research 股票研究员，公司
and information providers has focused on equity analysts rather than债的信用评级关注
CRAs
的较少
rating corporate debt. Studies on the equity markets address a broad range
of questions about research providers, ranging from whether analysts' opin-
ions convey value-relevant information to whether conflicts of interest and per-
sonal, strategic considerations influence the nature of the information provided.
CRAs present a different institutional structure for analysis. While the 信用评级机构已经成
same
为了有关信贷市场监
basic principles regarding information production apply, CRAs have 管的不可或缺的一部
become
分
integral to regulation pertaining to the credit market (see also the discussion
above). 信用评级机构角色理论研究
Research on the role of CRAs is more limited. Theoretical work has asked
what role CRAs play in the equilibrium pricing process. Boot, Milbourn, and
Schmeits (2006) highlight CRAs as a valuable coordination device whereby
CRAs provide little value-relevant information at the HY-IG boundary other
对于投机级债券,提供
than regulatory certification but provide useful valuation information about
了很多有用的价值信
riskier issues. Carlson and Hale (2005) point out that, when each息 investor's
optimal strategy is dependent on the strategy followed by other investors,
the public rating provided by the rating agency can serve to coordinate in-
vestor actions. Bannier and Tyrell (2006) introduce reputation and competition
among rating agencies. Under certain conditions, these features will stimulate
investors to search for private information and thus will not only restore a
unique equilibrium, but could even lead to a more efficient one.
Each of the three potential explanations for multiple ratings finds support
in existing academic literature. On the subject of information production, a
信息效应
17 This is not necessarily true when a rating agency rates too optimistically, but if credit spreads
do not decrease, there seems to be no benefit and thus no reason for rating shopping.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 125
number of papers look评级变化会向市场传

at t
递信息,并使得债券价
ample, Kliger and Sarig
值变化 (20
to show that rating chang
公司价值不变只 the value of debt. Howeve
有债券价值变
leaves the company's value
relative to the value of eq
effect of analyst dispersio
dispersion is associated wi
研究员分歧越大，信用利差越大——现金流不确定性
probably due to cash flow u
Jewell and Livingston (19
cally across rating agencie
more positive than Moody
once they restrict their s
investigate whether ratin
dence. Covitz and Harriso
C&H2003:
face between income result
声誉效益更重要，会
使得CRA避免选购评级
future fees from customer
tion concerns dominate and
Bannier, Behr, and Güttie
investigate possible adver
issuer incentives when CR
rating shopping Inspired by the financial c
金融危机之后
cent theoretical papers put
and Veldkamp (2009) devel
increase as product compl
show that naive investors
their ratings and that, in a
ping, which in turn aggra
sangiorgi 2009
评级中的偏差取决于
theoretical model of ratin
发行人之间的异质
性，以及不同评级者
upon heterogeneity across
同意的程度。
In research most closely
also look for evidence of th
closely related 1 certification effect. They
应该不需要梳理了 understand the motivatio
bond price and yield data t
18 Adverse selection may explai

ones. That is, firms that receive
anymore, whereas firms with an
On the other hand, CRAs could c
of companies in their unsolicited
solicited ratings subsequently. T
unsolicited ratings are lower tha
adverse selection of debt issuers.
is concentrated mainly among fi
ratings for U.S. corporate bonds,
several papers report a low inci
estimates an incidence rate of ap

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
第三个评级是否系统性的更乐观呢？
of the third rating. Like our paper, they find that the
systematically more optimistic. However, they fail to find e
of a third CRA is motivated by information, rating shopp
effects. Since the time of their study, bond price data have
available for research. This allows us to conduct more p
market response to an additional rating, and to underst
how market participants interpret and use ratings.
Another closely related paper is Becker and Milbourn (
closely related 2 the impact of the major growth in market share 竞争的加剧使得评级
of Fitch
市场的质量下降了
that more "competition leads to lower quality in the rating
机构发布的评级对发
bent agencies produce more issuer-friendly and less inform
行人友好，并传递出
competition is stronger." They explain this result更少的价值信息
by apply
model of Klein and Leffler (1981), who consider CRA in
评级机构有改善信息
生产质量的动机，以
information production in order to improve their reputati
达到改善声誉的目的
tives would be weaker if future rents are lower as a result of increased com-
1.竞争越激烈，未来
petition. Second, if demand is more elastic with greater competition, rent越少，改善信息
this may
质量的动机越低
force CRAs to spend less on expensive information production or tempt them
2.竞争加剧，替代品
to be more responsive to issuer demands, potentially inducing rating 增加，需求弹性增
shopping.
Blister et al. (1994) find evidence of a "superpremium" in yields of junk
加，可能导致评级选
购
bonds due to regulation around the HY-IG boundary. Based on only S&P rating
data, they find that yields increase disproportionally from a BBB to BB rating
异常的yield增加
relative to the increase in default risk. Moreover, in a recent paper, Kisgen and
相对于违约风险的增
加
Strahan (2009) find that credit spreads change in the direction of a Dominion
一个监管认证的
外部冲击 bond rating after the accreditation of Dominion as an NRSRO. They also find
that this effect is much stronger around the HY-IG boundary, indicating the
importance of regulatory certification. Finally, Kisgen (2006, 2009) investigates
whether discrete rating boundaries influence capital structure decisions多重评级是否影
before
and after rating changes. Kisgen (2006) finds evidence of reduced debt 响资本结构？
issuance
when ratings are close to an up- or downgrade, suggesting that credit ratings
如果多重评级使得债
directly affect capital structure decisions in a way not captured by traditional
务融资成本显著降低
了，那么公司是否会
capital structure theories. Moreover, this effect is especially pronounced around
更倾向于债务融资
the HY-IG boundary. Kisgen (2009) finds that managers lower leverage呢？ after a
rating downgrade, suggesting that managers target credit ratings rather than
debt levels or leverage ratios. This effect is again more pronounced around the
HY-IG boundary.
With respect to the nature of the certification effect that we find, our research
relates to earlier work on security design. Gorton and Penacchi (1990) consider
风险资产中，只有信息
不敏感的那一部分损失
a model in which uninformed investors are incentivized to transform risky
可以通过与知情投资者
assets into information-sensitive and information-insensitive parts, 交易来避免
where for
the latter category they can avoid losses due to trading with informed investors.
Boot and Thakor (1993), on the other hand, develop a model in which发行人通过使得信息敏
security
感的资产的盈利效果增
issuers lower funding costs by making informed trading more profitable. Our
加，来降低融资成本
setup motivating the exploration of the regulatory certification hypothesis uses
key insights of both papers. In particular, the nontrading region in our setup
is a result of the absence of the uninformed investor, whereas the uninformed
investor is needed to make trading profitable for the informed investor.
nontrading region的存在是由于不知情投资者不参与交易
如果要使得知情投资者从交易中获利，不知情投资者必须存在

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 127
II. Sample Construction and Methodology

A. Main Measures and Controls
We measure uncertainty or opaqueness by analyst dispersion of the fir

earnings per share (EPS) or by the dispersion between Moody's and S&P
ratings (like the HY-IG barrier dummies, we also require stability of the
uncertainty ference over at least one quarter).19 While rating dispersion is also a mea
两个核心解释变量of regulatory relevance, analyst dispersion is not, which gives us the require
identification. We consider two measures of rating dispersion: Notches of MS
Rating Dispersion , which is the absolute value of the rating difference betw
Moody's and S&P, and S&P and Moody's Disagree , which is a dummy equa
one if their ratings are not the same.
认证核心解释变量 Our main measure of the importance of regulations pertaining to the HY-I
boundary is denoted by Fitch Could Push IG , which is a dummy variable equ
惠誉决胜局
哑变量 to one if Moody's and S&P ratings are on opposite sides of this boundary suc
标普穆迪在分别 that a Fitch rating would be decisive for the HY-IG classification. In so
在投资和投机 regressions, this measure is interacted with the outcome from Fitch.
To avoid spurious results due to omitted variables in our regressions,
correct for several issue and issuer characteristics as well as for business cyc
effects. At the issue level, we correct for callability (using a dummy), s
(offering size), and term structure effects (duration and convexity). At the iss
level, we correct for credit risk (using the inputs of the Merton (1974) mode
namely leverage and volatility), profitability (ROA), systematic risk (equ
beta), and tangibility of assets (PPE/total book assets). Tangibility of asse
important since Moody's is the only CRA that incorporates expected recovery
in their ratings. We also include R&D intensity (R&D expenditure over b
assets) as an additional control. R&D intensity can be associated with sev
pricing mechanisms in the corporate bond markets. For example, high R
industries may have higher growth opportunities and therefore lower cr
spreads. On the other hand, R&D projects tend to be riskier than norm
projects, which may increase credit risk. We control for the aggregate effec
In the credit spread changes regressions, we also include time fixed effe
as controls for business cycle effects, since market-wide default probabilitie
liquidity, and risk premia are likely to vary with the business cycle.
B. Data and Filters
For our main analysis, we use corporate bond pricing data from the TRAC
database and merge it with bond characteristic and ratings data from Merg
19 To avoid capturing timing mismatches between (multiple) rating transitions, we require

any particular ratings situation exists for at least one quarter. This will also mitigate con
about not correcting for credit watches and credit outlooks (for these variables our data ar
sparse to be useful). Time variation in ratings is hard to measure since ratings are rather pers
Therefore, we do not explicitly include time variation in ratings as a variable in our regressio
rather analyze the correlation between having a Fitch rating and the likelihood of experien
rating changes.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Fixed Income Securities Database (FISD), equity data from

data from Compustat Industrial Quarterly and analyst data fr
time series ranges from July 1, 2002 to December 31, 20
data contain all trades in TRACE-eligible bonds by membe
Association of Securities Dealers (NASD) that were dissemin
Dissemination to the public happened in phases, resultin
universe of bonds. A more comprehensive description of t
as well as the dissemination process is given in Downing, Und
(2005).
We apply several filters to our data set to remove bonds wi
that we do not want to consider and to remove seemingly err
Next, we use the FISD characteristics to match the trades to b
tics using CUSIPs.22 We only use senior unsecured notes and b
筛选
all bonds that are exchangeable, putable, convertible, pay-
a nonfixed coupon; that are subordinated, secured, or guar
zero coupon bonds. Removing callable bonds would reduce our
tially, so we leave those in, but we correct for this feature in
using a dummy variable.
To decrease the impact of remaining data errors, we ave
all trades in each bond by trading day. To reduce the effect o
tion of very liquid bonds, we make monthly observations by
each bond the last available daily average credit spread of
then construct quarterly observations by only looking at the
quarter. To avoid issues with severe illiquidity and distressed
all issues with an average (based on average Moody's, S&P)
(B3). For all bond trades in our sample, we calculate yields
The benchmark rate that is used to construct credit sprea
interpolation of the yields of the two on-the-run governmen
the corporate bond with respect to duration.
Ratings data are obtained from FISD as provided by Me
ratings data provider confirmed that, due to changes in th
procedures, the ratings data before 2000 are incomplete. T
by Figure 1, which shows the number of rated bond issue
Moody's, S&P, and Fitch as well as the proportion of all
20 The TRACE database starts in July 2002.

21 We remove all trades that include a commission, have a settlement
5 days, or are canceled. Trades that are labeled as "corrected," we correct. M
trades for which we have a negative reported yield, since these will be ma
option premia in the yield. We also identify trades with a settlement dat
the maturity date and remove those trades. Furthermore, we find several re
to be duplicates, resulting from both parties involved in a trade reporting t
out duplicate trades that have identical prices, trading time, and volume.
yield changes are extremely high or low. We remove trades with credit
1,000 bps and credit spread changes of more than 500 bps. Finally, we d
duration of less than 1 year.
22 CUSIPs of nonexchange-traded bonds do not change in case of mergers,
http://www.cusip.com).

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 129
Figure 1. FISD database coverage

covered by S&P, Moody's, and Fitc
S&P data appear to be incomplete p
sample rated by each of th

While the number of rated b
three CRAs, the sudden jum
suggests that too many bond
issues had S&P ratings, but t
the percentage of all issues
jumps to 94% in 2000, and
Likewise, there is a significa
issues rated by Fitch, from 2
for the analyses that do not
second quarter of 2000 onwa
of these coverage patterns w
2002 and is dominated by dat
issues contained in TRACE
We follow convention and
Therefore, for Fitch and S&
the numerical scores correspo
AAA (Aaa), 2 for AA+ (Aal), 3
A (A2), 7 for A- (A3), 8 for
11 for BB+ (Bal), 12 for BB
(B2), and 16 for В- (B3). How
is, those implying lower ban
23 FISD confirmed that, as of mid-

CRAs, whereas before that time the
data errors in the earlier period.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Equity market data are obtained from CRSP. We calculate

historical daily idiosyncratic volatility and betas to the CR
index based on half a year of historical trading data. An A
to filter out bid-ask bounce in daily closing prices. For an
be included, we need at least 111 return observations in
year.
Company data are obtained from Compustat Quarterly. We download data
on firm size (total book assets), debt (long- and short-term debt), profitability
(earnings), tangibility of assets (PPE), R&D spending (obtained from Compus-
tat Annual, since usually reported in the annual file only), and industry (SIC
code). From these data, we construct variables for leverage (total debt over total
book assets), tangibility of assets (PPE/total book assets), R&D spending (R&D
expenses/total book assets and a dummy for missing values), and profitability
(total earnings over total assets). We also construct an SIC division variable
that is defined as the division that the two-digit SIC belongs to. Observations
with SIC codes 9100 to 9999 (Public Administration) are excluded because of
possible implicit government guarantees.
Analyst forecast data on annual EPS are obtained from I/B/E/S. We download
summary data including number of analysts, standard deviation of forecasts,
and minimum and maximum forecasts from the unadjusted file. Following
Güntay and Hackbarth (2010), we divide forecast dispersion measured by ana-
lyst standard deviation by the share price to end up with dispersion per dollar
invested.
We construct two samples with a quarterly frequency: a credit spread sam-
ple and a rating sample. Because the rating sample does not require trade
observations, this is a more inclusive panel, especially for the less liquid bonds.
Moreover, the period for which we have reliable data is also longer. Almost
all bonds in our final sample are rated by both Moody's and S&P (see also
Figure 1). Specifically, about 95% of all bonds in our database with at least
two ratings are rated by both S&P and Moody's. This lack of cross-sectional
variation in having an S&P or Moody's rating means that we can only study
the implications for having Fitch as a third rating.
Accordingly, we remove from our sample all bond issues that do not have rat-
ings from both S&P and Moody's. Using quarterly observations for 2000 to 2008,
we find that about 68% of observations in the sample of bonds rated by both
Moody's and S&P have a Fitch rating. As a result, the main focus of our paper
is to consider the "marginal" role of Fitch ratings, while controlling for S&P and
Moody's ratings. Table II presents summary statistics for the quarterly credit
spread sample. For completeness, the Internet Appendix24 presents summary
statistics for the quarterly ratings sample. Figure 2 presents the average credit
spreads over our sample by rating category. There is very significant variation,
especially starting the second half of 2007.
24 The Internet Appendix for this article is available online in the "Supplements and Datasets"
section at http://www.afajof.org/supplements.asp.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 131
Table II
Summary Statistics for Credit Spreads Sample

The table presents summary statistics and a brief description of the sample of bond issues that
have both a Moody's and an S&P rating in the quarterly credit spreads sample for 2002 to 2008.
Variable N Mean Std. Dev. Min Max Explanation
Fitch Could Break Tie 44,366 0.032 0.18 0 1 Moody's and S&P on
opposite sides of the
HY-IG boundary
Fitch Rated 44,366 0.68 0.47 0 1 Rated by Fitch
Fitch Rating 44,366 5.04 4.17 0 16 Fitch rating
Moody's Rating 44,366 7.33 3.6 1 18 Moody's rating
S&P Rating 44,366 7.17 3.55 1 17 S&P rating
Fitch Makes IG 44,366 0.016 0.12 0 1 Fitch pulls IG
Fitch Denies IG 44,366 0.0066 0.081 0 1 Fitch denies IG
MSP Rating Dispersion 44,366 0.43 0.69 0 12 Absolute value of MSP
rating difference
Moody's Upgrade 44,366 0.027 0.16 0 1 Moody's upgrade
(common sample)
Moody's Downgrade 44,366 0.034 0.18 0 1 Moody's downgrade
(common sample)
S&P Upgrade 44,366 0.029 0.17 0 1 S&P upgrade (common
sample)
S&P Downgrade 44,366 0.038 0.19 0 1 S&P downgrade
(common sample)
Fitch Upgrade 44,366 0.017 0.13 0 1 Fitch upgrade (common
sample)
Fitch Downgrade 44,366 0.024 0.15 0 1 Fitch downgrade
(common sample)
Fitch Added, Better 44,366 0.0039 0.062 0 1 Fitch added and < MSP
Fitch Added, Equal 44,366 0.005 0.071 0 1 Fitch added and = MSP
Fitch Added, Worse 44,366 0.0003 0.017 0 1 Fitch added and > MSP
Credit Spread 44,366 172.71 141.72 0.12 999.86 Credit spread
bond control Change in Credit Spread 44,366 18.39 68.33 -478.36 498.42 Credit spread change
Redeemable 44,366 0.69 0.46 0 1 Callable
Log of Offering Amount 44,366 12.03 1.88 0 15.42 Log of offering amount
Duration 44,366 6.32 3.72 1 18.91 Duration
Idios. Vol. 44,366 0.016 0.0084 0.0012 0.11 Idiosyncratic stock
volatility
Log of Total Assets 44,366 10.34 1.55 5.34 13.65 Log of total assets
Leverage 44,366 0.34 0.17 0 5.77 Leverage
issuer ROA 44,366 0.014 0.02 -0.46 0.26 ROA
control PPE/Tbtal Assets 44,196 0.36 0.24 0 0.95 PPE/Total assets
R&D/Tbtal Assets 44,366 0.012 0.023 0 0.23 R&D/Total assets
R&D Missing 44,366 0.43 0.5 0 1 R&D missing
Analyst Dispersion 43,923 0.0033 0.013 0 1.1 Analyst dispersion
Beta 44,366 0.95 0.39 -0.25 4.21 Equity beta
Convexity 44,366 53.83 56.58 1 357.47 Convexity
Turnover 43,292 13.15 15.27 0.018 85.99 Trading volume over
offering amount,
times IK

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
rating等级越高，credit spread越小
boundary:BBB和BB对应的credit spread
差距还是蛮大的
Figure 2. Average credit spreads by rating category.
III. Empirical Results

A. Rating Differences and Rating Information
Consistent with Cantor and Packer (1997), we show that Fitch ratings are
on average significantly more optimistic than both Moody's and S&P ratings
for the same issue in the same quarter. We present the results in Table III
and Figure 3. In general, S&P is also more optimistic than Moody's but the
difference is much smaller (both for the full sample and for the Fitch-rated
sample alone).
Next, we investigate the bond market reaction to the rating updates issued.
We are particularly interested in the informational content of Fitch rating
changes compared to the informational content of Moody's or S&P rating up-
dates. To minimize issues related to selection, we limit ourselves in this test to
the sample of issues that are rated by all three CRAs. Table IV presents the
results of regressing end-of-quarter credit spread changes on dummy variables
for these up- and downgrades for each CRA. All regressions on credit spread
changes in the paper use standard errors clustered by issuer (unless stated
otherwise) and include a large number of controls with time fixed effects.
The credit spread change regressions in columns 1 to 3 of Table IV indicate
that all CRAs appear to be highly informative in single CRA specifications.
However, in the joint specification in column 7, only S&P and Moody's rating
updates seem to contain relevant information associated with credit spread
changes. For example, Moody's and S&P downgrades are related to credit
spread increases of 8 and 17 basis points, respectively. However, Fitch rat-
ing updates are not statistically significantly associated with changes in credit
spreads. In joint significance tests, we reject the hypothesis that Fitch rating
看rating信息是否被反映，
可以用rating变化作为解释变量credit spread变化作为被解释变量

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 133
Table III
Average Rating Differences

Average rating differences for issues simultaneously rated by multiple CRAs, measured in rating
notches, and split up by rating categories. Rating categories are defined by average Moody's and
S&P ratings. We follow convention and use the numerical rating scale to convert ratings. For Fitch
and S&P (with Moody's rating in parentheses), the numerical scores corresponding to the rating
notches are, respectively, 1 for AAA (Aaa), 2 for AA+ (Aal), 3 for AA (Aa2), 4 for AA- (Aa3), 5 for
A+ (Al), 6 for A (A2), 7 for A- (A3), 8 for BBB+ (Baal), 9 for BBB (Baa2), 10 for BBB- (ВааЗ),
11 for BB+ (Bal), 12 for BB (Ba2), 13 for BB- (Ba3), 14 for B+ (Bl), 15 for В (B2), and 16 for
В- (B3). Therefore, a negative number means that the first-mentioned rating agency gives on
average a better rating than the other CRA in that comparison. Quarterly data for 2000 to 2008
are used, ¿-statistics based on robust standard errors clustered by issuer are in brackets. *, **, and
*** indicate statistical significance at the 10%, 5%, and 1% level, respectively.
Fitch vs. Fitch vs. Moody's vs. S&P Moody's vs. S&P
All Bonds Moody's S&P (Fitch Rated Sample) (Full Sample)
Difference -0.538*** -0.397*** 0.154*** 0.185***

[-5.57] [-3.99] [2.79] [4.41]
N 70,624 70,747 70,738 104,435
N. Issuers 450 449 452 818
AAA to AA-
Difference -0.111 -0.118 0.0217 0.0415

[-1.53] [-0.98] [0.14] [0.47]
N 3,557 3,549 3,601 8,619
N. Issuers 37 37 37 57
A+ to A -
Difference -0.751*** -0.599*** 0.165* 0.203***

[-5.85] [-4.63] [1.80] [2.68]
N 36,461 36,539 36,532 48,905
N. Issuers 232 232 233 475
BBB+ to BBB-
Difference -0.294*** -0.223*** 0.072 0.117**

[-5.03] [-3.82] [1.11] [2.42]
N 24,177 24,193 24,186 36,841
N. Issuers 277 278 276 451
BB+ to BB-
Difference -0.492*** -0.0574 0.478*** 0.471***

[-4.76] [-0.35] [3.30] [4.71]
N 6,429 6,466 6,419 10,070
N. Issuers 165 163 165 295
B+ to B-
Difference -0.800*** -0.457** 0.369* 0.430***

[-6.80] [-2.46] [1.82] [3.20]
N 3,279 3,272 3,282 6,200
N. Issuers 98 98 99 268

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Щ rv^^ö biž 2 fe ^
Ю , - , Ю , -
P4^_§<Dq3d£c4H+*"rCřC4 ^ <NI>t>O<N05(N000ïOI>0^^00 , - , , - , , - , „ S
O Cß § hfi О й . ^ ^ ^ W 00 Н^ЮН^ЮСО^ООСООО^ЙО g
Ш'ё'Зсв-Р'й^мЙййй W СО T-i !>• '- С*4 ' т^' rH t>» ' Tji СО H oj О ^ r-i '-К
řmSjdJwMWSoo II '- ' I I H ' - 'i I "тН| Й
älalll-SSslfl řmSjdJwMWSoo II -
91« ŠSĚSÍ Ïi-З Ì
3*Eog°®S
;?<!)•£ OimSîmINmï
ФНООФН<МЮ«00050)0
fi °'Г^й tí'2 S « ш £ ФНООФН<МЮ«0005
SiSgí* £§*§££ « ш
g S si »8-g s 6 «g - " -
^йь.т-СЙ'йЬвСсОнф
«■§ ^йь.т-СЙ'йЬвСсОнф a ® fc* ® .aa""®
®
f||Ná-||*.||í ^йь.т-СЙ'йЬвСсОнф i t
яorteS-ipS&^^See J £ S N
orteS-ipS&^^See &íSo^|'^.S ř*
J £ 03 (tNNflbn
О N H
S ÔSÓSIOH N ř* 03 О N H
3 §-g*<s ¿Mi а" 'Л« 7 л
CO
< ,S тз iS чз 'йчз ^ g ¿ > g
-в < ь тз S iS g чз I f « -e ^ s j.'ü| > g
-в ®^S1tì®S[BSSS3> -e
bJD
§ 2 # S 'S ^ S » t S " 5 01 £ 5? & ïo
Л
U
ТЗД^О^ aitíi-Pü'jKáS^M
ТЗД^О^ °Оь!х>Л ю # S 'S » t S ю Oí 5? Ю Ю ïo N
ъд о § P öS? ,и^сог0д>рО Ö'S шл:5и i 1 ■- 1
0 P
Т'тзтзрнтз.З 0
tì
•M с2 cößQ-мРн а> g "b
I
Tí 9 3 Шщя.щ
Š .tž Ь í"8.»
Рн diS â5 fi 9I 3
â fi iH773^^7
V I iH
1
OJ 'ü
•S|íS
ì £ ii
>> I Îli'« ì iiJJí
3 я
'S ф i
Йй. ф "° 2"S ф i ^ 2 - ¡ со , - , * ,- ,
GO »§£ "° 2"S I « M-SgS ^ 2 1 - ¡ s 3883 со
-M
o "TÍ >- < -M '"O .r -rH in o fv hû L-l
вр8
2 1 я S & 2 ® м К ° äg §• •§
ü
а
fil * .
•pH
^ S > A; "S ^ >* S^bOlV, ^ W <N Ю O CO <N
ä> °°Л"ьпщ°^^Й^Нщ§ > A; >* W Ю (N <N Ю
§ 8 s«í-rt
s«í-rt tí ."Йiff!*-»-?! tí,tí."Й
CD S п аз b tí
ív^ řrt CÖ ď S b c3 > ^ CD S п s аз 8-el! tí ,tí ^ U"
о lo^-S řrt S"'™.2|5.S^_r|
g I .S -S § S £ S -e £ » I g ï «
^M-S^^'O.S'ßüPwiä ri woiiHo
tí ^ tí tí S. N tí O TfrHTjî^
î2llrafl|2
I <a ã S á tís? ^« tíя tí
s ^S.^ Nл -e
3 ^ ^ 5S^ л tí O
•f -S « 'S •§>$ fei o-S ** ®
íg'003?ftrKr¡^tí0®'0 ф
|ô>-§^^â3Si=8ã|g íg'
и&353--б-§-з®я|» д
S ■§ il ЦЁ «б д.|2к -§ д 1
S 4 m -3 2 3 в ° &•& s •- a -S ? ® 'S «
t?| S 4 m -3 I 2 § I? 3 в S 1^1 &•& s S •- a g 1 -S I ? • 1 ® -S 1 'S -S* «
3a,SS!|-«?s¿^l«'tf
c.g g д s 8 -1 Q °I £
8 -15 ii -S-|° -S 111
o S^Qs ¡a
•I 1I 11 S -S 11 -S*
o ¡ft s ¿ ^
.tï||ï-î!l?!i,6l c.g g д s -1 5 -| £ ì f ^ £ ¡a S o < ¡ft ¿ a ? ^
áellS-SJIĚSlSé I I % S S ê £
202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 135
*
, - , * , - , £н , - , см co ^ ^
^ юаэсоаосон_нООсо©о i>
w X CD^C0^(MO^<Nin0000 О
w ¿riniNON^^ÒWÒ CO
<N 1 - 1 i-I 1 - 1 I rH <N CO
*
*
Oi i - i (N ^ ^
^ Ю <N ,-н 00 CO © ©
t> l> Г] (N ю 00 О
w ò и ^ Tř о 00 ю
I rH <Ñ i- I
不好于BBB+
*
^ 1>Ю1>ЮЮ<мРЧЮ©сО© t>
CO COTi<T*|><Nt>PQO*COH© 00
w ^hoíhooWCÍOÔ Ю
cs, - « ' ! r- i д ! i- ,
*
*
不差于A-
*
* O i - i . о
^ О i <N i . I 00 Ю t> О
LQ ^ O n ^ H CO
w О Ю VI H о H co
I rH rH Ю
*
* , - , * , - , 00 , - I CN СО ^ ^
^ «С0Н0)Ю00м00«оО (N
a
wн
w т*riioi>oqco(N^^<Nißoooq
m (N о ri ^ Tf о 00 o^ <N
oq
СО 1-1 rH 1-1 I 1-1 сч
•S
e
ï ^ Ю 00 , - I G0 СО o LO
co (N о J " N Ю 00 н
w Ò H " O n О
rH
I 1- I (N
Ž
00 I - I CN C0
^ ю H M 00 СО о
(M N N Г) N Ю 00
W ¿ ri ^ ^ ò W
i rH <N
±
*
*
00 r-i rH
^ Ю Ь rH ® СО о
rH N о ri M Ю 00
w о fi ^ Ö «
I rH <N
'oí
J2
'л
>
3
CD
Jh
PQ
I V fi- E
V £ДЪ" fi-
cd Ё
E
" <V) . о
^4 ^ Й . ""Ö
® ed ft £ fa
тз а> ся о II
Й £ +3 "ö Ф
й) ® ^ ^ ^
Î Ч 2 » & &
5 О „s S£ (D¿
Q
Q 1-1
1-1 О ^ "О
^ О "О „s
m<U (NřU
bD^ řU йй£
£ ^
33
i•1-1гЧ
• гЧ«J-H
«J-H .з ^•• pH
pHи mCd
Cd pH
pH bD^cd
i cd '5-M
vJ' řU. vJ* . 1£ *Гч
1 ^ ?irГч11
• fa гЧ «J-H £ fa • pH
202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
深2
深1
深3
Figure 3. Rating differences across CRAs. The figure plots rating diff
pairs of CRAs. We use a numerical scale for ratings, where a lower rating
rating (more optimistic or lower bankruptcy likelihood, see Table III for t
the figure, a negative number thus means that the first-mentioned rating ag
a better rating than the other CRA in负数：平均而言，前者给出的rating比后者给出的更高
that comparison.
we reject the hypothesis that Fitch rating
downgrade coefficients are equal to S&P or Moody's rating
ficients, though not for the equivalent rating upgrade coeffic
restrict ourselves to the upper end of the rating spectrum (se
we only use issues with an average rating of A- or better)
contain no information even in the single CRA specification. W
the hypothesis that the reactions to Fitch upgrades and do
tistically different from each other in the presence of up- and
the other CRAs, while we can for Moody's and S&P (see the In
for these tests on Moody's and S&P).
However, rating changes of Fitch at the HY-IG boundary do
when Moody's and S&P ratings are on opposite sides of the
and Fitch could be the tiebreaker and change the classifica
issue into IG versus HY. Economically, the credit spread c
with Fitch changing the classification to IG rather than HY
points in the full sample (column 4, p-value of 2.82%), abo
in a sample of bonds rated BBB+ or worse (column 6, p-va
again about 41 basis points in the full sample controlling for
rating updates (column 8, p-value of 6.07%). These results a
a regulatory certification effect and inconsistent with an infor
Table V presents regressions of price reactions to Fitch ad
bond has been in our sample for at least one quarter without

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 137
^
йл >íиьр
."Й ö tí2 й'S^ œ
§ фNíU I i'
I i' Gì , -l>
00, OÍ
- 00
- H -ooСО
- Ь-, ,ЮЮ,- ,-
<N,, g
P GO IO ю CO Oí !> СО S
w çooi-ноюо Tt©£
çooi-но
4 У ^ h ш <fH ^ M. ® + 4~
h Sí I ш <fH s ю й M. 2 ® m +
й s tí б-« «Й
^•Sce^bß.sS^gjH #
-£ ^ .й ft ^ Ì3 .H £ ÔÔiCDO О * # ^
SrtSS^'CüröSS .H ^ NOHIOO^ н <Л (N
w S l>i-i(NOCOO - - ai о о со
Sflêïs?,«!! w S l>i-i(NOCOO il " - - s 0> -
■'S^DU-eSSS10
S^'S? t. "«.!
0> 1 1
- Г-
^ c5 ^
S S
S >
ž ž
> <8
ъ ^®®ш ^
к- а
а 0
д
Moody's and S&Prating, it does not lead to a significantly
'SJh^'öS S ^ g"§
|Ë1 « «3-i^l »
When a Fitch rating is added that confirms the average
Gß
tí (D-м ^р-н^^Д^.^ ^ IOCDOO(NOOOOI>CD 1> <N 00
核心回归：三个机制
0 Ю (NCDOOOOi^NOOlOWN
5 âÍM й 14 S « w Ю юоооюоом ~ о о ю <n
•M 'Л ~ -eoj^-гн-
■d
« OO gQ.^ *^g22 £
-s £-Öí s И)
И) 1-1
1-1s »5títí-H
-H
3 ~ Q. g. -Й 3 .g -Ö s ™ ^ s 2
Ы)
tí
•fN
j-Sb-sl^ê! g. -Й 3 .g s ^ s 2
c2 hoíIljSí«1^-
c2 -rt « ^ ^« fe
^ ^ fe
,хГg ^g^CÛCOCD^HOOOCO^COtD Tf„oo„t>„î„ „„ ï „„
^ ^CÛCOCD^HOOOCO^COtD
I л-н^
ÍS^.íí
tí С* ЮНМ^С0
л-н^
00 0 05
o5w¿fi
0500)Ю
tí С* - o
Л g^l-SïïiJsS! o5w¿fi - 1
ü
-м
• PN
p.ëJë'l^ïïS-fS
®л<ер,8шРф-ае
fe
'S•■s
-Й « ^"S
¡g 2 «
3-я2í ф¡g
-S £g
^ «íо. fc -S •§ «
>«S О
О § *ё Я m a 'S ^ •* S
з %
S tí ',«
g +32 2й'S^ ¡2ГК
ГКfH
fH sg SЙ ^S гТЗ' ^г 'ЮЮООСООО^!>^ Ю (N (N
^ СО <М1>©ОО0>^Ю00С000<М
н S tí
bß^ - тз^! Кл м 50 öW w LodòoidòdcqdoLOco ~
lower credit spread

•M
СЙ S« сЛ » CS'S Кл "S II ~ "« i н-и-
Ю
1 llilÂTzl
И-^ШЦ l
SsSsSq?
л
Ф
âiie'Sgj-si'S'S a »»«et®
-tí 2 tí -w <И*^^ ^ ЮОООт^О
§S SS'S'-S^ÏÏ'â'S^» -w ' ± 1 ±
h
СО
A § р s 'S ^ g я ■« .s
ïïl р sil ^ 9 £.-.■§ g ■« IS
-M
^ i2.-Srï-e.g*-^ S
t! *15 »X о Stf1
2 Oí ■« h^á î S s S ® E
и
tírT.S'SSjtí^'tíli,^ ^ (NO
®£tí
ф 'S H
£tí
л ^H . ^1 . _g
§ $ О -3
- ^ ^ I - I
£¡£«*J§-S™3£ ^ ^ $ _g
О H hû £H _2 .in ho tí o -M
a Fitch addition is not relatedto
any change in credit spreads at
'Т!
o fe § §
О со tí ^ ^ T-I _Q Л СО Д Д -rH .ÍH
й.5 2'и^О <Й F-H a» «2 CO ÍH 2 _, Íh 2
Sr0tíô3tí>-Q_títí o F-H d a» M s а й й ^ g _, <я «
S"S Sr0tíô3tí>-Q_títí а-gS »à « % o & d I M 3 s 3 а 1 й й &.§
£ » I S 'S rt -ё '§ -s Ъ m H ^ s s I «afe HQ
"ОO)43<D
H ť M », ü ® í О Ф O) TÍOí
ТЗФTÍ'_TÍCD
PQM
TÍ>^-H
M йй^ r? tí
ф tí
2 2títí
âï'§ ф ' '»,
ф b'ï^ O Ô> ^"tí
T3 T3
^ "Ö "OТЗ
T5 T3
нн ^T3 T3
Л "Ö CÖo '_ T3^Q
^ s ^Нз-о ^ л ^ л л л с л
.■§ .■§ .■§ .■§ .■§ .■
öco СГСДтЗД^Д2 &н fe fe fe fe fe <3 fe < fe
交互项
202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
*
l> „ 00 „ °0 тЧ
^ tH<NIOO^h о 00 00
t> ÔfNCD^ylOlOCD
1-1 I г- 1 CO
*
*
Ю F- > тН Ю
^ CD CD гн л N И Oi
CD "N Oi >7 ® W 0
w Ö "tf Î ^tH
>7CO
^ Ö 50
±
*
BBS S
lõõâsl^gg Ni>^á".o®n
Ni>^á".o®n
o IO £ о а>
I
* PQ
t>
^ 1> 00 о о £ <М СО 00
^ w
w Tt I(Nt^-tJjgO^CDcoW
О (M* ^ ЮЛ ò ^ ю ^ о
I V PQ ^
± PQ
PQ
*
*
00 I - I 00 (N ^
1 w
^ ЮНрн O^wooco
^ СО (МООГЗ^ЮЮСОСЧ
■S
I гН СО
e
±
t *
*
*
1> Г-, 00 iH
^ ю О РН о ^ СО оо
I ^ (N (M Tt ^ о Ю ю CD
w © т* ^ ^ © Ф
I IH co
L„ 00.
^ LO OÍ i- i о ^ 05 00
^ iH (N CO i- rj i ¿ о ю ю CD
w o Tji ^ ^ Tf о ^
1 i-H со
/ô>
¿3
13
>
3
Ě
Tf
'S
43 fa
2 II
j? a d5 О
Ì j? S a ! 5
I I ! „I
'Ü 0) О ü for03
S 2 8n ^ o) 041 s
Iti! S 2 8n ^ Il o) 041 5^1 s
202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
cox regression
https://baike.baidu.com/item/COX%E5%9B%9E%E5%BD%92%E6%A8%A1%E5%9E%8B/8894307?fr=aladdin
Tiebreaker 139
with both Moody's and S&P

by both Moody's and S&P, an
rating as in the sample use
directly by modeling the a
hazard model regressions. I
expect the event of a Fitch
in credit spreads. For examp
prospects to request a (gener
a Fitch rating addition to b
However, column 1 in Table V indicates that a Fitch addition is not related
to any change in credit spreads at all (coefficient of -2.28 basis points with a
¿-statistic of -0.65). This lack of an effect mitigates selection issues, although
we do find a mild adverse selection effect in a robustness test in the Internet
Appendix, where we restrict the sample to the precrisis period.
Table V also fails to show any evidence in favor of an information production
effect. When a Fitch rating is added that confirms the average Moody's and S&P
rating, it does not lead to a significantly lower credit spread (see columns 2 to 7).
交互项 The interaction of the added Fitch rating with measures of uncertainty also fails
to show a significant effect (see the interaction with Analyst Dispersion in col-
umn 6 and the interaction with Notches of MSP Rating Dispersion in column better_neg:说
7).
明better使得
The negative (positive) sign of an added Fitch rating that is better (worse) than
spread降低了
the average Moody's and S&P rating is consistent with rating shopping and
worse_pos:说明
information production, but is not statistically significant. Likewise, worse使得
the co-
spread提高了
efficients on Fitch Added, Better and Fitch Added , Worse are not statistically
different from each other.
However, columns 3 to 5 provide strong evidence in favor of the regulatory
certification hypothesis. In the cases for which Moody's and S&P ratings are
on opposite sides of the HY-IG boundary, an added Fitch rating that makes
the issue qualify for an IG rating is associated with a substantial drop in credit
spread. The difference between Fitch classifying such bond issues as IG rather
than HY is associated with a difference of about 41 basis points (p-value of
3.23%) in the credit spread. This result is robust to using either the full sample
or only those issues with average Moody's and S&P ratings between BBB+
and BB- (columns 3 and 4, respectively) as well as to double clustering credit
spread changes in both issuer and time dimensions.25
25 As suggested by Cantor and Packer (1997), ratings by Fitch could be inflated. To address this
issue, we repeat the analysis of Table V, where we correct all Fitch ratings by one notch (except for
the tiebreaking at the boundary). The results can be found in the Internet Appendix. These results
are consistent with the results reported in Table V. Furthermore, we show similar results in levels
in the Internet Appendix, exploiting also observations that already had a Fitch rating when they
entered the sample. If anything, the results are even stronger since an agreeing Fitch rating is
associated with a higher credit spread. The effect of certification is statistically and economically
very similar. Finally, one might be concerned that the large movements in credit spreads at the
onset of the crisis might drive some of our results. Results in the Internet Appendix for the sample
ending in June 2007 indicate that this is not the case.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
further evidences
Table VI
likelihood to get Fitch rating
Cox Regressions for Time to Adding Fitch Rating
Cox proportional hazard model regressions of the time to adding a Fitch rating on rating category
dummies based on average Moody's and S&P (MSP) ratings, measures of uncertainty Analyst
Dispersion (standard deviation of analyst earning forecasts normalized per dollar share value) and
MSP Rating Dispersion (absolute value of the notches difference between Moody's and S&P), and
whether the Fitch rating "could push" the issue to IG or A-. F Could Push IG and F Could Push
A- are dummies indicating whether the Moody's and S&P ratings are on opposite sides of the
IG or A- boundary, respectively. Other controls that are included but not shown are firm beta,
leverage, PPE/assets, R&D expenses/assets, ROA, log of offering amount, time to maturity, time to
maturity squared and redeemable (see Table IV for descriptions). Quarterly data for 2000 to 2008
are used. The sample consists of all issues with both Moody's and S&P ratings that on average are
rated B- or better. Coefficients on the covariates in the partial hazard function are reported, and
¿-statistics are in brackets (using robust standard errors clustered by issuer). *, **, and *** indicate
statistical significance at the 10%, 5%, and 1% level, respectively. Pseudo R2 refers to McFadden
(1973) pseudo R2.
(1) (2) (3) (4) (5)
MSP A+ to A- Rating 1.553*** 1.548*** 1.568*** 1.524***

[3.04] [3.06] [3.08] [3.06]
MSP BBB+ to BBB- Rating 1.146** 1.156** 1.119** 1.071**
[2.28] [2.32] [2.22] [2.19]
MSP BB+ to BB- Rating 1.346** 1.481*** 1.408*** 1.292**
[2.53] [2.79] [2.60] [2.49]
MSP B+ to B- Rating 1.131** 1.114** 1.159** 1.073**
[2.06] [2.06] [2.12] [2.00]
Fitch Could Push IG 0.717*** 0.693*** 0.702**
[2.83] [2.83] [2.44]
Fitch Could Push A 0.248 0.221 0.231
[1.23] [1.13] [1.00]
Avg MSP BB+ 0.292
[1.27]
Avg MSP BBB- 0.235
higher uncertainty lower likelihood [1.41]
MSP Rating Dispersion -0.128
[-0.72]
S&P and Moody's Disagree -0.307 -0.267 -0.254 -0.251
[-1.16] [-0.90] [-1.03] [-1.02]
Analyst Dispersion -4.988 -4.262 -6.275 -6.479 -5.979
[-0.58] [-0.53] [-0.69] [-0.70] [-0.64]
Idiosyncratic Volatility -26.11*** -23.85*** -25.62*** -26.08*** -26.61***
[-3.95] [-3.86] [-3.86] [-3.93] [-3.88]
Log of Total Assets 0.296*** 0.200** 0.290*** 0.299*** 0.289***
[5.53] [2.57] [5.39] [5.67] [5.26]
N 38,351 38,351 38,351 38,351 38,351
Pseudo R2 0.029 0.019 0.029 0.029 0.028
N. Issuers 813 813 813 813 813
B. Adding a Fitch Rating

This subsection considers t
issues in our sample are r
In Table VI, we model the

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 141
hazard regressions, which

the three hypotheses. In t
getting a Fitch rating. Th
can focus on the relative ra
the baseline hazard rate an
The baseline hazard rate c
model), and therefore need
results.
In our analysis, we employ several proxies for information uncertainty:
(i) the absolute difference in the number of notches between Moody's and S&P
ratings and a dummy equal to one if S&P and Moody's ratings are different,
(ii) idiosyncratic volatility of daily stock returns, and (iii) equity analyst disper-
sion. We further include variables related to the relative importance of ratings,
such as leverage, firm size, and issue offering size. A positive coefficient on the
variables related to information uncertainty could be interpreted as evidence
for information or rating shopping effects.
We investigate the certification effect by including the Fitch Could Break Tie
dummy, which equals one if Moody's and S&P ratings are on opposite sides of
the HY-IG boundary. This approach exploits the fact that regulations typically
prescribe that, if an issue has three ratings, the median rating should be used to
determine the issue's rating, while the worst rating should be used if there are
two ratings. Therefore, if Moody's and S&P ratings are on opposite sides of the
HY-IG boundary, an additional Fitch rating would be decisive about whether
the issue becomes IG. As a robustness check, we also include a dummy vari-
able indicating whether Moody's and S&P ratings are on opposite sides of the
A- boundary. The A- boundary obviously does not have the same regulatory
importance as the HY-IG boundary, and thus its coefficient is expected to be
insignificant.
Finally, we add several other controls that influence bond prices, such as
rating group dummies based on the average Moody's and S&P rating, whether
the issue is redeemable, maturity, liquidation values (using proxies for fixed
assets and R&D expenses), time to maturity, the square of time to maturity, and
industry dummies. Standard errors are again clustered by issuer. The sample
consists of all issues that are rated by both Moody's and S&P over 2000 to 2008.
结论 Empirically, we find that the coefficients on variables related to uncertainty
(i.e., analyst dispersion, idiosyncratic equity volatility, and a dummy indicating
Moody's and S&P rating differences as well as a variable measuring the size
of the dispersion in notches) are insignificant or have the wrong (i.e., negative)
sign for an information or rating shopping effect. For example, we find that
issues with greater idiosyncratic volatility are less likely to get a Fitch rating,
even though further information production may be relatively useful for those
issues. We therefore find no support in the data for either the information or
the rating shopping effects.
On the other hand, columns 1, 2, and 5 show that, if an issue has Moody's
and S&P ratings on opposite sides of the HY-IG boundary, the (conditional)
likelihood that the issue gets a Fitch rating increases considerably. The

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
economic significance of the coefficient on Fitch Could Pu

able. For example, the coefficient of 0.717 in column 1 im
which Fitch is the tiebreaker have about twice (2.05 = exp
rate, that is, are about twice as likely to get a Fitch rating. W
result as strong evidence in favor of a certification effect: it
cases in which the marginal rating (i.e., Fitch) is decisive for
lation classification of the bond issue into IG and HY, that Fi
likely to (be asked to) give a rating.
The downside of the Cox regression is that it basically disca
tions that already have a Fitch rating, thereby ignoring som
information. We therefore try to corroborate the result that r
tion is an important explanation for having a Fitch rating by
the existence of a Fitch rating using a logistic regression
regression can be found in the Internet Appendix, which con
Could Push IG is strongly positively associated with havin
while none of the three main measures of uncertainty provid
support of the information production hypothesis. With a 12
ity of having a Fitch rating if the other two CRAs split at the
these results are also economically large. The Internet Ap
these results are robust to double clustering.
C. CRA Performance
This subsection investigates the general performance of
fault prediction. The main purpose is to corroborate our prev
in general, Fitch rating changes or Fitch rating additions
with credit spread changes unless Fitch is the tiebreaker
boundary. This finding predicts that Fitch rating difference
S&P ratings do not significantly improve default predictio
conjecture, we perform two tests. First, we run logistic r
defaults on 1-year lagged credit ratings. Second, we calcul
of the 1-year-ahead default prediction (i.e., Gini coefficien
rating performance of all three CRAs. This method is also
CRAs for self-evaluation in their annual default study; howev
periods and rated populations typically differ among CRAs, t
results are not useful for comparative purposes.
The sample that we use in this analysis is different from
we used for our other analysis. Since defaults are relative
maximize the size of our sample by incorporating as many
Therefore, we include bonds from issuers for which we h
CRSP, or I/B/E/S data as well as bonds with ratings worse th
restrict our attention to only senior unsecured U.S. bonds
before, we exclude bonds that are putable, exchangeable, co
ual, asset backed, or floating rate. We collect ratings for all t
that are rated by all three CRAs between 2000 and 2008 a
with the Moody's Default Risk Services Corporate database

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 143
default events. To avoid ov

that are likely to default at
in time with the inverse of t
Table VII reports the result
that default prediction is b
holds true in terms of the
default prediction relative t
flexible specifications with d
(i.e., AAA, AA+, AA, AA-, an
(37%, column 1), followed by
column 5). In columns 2, 4,
of the CRA ratings in notch
The resulting pseudo i?2s are
31% for Moody's, S&P, and F
specification is quite reason
In subsequent columns, we
Fitch ratings in columns 7
and Moody's and S&P ratin
find that the pseudo R2 is n
R2 in columns 1 to 6. For ex
Moody's and Fitch ratings eq
Moody's ratings by themselv
R2 of Fitch ratings by them
Fitch and Moody's ratings
in column 7. Thus, a Mood
while the reverse is not the
10 for the comparison of Fi
that, conditional on a Fitch
additional information for 1
is not the case.
Consistent with Panel A, the accuracy ratios (i.e., Gini coefficients) are high-
est for Moody's and lowest for Fitch. In Figure 4, we plot the cumulative fraction
of default over the next year against the cumulative fraction of ratings (from
worst to best); this curve is also called a Cumulative Accuracy Power (CAP)
curve. Here, a smaller area in the upper left-hand corner of the graph implies
greater prediction accuracy. The accuracy ratios are the fractions of the ar-
eas underneath the plotted lines minus the area under the 45° line multiplied
by two, such that the accuracy ratio converges to one as prediction accuracy
improves and is equal to zero if ratings are assigned in a completely random
fashion. The graph shows that the Fitch line is clearly below (and is thus worse
than) the other two over most of the rating spectrum.
More formally, we find that accuracy ratios of Moody's (77.9%) and S&P
(76.5%) outperform that of Fitch (71.8%), and that these differences are
26 That is, the weighting is done in a weighted least-squared sense.

27 For all nonlinear regressions in the paper, we report McFadden (1973) pseudo R s.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
и и и A bí Ci1
O
£OJ
°_o
^ -m
.2cdtí
'C *
tí
1-1
S 05
r^r ¡M
.
tn
^
Л O OJ £ _o 'S .2 -m c
Л "С
С>с35св§ wЖ m £
w гН ^ £ "СJ
^ онЛАЛЯоя- Ж ^
^ --- tí > ^ ? v< ^ 'J- 14 О
tí 3 0) ^ tP
tí О prt Р
п.св а>
-м Я tí Я g od ^
5 <3
JÍŽ £ ^ fY* tí * * ^
0 d Щ О ^ Й rH (N ю i>to°°o°°^ï
N § ^ о у > Й *Н (N ю юоййймсоо
S ° О) - 2 '3 ^ СО " ^ ^ О N
Д S ш л О) Н 2 ^ ^ "
Д X! й " Н О)
^3 " ^ р-н 4 - ' ^
S £ я ^ ^ 1Р ^ ír-, ^ St>cD
з 9 ®5 S¡ § ž Ž Ž S ^ ^
У .Îh w *7
qi *7
OÍ нО^
-I -I
tí O <« í ü qi í -
o CD <« T3 tí W
o гн fc < 2 tí 5 w
O
•и
гн 4iß
O 2
_jjлwO Сw
ей
Cl)Cl)
л
CD
гн CD
2r-í-i
* * * Xí
i *
*N *
l-l N ^ 05
rr'
^ |> rr'CO
CO
•и ü ^ Л1 _jj ^ S С 5 гн Яч 00 * r- t> i * ^ l-l см о о о ce со 2
Ittici ü Л1 Яч ® =9 •> Й Й Й <м « Я. 2
Ш tí^ Я ^ «в Ю
L 'H J О 'Й кП О
L^ кcd^ 'H
<И-g^ <о)JXiа XiЯ 'Й ^ кП Sf Я
■*> и ^ и +» д *
2 13 ° S 'S d * ^ CD „ S СО
■S
ТЗ2 §
»ч<§ rH
ä РЗ SРЗ
'S СО
СО§ CQ
CQ¡ dM
M ^ ¿н
ss 8 ^* ^^ 8^SĚ2 /к-»ч CD
/к-»ч „ llllsl
¿-«ч Ä Ä(^
(^ »v
СО
1 ТЗ 8 »ч rH II РЗ -J СО 2 CQ I ¿н ^ sa ^ ^ ?? /
У^'tíM^cdrg
3 CD <D *H +j» ® *H $ ® "
tí £ 2tífc
о A § ° ^ Ш ^
ü ^£ ^ faß
£ H H ***
* (MGì
N CD
ü> ^
h °ш.2h2й12
2 bL РнРн
ij bL S P
С.юо?^Р
P (M о
ЮŽas
о (N
Ž о^ те
Ž Ž£
•M
-M
á h 1 ш h 8 I § Î I '"ä ч! 2É eó á (n
О
•F*
^w_tí «3^w Я
« 2 řD ^w_tí bcS ¿5 > ,ф Ï * Ci
« ?£ öřD Ontí
VC Гл ptí ^ ¿5.2tí
tí CD СО О
00 ,>
о ,,ф
о ftЮ*
л о о tí
ft « *
о £, - , ft о л о
ü s
cd VC cd Гл ф n ptí tß ^ tí CD h -м W СО оъ со ^ ^ ^ çg
C^-tíWpQ'S^-r^
Г ~ О <D " £¿ Г 0>
" ЮЮ
ТЗ 111 -i тн О <N
1 тн
a 9)
0 W
<2 0 а> W +J о 13 р- $ ¡ ^ tcd § к. ф s
®cdtì"g§2^
+J р- ¡ tcd
со
к.
„
ф
.
s
Л Ю .2 tí &н £ ^ ^ ^ О « О 00 N 00
О
^hooi.s^bû^giS
« tí Ь ^ ft tí <2^^ <2
й íSо йо«1-
« ®i
2 " ! 5 S i = э- s
и
1 g? iti S Í л í - о о о I "g
|js'SÍ5g*
§tí>aj^5°§tí>aj^5°
i0 i0
- s
(P Ф h n M M H Рн - ^
^ л^s о
- llž^3|
^ о о о ^ о
S (P Ф Í! h ï n 2 Л M g Ф
QjT ^ ^3 5 § y 5 m ^ СО OÍ
og 2 § ü g -g § ž ž ^ m ^ со S
5 ^ чэ H ! S i 2 S ^ Si d '-r
g чэ H S £ 2 о 2
S«¡S^«»fcü
2 S ttí "2 ^ ^ ^ со ^
о ОïО ScoCO
О СО со со
^ ^(NCq со о ^^^(NCOO со
tí о d +3 S ^ œ "
tí 8 о Z 2 d Й -S 2 Â
В-|г1»юЯ - «ooiSS
- w ÍSŽŠ5-S8
ctíbii^^ôS"4
о о T- i
I ctíbii^^ôS"4 tesali
^ ^ S .2 42 тЗ to t to en
1I1|31§ М $I|«|H
S;si! TÍS-íJûřn 1!.cd
TÍS-íJûřn I cd M
М M5P
5P tí
. tí ^ptí
I^Sál^Se«ptí
^ptí Pncd^fa
ptí Pncd^fa
Sí в
>i í 2í¿ ^
S ^в S cd
® J° ^ îtí£
"štíťw •■§
tíCQai^ cd i
^SžS¿8-S8=8«
w^^ltífeS >i 2 ¿ ® b J° « cd ^ 5 2^ tí£ 2b tíť a-eb«« tíCQai^ cd -ss
•Si's i® It 1 ä á î§ s ¡s issliä g я
&2^айс!.^ S m E S S S §Eœfe;t£z
202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 145
ti Ю I> Ю Ю 1>
та О Oí H (N Oí 00
Cp r-i Ю O <N i-i o
N H N
tí
O
N
'С
O
ptí
*4
ев
<u
<Ñ
* * *
ni * * * * *
ü¡ *****
tí СО t> t> l> Oí
й 1-H ю О Ю ^ o
¡> I> СО l> o o o
o o o o o o
со
ai
•-м
2 «"йСО
00Ю
<NHCD СО 1-1
Tf 00 СО
a
•S f2 (M
M"1 M"1
rH (N cp Ю 0Ó Tji N ri H
с &
!
cd
*4
tí
8 tí
S <¡ §
<D м §
I % í
tí Ъ
ой 2
Рн >>
1- I
(Г) * * * *
s *****
tí OÍ 00 Ю <N 1> Ю
'Trt t> H СО СО ^ H
t> l> О О О О
О О О О О О
Й> ьо
.9 я
SIS
£ -s|§
'S '-Ö i .-a i
■g-s a.'síi's
О # cg О og О
S fe M S СО S
202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Figure 4. Default prediction accuracy. The figure plots the cumulative

U.S. corporate bonds over a 1-year horizon against the cumulative fractio
to best) for Moody's, S&P, and Fitch based on data from 2000 to 2008
rating agencies use for self-evaluation and that we report in Table VII
areas under the graphs. A larger area under the graph corresponds to
of graph is also known as a CAP curve.
statistically significant. There is no statistically significan

the accuracy ratios of Moody's and S&P (with standard erro
a Jackknife with re-sampling based on issuer). We conc
based on accuracy ratios and default prediction confirm th
the information production hypothesis in our data.28
D. Further Explorations of Regulatory Certification and

Arguably, rating shopping is more worthwhile around th
thus leading to more rating shopping at the boundary.
under a certification effect, even for issuers at the bound
28 As a robustness check, we also investigate the predictive power of

(i) Moody's and S&P disagree and (ii) it breaks a tie at the HY-IG bo
found in the Internet Appendix. For disagreements, the additional pre
the correct sign, but is not significant at conventional levels. Also, th
the HY-IG boundary does not yield a significant effect and has coeffic
regardless of whether Fitch pulls over or under the boundary.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 147
Table VIII
Relative Rating Levels

The table shows logit regressions of achieving a regulatory gain on rating category dummies,
Analyst Dispersion (standard deviation of analyst earning forecasts normalized per dollar share
value) as a measure of uncertainty, and whether the Fitch rating "could push" the issue to IG.
F Could Push IG is a dummy indicating whether Moody's and S&P ratings are on opposite sides
of the HY-IG boundary. Other controls that are included but not shown are firm beta, leverage,
PPE/assets, R&D expenses/assets, ROA, log of offering amount, time to maturity, time to maturity
squared, and redeemable (see Table IV for descriptions). Quarterly data for 2000 to 2008 are used.
The sample consists of all issues rated by Moody's, S&P, and Fitch that on average are rated
B- or better and for which Moody's and S&P disagree. Marginal effects are reported for the
regressions and all ¿-statistics are in brackets (using robust standard errors clustered by issuer).
*, **, and *** indicate statistical significance at the 10%, 5%, and 1% level, respectively.
(1) (2) (3)
A+ to A- 0.0691 0.0698
[0.43] [0.42]
BBB+ to BBB- -0.155 -0.155
[-0.76] [-0.75]
BB+ to BB- -0.0823 -0.016
[-0.34] [-0.08]
B+ to B- -0.0516 -0.052
[-0.19] [-0.19]
Fitch Could Push IG 0. 187*** 0. 196***
[3.73] [3.62]
Analyst Dispersion 5.461 5.366 5.192
[1.44] [1.60] [1.41]
Idiosyncratic Volatility -1.432 -2.132 -1.735
[-0.34] [-0.44] [-0.40]
Log of Tbtal Assets 0.0351 0.0476 0.0316
[0.90] [1.41] [0.80]
Other Controls Included Yes Yes Yes
Only with MSP Disagreement Yes Yes Yes
Fitch Added Yes Yes Yes
Industry FE Yes Yes Yes
N 569 569 569
Pseudo R2 0.15 0.114 0.131
N. Issuers 170 170 170
qualify for IG and th

IG status by sheer l
rating. Thus, at the
added Fitch ratings
shopping.
As a proxy for the l
consider whether th
as the difference bet
Table VIII presents r
mies indicating the
are conditioned to

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Table IX
Logistic Regressions for Having a Rating Transition

Logit regressions of having a rating transition next quarter on rating category dummies (AAA and
AA+ are merged to avoid singularities), a dummy indicating whether the issue has a Fitch rating,
measures of uncertainty as idiosyncratic volatility, beta, Analyst Dispersion (standard deviation of
analyst earning forecasts normalized per dollar share value) and MSP Rating Dispersion (absolute
value of the notches difference between Moody's and S&P). See Table IV for descriptions of bond-
and firm-level control variables. Quarterly data for 2000 to 2008 are used. The sample consists of
all issues with both Moody's and S&P ratings. Only marginal effects are reported, and ¿-statistics
are in brackets (using robust standard errors clustered by issuer). *, **, and *** indicate statistical
significance at the 10%, 5%, and 1% level, respectively.
Moody's Rating Change S&P Rating Change
Fitch Rated 0.0128*** 0.0102***

[3.68] [2.94]
Idiosyncratic Volatility 1.509*** 0.712***
[5.95] [3.17]
Beta -0.00562 0.00528
[-1.21] [0.96]
Analyst Dispersion 0.0459 0.0549
[1.27] [1.51]
MSP Rating Dispersion 0.0128*** 0.00611**
[4.55] [2.16]
MSP Rating FE Yes Yes
Year FE Yes Yes
N 104,629 104,629
Pseudo R2 0.082 0.067
N. Issuers 818 818
regulatory gain is
(additional) Fitch
thus that some rat
Moreover, the re
largest if the Fitch
and S&P ratings a
about 20% more l
more optimistic, t
tion in the event
average. As a resul
the Fitch rating co
for all else).
Next, one of the p
ing. Given the dem
want to hedge th
precautionary mot
estimate the freque
ever, due to data c
29 We thank our NBER

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
如果认证效应强，那么HY债券的流动性差，IG债券流动性好
Tiebreaker 149
Table X liquidity
Turnover Regressions
OLS regressions of quarterly bond turnover, measured as aggregated trading volume over total
value of outstanding bonds on rating category dummies, a dummy indicating whether the issue
has a Fitch rating, and controls for off-the-run vs. on-the-run effects (Age). F Makes (Denies) IG is a
dummy equal to one if Moody's and S&P ratings are on opposite sides of the HY-IG boundary and
the Fitch rating is IG (HY). All other control variables are dropped due to the use of both time and
issue fixed effects. Monthly data for July 2002 to December 2008 are used. The sample consists of
all issues with both Moody's and S&P ratings, i-statistics are in brackets (using robust standard
errors clustered by issuer). *, **, and *** indicate statistical significance at the 10%, 5%, and 1%
level, respectively.
(1) (2) (3)
MSP AAA to AA- Rating -2.889*** -2.887*** -2.891***

[-2.92] [-2.91] [-2.91]
MSP BBB+ to BBB- Rating 0.722 0.749 0.745
[0.97] [1.00] [0.94]
MSP BB+ to BB- Rating -2.434* -2.543* -2.554*
[-1.83] [-1.78] [-1.77]
MSP B+ to B- Rating -5.368*** -5.270*** -5.280***
[-3.51] [-3.40] [-3.38]
Age -4.035*** -4.033*** -4.022***
[-6.91] [-6.92] [-6.91]
F Makes IG 4.208*** 4.213***
compensate
[2.67] [2.67]
F Denies IG -5.801** -5.767** drops
[-2.18] [-2.16] dramatically
Fitch Rated 0.952
[1.40]
Time FE Yes Yes Yes
Issue FE Yes Yes Yes
N 43,292 43,292 43,292
Adj. R2 0.278 0.279 0.279
N. Issuers 739 739 739
can analyze this pre

mating the associati
change and a dummy
and volatility. The r
ing change is positiv
usual measures of v
with a quarterly tran
economically sizable
and 5.5%, respective
If the certification
sensitive and -insen
ket and an illiquid
should have low liqu

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
information-insensitive. However, if Fitch gives an HY rat

boundary will truly fall into the no-trade region and h
liquidity. Table X confirms these predictions empirically. B
HY based on their Moody's and S&P ratings have substan
than those that qualify for IG. However, if Fitch pulls them
this effect is compensated. On the other hand, if Fitch cou
IG category but instead gives an HY rating, liquidity drops
after correcting for issue and time fixed effects as well as t
(corrected for by age).
IV. Conclusion
Credit ratings play an important role in the capital markets. They are us
regulators and market participants to establish capital requirements an
legal setting, to provide safe harbor for fiduciaries. This widespread depe
upon credit ratings has the potential to influence how CRAs are used by
and how their ratings are evaluated by the market. A number of theorie
been proposed regarding how such dependency will affect the use of mul
CRAs, how the type of rating issued by CRAs depends upon their str
position, and how the market interprets the informational output of
agencies though the price formation process.
In this paper, we use bond issue credit ratings, characteristics, and
ket prices to empirically evaluate some of these proposed theories. W
three hypotheses: (i) "information production," which posits that the third
adds value-relevant information, (ii) "rating shopping" which proposes th
suers shop for a better rating conditional on receiving a disappointing on
(iii) "regulatory certification," which conjectures that a third agency play
role of tiebreaker at the boundary of being classified as IG versus HY. The
fication effect could arise naturally as an equilibrium outcome in a settin
information-sensitive and -insensitive investors and assets along the l
Gorton and Pennacchi (1990) and Boot and Thakor (1993). An extra rat
dicating the potential value to be gained from research could (partially) r
a no-trade region around the HY-IG boundary.
Our empirical work contains several results. First, we find that signific
differences exist across multiple credit ratings of the same bond issue
same point in time, with Fitch ratings on average clearly more positiv
Moody's and S&P ratings. This is consistent with Fitch playing a strategic
that reduces the threat that the other two CRAs could withhold IG ratin
extract compensation for regulatory certification, that is, Fitch being av
to push bonds into the IG classification when the other two firms may dis
Bond price data reveal how the market regards a rating by the third ag
In general, CRAs provide useful information to the market about credit r
However, we find no robust evidence that Fitch ratings provide addi
information incorporated in bond prices, relative to the information
contained in Moody's and S&P ratings. Thus, even though Fitch ratin
on average clearly better (i.e., more optimistic) than Moody's and S&P rat

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Tiebreaker 151
there appears to be little inf

market incorporates. This re
rating shopping hypotheses.
We find strong evidence th
fect. The likelihood of gettin
and S&P ratings being on op
that, in equilibrium, Fitch r
cases. We find some suggesti
if the Fitch rating is decisiv
Fitch ratings. In particular,
are on opposite sides of th
more likely than otherwise
classification (in this partic
sult provides some evidenc
or alternatively of the marg
In the cross-section of bon
strongly associated with cr
and S&P rating, for issues w
of the HY-IG boundary, a F
has credit spreads that are a
would push the issue into th
relatively many rating cha
Fitch rating, suggesting a pr
results combined with addit
bonds, are consistent with
that resolves a no-trade region in a setting with information-sensitive
and -insensitive investors and assets.
REFERENCES
Bannier, Christina E., Patrick Behr, and André Güttier, 2010, Rating opaque b
unsolicited ratings lower? Review of Finance 14, 263-294.
Bannier, Christina E., and Marcel Tyrell, 2006, Modeling the role of credit ratin
spark of virtuous circle? Working paper, Goethe-University Frankfurt.
Basel Committee on Banking Supervision, 2000, Credit ratings and complem
credit quality information, Working paper, Bank of International Settlement
Becker, Bo, and Todd T. Milbourn, 2009, Reputation and competition: Eviden
rating industry, Working paper, Harvard University.
Bolton, Patrick, Xavier Freixas, and Joel Shapiro, 2009, The credit ratings gam
Columbia University.
Boot, Arnoud W.A., Todd T. Milbourn, and Anjolein Schmeits, 2006, Credit ratin
mechanisms, Review of Financial Studies 19, 81-118.
Boot, Arnoud W.A., and Anjan V. Thakor, 1993, Security design, Journal of
1378.
Blister, Bill M., Robert E. Kennedy, and Pu Liu, 1994, The regulation effect of credit ratings on
bond interest yield: The case of junk bonds, Journal of Business Finance and Accounting 21,
511-531.
Campbell, John Y., and Glen Taksler, 2003, Volatility and corporate bond yields, Journal of Fin
58, 2321-2349.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC
Cantor, Richard M., and Franck Packer, 1997, Differences in opinion

credit rating industry, Journal of Banking and Finance 21, 1395-14
Carlson, Mark A., and Galina В. Hale, 2005, Courage to capital? A mode
agencies on sovereign debt roll-over, Working paper, Yale University.
Chen, Long, David A. Lesmond, and Jason Z. Wei, 2007, Corporate yield s
Journal of Finance 62, 119-149.
Chernenko, Sergey, and Adi Sunderam, 2010, The real consequences of
Working paper, Harvard University.
Covitz, Daniel M., and Paul Harrison, 2003, Testing conflicts of inte
cies with market anticipation: Evidence that reputation incentives d
Federal Reserve Board.
Downing, Christopher, Shane Underwood, and Yuhang Xing, 2005, Is liquidity risk priced in
corporate bond market? Working paper, Rice University.
Ellul, Andrew, Pab Jotikasthira, and Christian T. Lundblad, 2010, Regulatory pressure and
sales in the corporate bond market, Working paper, University of North Carolina, Chapel Hi
Gorton, Gary В., and George G.Pennacchi, 1990, Financial intermediaries and liquidity creati
Güntay, Levent, and Dirk Hackbarth, 2010, Corporate bond credit spreads and forecast dispersio
Journal of Banking and Finance 34, 2328-2345.
Jewell, Jeff, and Miles Livingston, 1999, A comparison of bond ratings from Moody's, S&P,
Fitch, Financial Markets , Institutions and Instruments 8, 1-45.
Kisgen, Darren J., 2006, Credit ratings and capital structure, Journal of Finance 61, 1035-10
Kisgen, Darren J., 2009, Do firms target credit ratings or leverage levels, Journal of Financial a
Quantitative Analysis 44, 1323-1344.
Kisgen, Darren J., and Philip E. Strahan, 2009, Do regulations based on credit ratings affect
firm's cost of capital?, Working paper, Boston College.
Klein, Benjamin, and Keith B. Leffler, 1981, The role of market forces in assuring contract
performance, Journal of Political Economy 89, 615-641.
Kliger, Doron, and Oded Sarig, 2000, The information value of bond ratings, Journal of Fina
55, 2879-2902.
McFadden, Daniel, 1973, Conditional logit analysis of qualitative choice behavior, in Paul
Zarembka, ed. Frontiers in Econometrics (Academic Press, New York).
Merton, Robert C., 1974, On the pricing of corporate debt: The risk structure of interest ra
Partnoy, Frank, 2006, How and why credit rating agencies are not like other gatekeepers, Work
paper, University of San Diego.
Poon, Winnie P. H., 2003, Are unsolicited ratings downward biased? Journal of Banking
Finance 27, 593-14.
Poon, Winnie P. H., and Michael Firth, 2005, Are unsolicited credit ratings lower? Internatio
evidence from bank ratings, Journal of Business Finance and Accounting 32, 1741-771.
Sangiorgi, Francesco, Jonathan Sokobin, and Chester Spatt, 2009, Credit-rating shopping, sel
tion and the equilibrium structure of ratings, Working paper, Carnegie Mellon University.
Skřeta, Vasiliki, and Laura Veldkamp, 2009, Ratings shopping and asset complexity: A theory
ratings inflation, Journal of Monetary Economics 56, 678-695.

202.115.120.41 on Wed, 07 Sep 2022 07:45:07 UTC

Bongaerts Tiebreaker

Uploaded by

Copyright:

Available Formats

You might also like

Bongaerts Tiebreaker

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bongaerts Tiebreaker

Uploaded by

Copyright:

Available Formats

一决胜负！

Tiebreaker: Certification and Multiple Credit Ratings

Stable URL: https://www.jstor.org/stable/41419673

This content downloaded from

Tiebreaker: Certification and Multiple Credit

DION BONGAERTS, K. J. MARTIJN CREMERS, and WILLIAM N. GOETZMANN*

*Bongaerts is with Finance Group, RSM Erasmus University Rotterdam, and Cr

This content downloaded from

demand for HY bonds can significantly increase the cost of

This content downloaded from

Under the information prod

This content downloaded from

and S&P ratings, investors do not appear to incorporate the

This content downloaded from

role in the long run. If th

A. Credit Rating Agencies and Regulation

This content downloaded from

U.S. insurance companies explicitly rely on NRSRO rating

7 See http://www.occ.treas.gov/law/basel.htm for an overview of legal and regulatory news per-

This content downloaded from

institutions center self-reg

B. Why Multiple Ratings M

This content downloaded from

This content downloaded from

witnessed in the case of AI

This content downloaded from

This content downloaded from

where the unconditionally e

16 An alternative way to hedge is

This content downloaded from

while more uncertainty will make an additional rating m

As asset pricing relies fundamentally on the production and disseminati

This content downloaded from

number of papers look评级变化会向市场传

18 Adverse selection may explai

This content downloaded from

126 The Journal of Finance®

This content downloaded from

II. Sample Construction and Methodology

We measure uncertainty or opaqueness by analyst dispersion of the fir

B. Data and Filters

19 To avoid capturing timing mismatches between (multiple) rating transitions, we require

This content downloaded from

Fixed Income Securities Database (FISD), equity data from

20 The TRACE database starts in July 2002.

This content downloaded from

Figure 1. FISD database coverage

sample rated by each of th

23 FISD confirmed that, as of mid-

This content downloaded from

Equity market data are obtained from CRSP. We calculate

This content downloaded from

Summary Statistics for Credit Spreads Sample

Variable N Mean Std. Dev. Min Max Explanation

This content downloaded from

Figure 2. Average credit spreads by rating category.

III. Empirical Results

This content downloaded from

Average Rating Differences

Difference -0.538* -0.397* 0.154* 0.185*

Difference -0.751* -0.599* 0.165* 0.203***

Difference -0.294* -0.223* 0.072 0.117**

Difference -0.492* -0.0574 0.478* 0.471***

Difference -0.800* -0.457 0.369* 0.430***

MSP A+ to A- Rating 1.553* 1.548* 1.568* 1.524*

Fitch Rated 0.0128* 0.0102*

MSP AAA to AA- Rating -2.889* -2.887* -2.891***