The Ratings Oligopoly: How Three Firms Hold a Veto Over Sovereign Borrowing

When a government wants to borrow, it must first submit to a private judgment. Before a finance ministry can place a bond, before a pension fund or a bank treasury can buy it on favorable terms, three firms headquartered largely in lower Manhattan and London assign the issuer a letter grade. That grade does not merely describe risk; it sets the price of money. A single notch can move the interest a country pays by hundreds of millions of dollars over the life of its debt, can trigger forced selling by institutions that are contractually barred from holding sub-investment-grade paper, and can shut a sovereign out of capital markets entirely. S&P Global, Moody’s and Fitch are nominally publishers of opinions. In practice they operate something close to a veto over the cost of capital for every state and large corporation on earth, and they have held that position, with remarkable stability, for the better part of a century.

The shape of the cartel

The concentration is not subtle. The three agencies collectively account for roughly 95 percent of the global ratings market. S&P and Moody’s are the two heavyweights, together responsible for the large majority of all ratings outstanding, with Fitch occupying most of the remainder. A filing by the US Securities and Exchange Commission, which supervises the agencies under the Nationally Recognized Statistical Rating Organization (NRSRO) regime, found that the three firms produced more than 96 percent of all credit ratings outstanding. Roughly half a dozen smaller agencies are licensed, but they cluster in niches and command only fractions of the assets that matter.

What makes this durable is that the product is not really information; it is acceptance. A rating is valuable only insofar as the buyers of a bond, and the regulators who supervise those buyers, treat it as authoritative. That is a self-reinforcing property. An issuer pays the established agencies because investors and rules already recognize them, and investors recognize them because issuers already pay them. A new entrant with a more accurate model faces a coordination problem it cannot solve alone: no amount of analytical quality substitutes for the network of acceptance that the incumbents inherited.

The economics that follow are extraordinary. The ratings divisions of the largest agencies have reported operating margins above 60 percent, a figure more typical of a tollbooth than a research house. These are not the returns of a competitive industry; they are the rents of a position that cannot easily be contested.

Hardwired into the rulebook

The agencies did not seize this power; regulators handed it to them. Beginning in the 1930s and accelerating through the regulatory build-out of the late twentieth century, governments wrote credit ratings directly into law. Bank capital rules under the Basel framework used external ratings to determine how much capital an institution must hold against a given asset. Insurance regulators, money-market fund rules and central bank collateral frameworks all referenced the same letter grades. A pension fund’s investment mandate would permit it to hold only paper rated investment grade by a recognized agency.

This is the mechanism that converts a private opinion into a binding constraint. When a regulator says an asset is safe for capital purposes if and only if a recognized agency blesses it, the agency’s judgment ceases to be advisory. It becomes a license to access the cheapest pools of capital in the world. The Dodd-Frank Act recognized the danger explicitly: Section 939A ordered every US federal agency to strip references to credit ratings out of its regulations and substitute its own creditworthiness standards. Many years on, that excision remains incomplete, and ratings continue to function as critical inputs across the prudential system. The state created the dependency and has struggled to unwind it.

The conflict at the center

The business model contains a flaw so plain that it would be disqualifying in almost any other audit function. The agencies are paid by the issuers they rate. This was not always so. Until the early 1970s, the agencies sold ratings to investors through subscriptions. The shift to the issuer-pays model is usually attributed to two forces: the spread of the photocopier, which let subscribers share rating manuals freely and undermined the old revenue base, and the shock of the 1970 Penn Central collapse, which sharpened issuers’ demand to be certified. Charging the issuer solved the free-rider problem. It also created a permanent incentive to please the party writing the check.

The consequences are not hypothetical. In the structured-finance boom that preceded the 2008 crisis, the agencies awarded top grades to pools of subprime mortgage securities that later collapsed, and issuers shopped among them for the most generous treatment. The legal reckoning was eventually substantial: S&P settled with the US Department of Justice, a coalition of states and a large public pension fund for roughly $1.5 billion, and Moody’s reached a comparable settlement of about $864 million. Neither admitted a violation of law. The settlements were large in absolute terms and modest against the cash the franchise throws off each year. The structural conflict that produced the conduct was left intact.

The asymmetry over sovereigns

The power is felt most acutely by emerging-market governments, which have the least leverage and the most to lose. A sovereign downgrade does not simply raise the cost of the next bond. It cascades. Domestic banks and corporates, whose own ratings are typically capped at or near the sovereign ceiling, are repriced regardless of their individual balance sheets. Foreign funds tracking investment-grade indices are forced to sell. Empirical work has found that yield spreads tend to widen by around two percentage points in response to a downgrade, and that the agencies behave procyclically, upgrading in booms and downgrading into stress, which deepens the very crises they are meant to measure.

There is also a persistent question of method. The United Nations Development Programme has estimated that African sovereigns could save as much as roughly $75 billion if ratings rested on less subjective assessment, citing foreign-currency bias and asymmetric treatment of African risk. A Brookings analysis decomposed a comparable figure into more than $24 billion in excess interest and over $46 billion in forgone lending attributable to subjectivity in the rating of African states, noting that analysts at the global agencies frequently do not even visit the country in question. Whether one accepts those specific figures or treats them as upper bounds, the direction is consistent and the asymmetry is real: the same analysts who can downgrade a wealthy democracy and absorb a polite rebuttal can, with one decision, divert a meaningful share of a poorer country’s government revenue toward debt service.

Why every challenge has failed

The reform record is a study in entrenchment. The most direct attempt to break the issuer-pays conflict, advanced during the Dodd-Frank debate, would have had a public board assign agencies to issuers rather than letting issuers choose. It was diluted into a mandated study and never implemented. Europe explored a public ratings agency after its own debt crisis and abandoned the idea. The African Union has moved to stand up its own continental agency precisely because the existing structure is judged unfair, but a regional body still must win the acceptance of global investors and the recognition of global regulators, the same network effect that defeats every newcomer.

The reason challenges fail is that the barrier to entry is not technological or financial. It is institutional. The incumbents are written into statutes, index rules and investment mandates across dozens of jurisdictions. Dislodging them requires not a better firm but the simultaneous rewriting of the global regulatory apparatus that made them indispensable, and that apparatus moves slowly, fears the disruption of a transition, and includes regulators who quietly prefer outsourcing the judgment.

The deeper lesson is about how durable power is actually constructed. The agencies do not own assets, command armies, or print currency. They sell letters of the alphabet. Their leverage comes entirely from having been embedded, decades ago, into the plumbing through which capital is permitted to flow, and from a network of acceptance that competitors cannot replicate merely by being right. Until governments are willing to take back the creditworthiness judgments they delegated, and to bear the cost of building substitutes, three private firms will continue to hold a quiet veto over the price every state pays to fund itself. That is not a market failure to be corrected at the margin. It is the architecture working exactly as it was built.

Read our full Report Disclaimer.