Centre for Economic Policy Research

4.4
The evolution of the rating industry 
The origins of the industry can be traced back to the mid-nineteenth century,
when various firms began publishing analyses of the creditworthiness of 
commercial counterparties (Cantor and Packer, 1994). These reports responded to
a need on the part of suppliers of goods to judge how much credit to extend to
customers. The next step was to provide credit assessments of traded debt 
instruments. In 1890, Poor’s Publishing Company (the predecessor of Standard
and Poor’s) began to publish Poor’s Manual, which analysed various types of
investment, including bonds. Moody’s Manual of Industrial and Miscellaneous
Securities followed in 1900 (Moody’s, 2003).
With the growth in bond markets and the increasing number of investors, a
demand arose to have a simple means of ranking the investment quality of 
public issues. By the early 1900s, securities assessment had already moved beyond
analysis and had begun to classify bond issues into different quality groupings.
The beginning of ratings as they are known today can be traced to John Moody,
who in 1909 developed a methodology for translating various statistical measures
of credit quality into a single rating symbol. Moody’s Investors Services was 
incorporated in 1914, and was shortly followed by Poor’s, the Standard Statistics
Company, and Fitch Publishing Company (Cantor and Packer, 1994). 
Initially, credit assessments were provided only to paying subscribers. After the
establishment of the main rating agencies, these were generally institutional or
large private investors. Even in the early days, however, ratings came to be relied
on by users who were not subscribers. It became relatively common for private
contracts to limit fiduciary agents’ ability to invest in obligations below a certain
credit quality, as reflected in their ratings grade. Regulatory oversight of financial
institutions also took investment ratings into account. From at least 1930, bank
regulators made a distinction between investment and non-investment grade
securities in their assessment of banks’ portfolios. The SEC, under its net capital
rules for broker-dealers, required ‘haircuts’ from net worth to take account of the
credit quality of securities held. In 1936, the comptroller of the Currency’s Office
prohibited federally chartered banks from holding non-investment grade bonds
(Wakeman, 1984). At roughly the same time, US insurance supervisors also began
to use ratings in their oversight of insurance underwriters (Partnoy, 1999).
Although ratings had thus grown in importance in the interwar years, their
influence seems to have waned during the 1940s and 1950s. In a large part, this
may have reflected the fact that bond prices were not volatile and defaults were
few. This diminished both the demand for and the supply of relevant 
information. Partnoy (1999) characterizes this period as one of ‘austerity and 
contraction’ for rating agencies.
This situation began to change in the late 1960s and 1970s, when interest rates
and credit spreads became more volatile again, and the number and size of debt
issues began to increase. A defining event was the bankruptcy of the Penn Central
Railroad in 1970. From that time onward the demand for informed credit 
analysis has grown continuously.
An important change in the structure of the ratings industry, with potentially
significant implications for conflicts of interest, took place in the early 1970s.
44 Conflicts of Interest in the Financial Services Industry

---

Starting at this time, the major rating agencies began to charge issuers for ratings
assessments. One reason was the diminishing viability of investor subscriptions as
a source of financing ratings. Technological changes in the dissemination of 
information, including the spreading use of photocopying (White, 2001), had
made it increasingly difficult to restrict the beneficial use of ratings assessments
only to paying subscribers. Since ratings information quickly became available to
most market participants, the free-rider problem intensified and the value of 
subscribing to a rating service was diminished.
It might seem that in such circumstances the effective demand for ratings
would die away because they were becoming more of a ‘public good’. That is, they
are costly to provide, but once created they benefit everyone and not just those
who have paid for their provision. Under such circumstances, it is well known
that the product or service in question will be undersupplied unless some other
financing mechanism can be developed.
In the case of credit assessment, however, asymmetric information means that
issuers, as well as investors, have an interest in a credible certification mechanism
for loan quality. As discussed earlier, the dynamics of the adverse selection 
mechanism mean that all debt issuers have a need to certify the quality of their
debt. How much they are prepared to pay for a rating is an empirical matter, and
depends on their assessment of how much a credible assessment will result in a
reduction of their credit spread. In turn, this depends on the capacity of the 
rating agency to supply the market-place with credible additional information
and analysis.
The fact that issuers are prepared to pay for ratings is consistent with all or one
of three hypotheses. It could be that the rating assessment adds additional 
information or analysis that reduces information asymmetries. Investors are 
prepared to pay for this by accepting a lower yield on their investment, and 
borrowers can use part of the resultant interest saving to pay for the rating. It
could also be that, because of custom or regulation, a rating is a necessary 
precondition for access to capital, regardless of its informational value. Or it could
be that payment confers special benefits on the payer beyond simply the 
dissemination of credit information to the market. 
While these explanations suggest how rating agencies were able to move to an
issuer fee model as the scope for subscriber financing diminished, they do not 
adequately explain why the main agencies (Standard & Poors and Moody’s) rate
virtually all issues, regardless of whether the issuer pays. Nor does it explain why
98% of rated issuers nevertheless pay for their ratings (Partnoy, 1999). 
One possible explanation is that rated companies are afraid that, in the absence
of payment, the rating would be lower than justified by the underlying facts. The
SEC reports allegations that rating agencies have used ‘strong-arm’ tactics to
induce payment by issuers for unsolicited ratings (Securities and Exchange
Commission, 2003a). The Economist (1996) has suggested that ‘the suspicion exists
that a borrower who asks and pays for a rating may receive more favorable treat-
ment than one who merely attracts an agency’s uninvited “hostile” attention’.
The US Justice Department has investigated allegations on occasions in the past
(Economist, 1996) but no prosecutions have resulted.
Even in the absence of abusive tactics by the agencies, there may still be 
reasons why issuers may choose to pay for ratings. For example, it may be that the
issuer is in possession of favourable proprietary information on its prospects,
which for various reasons it is unwilling or unable to communicate directly and
credibly to the market. The ‘screen’ of an intermediary that has no direct finan-
cial interest could allow such background assessment to get into market percep-
tions and prices without revealing proprietary information, or exposing the issuer