Rating Agency Conflicts: Issuer-Pays and Dodd-Frank

Key Takeaways

• The issuer-pays shift happened in the early 1970s; before that, investors subscribed to rating reports, aligning agency incentives with users rather than issuers.
• Rating shopping allowed structured-finance arrangers to seek indicative ratings from multiple agencies and hire only those producing the highest grades, creating a race to the bottom.
• Dodd-Frank Section 932 mandated annual SEC examinations of every NRSRO and required separation of sales staff from rating analysts.
• S&P's $1.375 billion 2015 settlement and Moody's $864 million 2017 settlement were the largest penalties ever paid by rating agencies, though neither admitted legal violations.

What It Is

Until the early 1970s, rating agencies were paid by investors who bought subscriptions to rating reports. Moody's and S&P then shifted to the issuer-pays model: the bond issuer pays a fee (typically a few basis points of deal size) to have its debt rated. Fitch followed. The change let agencies scale revenue with bond issuance rather than with subscriber counts.

Issuer-pays creates a structural conflict. The customer writing the check is also the subject of the opinion. If the opinion disappoints, the customer can take future business elsewhere. The conflict is most acute for complex products where the issuer can restructure a deal to chase a target rating.

The Intuition

Two forces keep the conflict partly in check. First, a rating agency's franchise depends on reputation. If ratings drift away from actual default rates, bond investors stop paying attention and the franchise collapses. Second, competition among Moody's, S&P, and Fitch means an agency that gets too lax risks being exposed by the other two.

Both forces weakened in structured finance before 2008. The reputational cost was slow to arrive because default experience on newly issued RMBS and CDOs took years to materialize. Competition became a race to the bottom as issuers could shop between agencies and choose whichever gave the highest rating for a given tranche structure. By 2007, structured-finance ratings accounted for nearly half of Moody's ratings revenue, concentrating the conflict in exactly the products that later failed.

How It Works

Issuer-pays fee flow. The issuer hires one or two NRSROs, shares the data, iterates the deal structure, and receives the ratings. Fees are paid regardless of rating outcome, but future mandates depend on analyst relationships and the issuer's sense of being treated "fairly."

Rating shopping. For structured deals the issuer can ask multiple agencies for indicative ratings, then hire the ones that give the best grades. Agencies that repeatedly produce lower grades lose market share. A 2011 SEC study documented this dynamic in the RMBS market.

Pre-rating the structure. Agencies publish criteria detailing how tranching, subordination, and credit enhancement map to ratings. Sophisticated arrangers reverse engineer the criteria: they build deals that hit AAA with the minimum enhancement the criteria allow. Small errors in agency model assumptions translate to large errors in rated outcomes.

Dodd-Frank rule responses.

• Section 932 required NRSROs to separate sales and marketing from rating analysts and to report when an analyst who worked on a rating later took a job at the rated issuer within 12 months.
• Rule 17g-5 (2010 amendment) required structured-finance issuers to share underlying data with all NRSROs, not just the hired ones, so unsolicited ratings could serve as a check on shopped ratings.
• Section 939F commissioned a study on replacing issuer-pays with an assignment system for structured finance. The study was completed; the assignment system was not adopted.
• Section 939G rescinded the Rule 436(g) safe harbor, exposing NRSROs to expert liability when ratings are included in registration statements.
• Form NRSRO disclosures now include separate performance statistics by asset class so investors can see whether structured-finance ratings have historically performed worse than corporate ratings.

Worked Example: The 2015 S&P Settlement

In February 2015 the Department of Justice and 19 states reached a $1.375 billion settlement with Standard & Poor's and parent McGraw Hill Financial. Half ($687.5 million) went to the federal government and half to the states and the California Public Employees' Retirement System.

The DOJ complaint alleged that from 2004 to 2007 S&P knowingly inflated ratings on RMBS and CDOs to win issuer business. Internal emails cited in the complaint showed analysts discussing how to soften criteria under competitive pressure. S&P did not admit violations of law but signed an agreed statement of facts acknowledging the conduct. The penalty was the largest ever paid by a rating agency at the time.

Moody's reached a separate $864 million settlement with the DOJ and state attorneys general in 2017 on similar allegations. Fitch was not sued, having had smaller market share in the most problematic 2005 to 2007 vintages.

Common Mistakes

• Believing Dodd-Frank fixed the conflict. The issuer-pays model is still the norm. Dodd-Frank added disclosure, exam oversight, and a conduct rulebook; it did not replace the business model.
• Dismissing investor-pays agencies. Smaller NRSROs (Egan-Jones, Kroll in some lines) operate on investor-pays or hybrid models. Their share is small, but their ratings can serve as a check, especially on issuers whose paid ratings look stale.
• Confusing the 2015 S&P settlement with a court ruling. The settlement included no judicial finding of liability. Investors sometimes cite the $1.375 billion as "proof S&P was found guilty of fraud." It is not.
• Ignoring structural fixes outside the U.S. Europe's ESMA supervises rating agencies directly and has fined S&P, Moody's, and Fitch for various breaches. Cross-border issuers face both regimes, and the rules are not identical.
• Treating unsolicited ratings as equivalent to solicited ones. NRSROs disclose when ratings are unsolicited. Unsolicited ratings often rely on public data only and may be less informed, though they can also be less conflicted.

Frequently Asked Questions

Why did the shift to issuer-pays in the early 1970s create such a persistent conflict? When investors paid for ratings, the agency's revenue depended on subscribers who had no stake in seeing inflated grades and every reason to demand accuracy. When issuers began paying, the agency's revenue became tied to deal flow from entities that had a direct financial interest in higher ratings. The conflict is structural: an analyst who repeatedly disappoints an issuer risks losing that mandate, while an analyst who accommodates rating-sensitive deal changes protects revenue. The conflict is self-reinforcing in boom years when the gap between inflated ratings and underlying credit quality is slow to show up in defaults.

What is rating shopping and why is it most damaging for structured products? Rating shopping occurs when an issuer solicits preliminary assessments from multiple agencies before engaging any, then hires only those that give the most favorable result. For standard corporate bonds, the practice is constrained because a company has a visible credit history that limits how far ratings can diverge. For structured products, the underlying pool of assets is assembled after the rating criteria are known, so arrangers can tailor the collateral, tranching, and credit enhancement to precisely hit the target rating from the most accommodating agency. The result is a floor, not a ceiling, on how lenient criteria can become.

What did Dodd-Frank Section 939F require and why was the assignment system never adopted? Section 939F directed the SEC to study replacing issuer-pays with a system where an independent board or the SEC itself would assign rating mandates for structured-finance deals, breaking the direct client relationship between issuer and agency. The SEC completed the study but concluded that an assignment system would be complex to design and administer without introducing new inefficiencies, and Congress did not mandate the system legislatively. The issuer-pays model therefore persists, with behavioral rules as the primary constraint rather than structural separation.

How does rating shopping interact with the 17g-5 data-sharing requirement? Rule 17g-5 requires structured-finance issuers to post underlying data on a password-protected website accessible to all registered NRSROs. The intent was to allow non-hired agencies to issue unsolicited ratings on deals, providing a check on the shopped rating. In practice, few NRSROs have the coverage capacity to issue meaningful unsolicited ratings on complex ABS deals, so the rule has had limited impact on rating shopping behavior. However, the data availability does allow investors and researchers to reanalyze deal collateral independently, which increases market scrutiny if not internal agency competition.

Is the issuer-pays conflict meaningfully worse for structured products than for corporate bonds? Yes, for two reasons. First, structured products are assembled after the rating criteria are published, meaning arrangers can reverse-engineer the criteria to squeeze maximum leverage into a structure that technically achieves the target rating. Corporate bonds cannot be structurally redesigned the same way because the issuer's business is fixed. Second, default experience on structured products takes longer to materialize than on corporate bonds, delaying the reputational feedback loop that normally disciplines ratings inflation. This combination made the structured-finance sector the focal point of both the 2008 rating failures and the subsequent regulatory enforcement.

Sources

1. DOJ. Justice Department and State Partners Secure $1.375 Billion Settlement with S&P (February 3, 2015). https://www.justice.gov/archives/opa/pr/justice-department-and-state-partners-secure-1375-billion-settlement-sp-defrauding-investors
2. SEC. Dodd-Frank Act Rulemaking: Credit Rating Agencies. https://www.sec.gov/spotlight/dodd-frank/creditratingagencies.shtml
3. Congressional Research Service. Credit Rating Agencies: Background and Regulatory Issues. https://www.congress.gov/crs-product/IF11916
4. Capital Markets Law Journal (Oxford). Dealing with the conflicts of interest of credit rating agencies. https://academic.oup.com/cmlj/article/17/3/334/6609964
5. S&P Global. McGraw Hill and S&P Ratings Reach Settlements with DOJ, State AGs, and CalPERS (February 3, 2015). https://investor.spglobal.com/news-releases/news-details/2015/McGraw-Hill-Financial-And-SP-Ratings-Reach-Settlements-With-DOJ-Attorneys-General-Of-19-States-And-District-Of-Columbia-And-With-CalPERS-2015-2-3/default.aspx

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.