Rating Shopping

Issuers and underwriters engaging in rating shopping approach multiple rating agencies and publish only the most favourable outcomes. This conflicts with a ratings’ integrity because agencies operating in competition for business have incentive to be lenient. The practice became notorious in structured finance before and during the 2008 crisis.

The fundamental conflict

A solicited rating involves the issuer hiring and paying an agency. This fee-for-service arrangement creates an obvious tension: the agency is paid by the entity it is assessing. If the agency issues a harsh grade, the issuer may shop elsewhere. If it issues a generous grade, the issuer is more likely to hire it again.

Rating shopping exacerbates this pressure. Instead of selecting one agency and living with the result, issuers can approach two, three, or four agencies simultaneously. They shop the deal or company, see which agency gives the best rating, and then publish only that best outcome—perhaps supplementing it with solicited ratings from a second agency if the first falls short of the desired investment-grade threshold.

Rational issuers will always prefer higher ratings: a BBB rating allows access to lower-cost capital than a BB, and an A rating opens doors that BBB does not. Rational agencies, knowing they risk losing business, have incentive to shade their estimates optimistically. The result is a downward bias in published ratings—they tend to be softer than the underlying credit quality justifies.

The structured-finance reckoning

Rating shopping reached its apex in the 2000s structured-finance boom. Investment banks would sponsor the creation of mortgage-backed securities and collateralized debt obligations backed by thousands of mortgages, and would approach agencies to rate the resulting tranches. The banks and underwriters knew that if one agency came back with a BB rating on a subordinated tranche, they could shop elsewhere until an agency assigned AAA to the same piece of paper.

Consider a simplified example: a bank structures an MBS with junior bonds that should, based on historical mortgage default rates and model assumptions, be rated BB. But one agency (under pressure to win business) rates it BBB. Another, even more competitive, rates it A. The bank publishes the A rating in its deal prospectus. Investors buy the bond expecting single-A credit quality. Mortgage defaults spike; the bond collapses. The agency that assigned A had the same data as the one that assigned BB—the difference was not information but competitive pressure.

This became systemic. Nearly all structured securities issued in the 2000s were shopped. The major agencies all knew they competed on “how good a rating can we justify for this deal.” The result: structured products were chronically overrated. When the housing market turned, massive losses followed, not primarily because of unexpected events but because the ratings had been inflated through competitive pressure and shopping.

Empirical studies bear this out. Securities that were shopped multiple times had higher subsequent default rates than those with single ratings. Agencies that lost deal volume due to conservative ratings faced pressure from senior management to loosen standards to regain market share. The Dodd-Frank Act post-crisis investigations exposed internal emails showing senior bank officials explicitly discussing which agencies to approach based on “who will give us the best rating.” It was not subtle.

How rating shopping works in practice

The mechanics: An underwriter or issuer hires a bank to arrange a bond offering or securitization. Before filing with the SEC, the bank approaches one or more agencies with a preliminary prospectus and loan data. Agencies assign a preliminary rating. If all agencies come back with the desired grade, the process is smooth. If one gives a lower grade, the issuer may:

1. Modify the deal (add more collateral, lower the coupon offer) to try to convince the conservative agency to upgrade.
2. Hire an underwriter that the lower-rating agency respects, hoping for a different conclusion.
3. Simply publish only the higher rating(s) and hope the lower-rated agency does not issue a public unsolicited rating to contradict.

In structured finance, option 3 was most common. The SEC required disclosure of all solicited ratings, but the rule had loopholes. If an issuer formally withdrew a request before the agency issued a final rating, the preliminary assessment need not be disclosed. Underwriters would pressure an agency with a preliminary harsh rating: “Are you sure about that? Do you really want to be out of step with Moody’s and Fitch?” The agency, facing the prospect of losing the deal, sometimes capitulated. If not, the issuer could simply decline to formally solicit a final rating, making the unfavourable preliminary finding non-disclosable.

In corporate bond markets, rating shopping was less blatant but still present. A firm issuing a bond would approach Moody’s and S&P. If S&P came back BBB and Moody’s came back BBB-, the firm might emphasize the S&P rating in its marketing materials to investors (the higher one). Both ratings would technically be published, but the narrative around the deal would highlight the higher grade.

Regulatory response and ongoing concerns

The Dodd-Frank Act (2010) required fuller disclosure of the rating process:

• Agencies must publish all methodologies and rationale for rating changes.
• Issuers must disclose that they solicited ratings and, in many cases, disclose what rating(s) they received even if unpublished.
• Agencies cannot tie compensation to rating outcomes or apply sales pressure internally based on rating leniency.

These reforms reduced but did not eliminate shopping. Modern bond prospectuses disclose multiple ratings more transparently. Yet the incentive remains: issuers prefer higher ratings, agencies compete for business, and some flexibility in applying methodology persists.

Securitization markets, which are less regulated than public bond markets, still experience shopping. A private fund raising collateralized by mortgages or commercial loans may shop for favorable ratings from smaller regional agencies if major agencies are cautious. Investors in these less-transparent deals have little visibility into whether shopping has occurred.

The broader epistemological problem

Rating shopping points to a deeper issue: when an agency is hired and paid by the issuer, can it ever be truly independent? The fee arrangement itself, regardless of formal protocols, creates psychology and incentive misalignment. Studies show that agencies’ empirical accuracy has not improved markedly since Dodd-Frank; ratings continue to be optimistic on average, and downgrades often cluster in or near recessions when losses mount.

One alternative—a system in which investors (not issuers) pay for ratings—would reverse the incentive. But such a model has never scaled; rating agencies built their business on issuer fees, and investors lack the coordination to fund an independent competitor. Some proposals call for a “public utility” rating model, where government funds neutral assessment. Others suggest mandatory rotation of agencies or limiting which agencies can rate which types of securities. None has been adopted at scale.

Rating shopping and systemic risk

The 2008 crisis exposed rating shopping as a vector for systemic risk. Because so many structured securities were overrated due to competition and shopping, the apparent credit quality of the financial system was illusory. When housing prices stopped rising and defaults accelerated, the real risk surfaced. Banks, pension funds, and insurance companies that had relied on ratings discovered their exposures were far riskier than advertised. The domino-like failures that followed—the collapse of mortgage-backed securities, the near-failure of insurance giant AIG, the bankruptcy of Lehman Brothers—can be traced partly to mispriced risk caused by shopping and competitive pressure on rating agencies.

Post-crisis, there is broader awareness of rating limitations. Sophisticated investors now view ratings as one input among many, cross-checking against market prices, credit spreads, and agency competitor views. Retail investors, by contrast, often remain overly reliant on ratings, a situation that persists in municipal bonds and some secured lending.

See also