The Hierarchy of Evaluative Authority: Credit Rating Agencies and the Limits of Evaluative Autonomy
Executive Summary
In the spring and summer of 2026, three of the world’s most consequential financial institutions found themselves on the receiving end of a contest usually reserved for the governments and companies they evaluate. On 22 April, twenty-three Republican state attorneys general wrote to Fitch, Moody’s and S&P Global Ratings, copying the Securities and Exchange Commission, alleging that the agencies had allowed environmental, social, and governance (ESG) considerations to distort their fossil fuel ratings and demanding that the downgrades be reversed or justified on narrower, non-ESG grounds. On 27 August, a coalition of twenty attorneys general led by Letitia James of New York and Rob Bonta of California wrote to the same Commission disputing the factual basis for that letter and arguing that its underlying demand risked pressuring the agencies to abandon independent, fact-based methodology. Read together, the two letters look like another instalment in the American argument over ESG. This essay argues that they are something more analytically interesting than that.
Credit rating agencies are conventionally studied as evaluators: institutions whose opinions shape the cost of capital for governments, corporations and municipalities, and whose judgements those actors cannot simply decline to anticipate. This essay inverts the angle of enquiry. It asks what happens when the evaluator itself becomes the object of scrutiny by actors who command genuine legal and political leverage. The answer rests on a distinction between evaluative authority - the capacity to render consequential judgements about others - and evaluative autonomy - the capacity to determine the terms on which those judgements are reached without an external party substituting its own preferred conclusion. Credit rating agencies possess the former in abundance. The two 2026 letters expose how conditional the latter actually is.
The essay does not adjudicate the underlying climate dispute, nor does it treat the two interventions as mirror images pursuing opposite substantive outcomes through identical means. The April letter makes explicit demands for changed behaviour. The August letter argues chiefly against the legitimacy of enforcement pressure and states plainly that its signatories take no position on the agencies’ actual conduct. What both letters share is a structural feature rather than a shared aim: each is compelled to argue, from opposite positions, about which evidentiary register should govern legitimate forward-looking credit judgement, exploiting the fact that an agency’s own procedural language, such as sufficient visibility or reasonable certainty, cannot fully determine that question on its own. The essay traces how American and European law each try to hold open a narrow space in which agencies can be governed without being told what to conclude, and asks how stable that space actually is. It closes with a more precise account of hierarchy: not a single vertical order with credit rating agencies fixed permanently at the top, but a relational structure in which the same institution can hold authority over one set of actors while remaining subject to another’s capacity to govern the terms of its own judgement.
A Familiar Institution in an Unfamiliar Position
Credit rating agencies are used to being the ones doing the judging. Governments structure fiscal announcements with an eye to how the major agencies will read them. Corporate treasurers plan debt issuance around rating thresholds that determine borrowing costs for a generation. Bruner and Abdelal note that many sovereign governments have sought a rating even without intending to issue debt, purely as a signal of transparency and orthodoxy to markets and other governments (Bruner and Abdelal, 2005). This is the ordinary grammar of the industry, and it has produced a substantial literature examining the agencies as private authorities whose opinions carry weight independent of any formal governmental sanction (Sinclair, 2005).
What is less familiar, and considerably more revealing, is the spectacle of the agencies sitting on the other side of an evaluative relationship. In the space of four months in 2026, Fitch, Moody’s and S&P Global received two letters from state attorneys general that did not simply criticise particular ratings. They contested the terms on which the agencies were entitled to reach those ratings in the first place, and each did so by invoking instruments with real legal consequence. The first, sent on 22 April by twenty-three attorneys general led by Nebraska’s Mike Hilgers and including Texas, Alaska, Florida and nineteen others, accused the agencies of allowing ESG commitments to override stated methodology in their treatment of fossil fuel companies and fossil fuel producing states, framed around specific statutory categories: material contravention of stated methodology under Section 15E of the Exchange Act, undisclosed conflicts of interest, and coordinated conduct among three firms that together control the overwhelming share of the market (Hilgers et al., 2026). It closed by threatening referral to the SEC’s Office of Credit Ratings, to state consumer protection enforcement, and to the Department of Justice if the demands it set out were not met within specified deadlines.
The second, sent on 27 August by a coalition of twenty attorney generals led by James and Bonta, took a narrower and more defensive posture. It did not offer its own assessment of whether the agencies’ ESG related conduct was appropriate. It states this explicitly: the signatories take no position on the conduct of the ratings agencies themselves, reserving their argument for a single claim, that the April letter does not supply a legitimate basis for the SEC to open an enforcement investigation, and that pressuring the agencies toward a particular substantive conclusion would itself risk violating the statute protecting analytical independence (James and Bonta, 2026). It disputes several of the April letter’s factual premises along the way, most pointedly its selective use of a single scenario from the International Energy Agency’s 2025 World Energy Outlook, but its central argument is procedural: political actors should not use the machinery of enforcement to compel a change in independently determined ratings, regardless of which direction that change would run.
That asymmetry between the two letters - one demanding a specific outcome, the other resisting the demand for any outcome at all - matters for how the episode should be read. Two coalitions did not race to install opposite ratings by identical means. Each was drawn onto common terrain: both are forced to argue - using the same statutory vocabulary of methodology, disclosure and internal control - about where the line falls between legitimate oversight of how an agency reaches its conclusions and illegitimate interference with what those conclusions say. Neither side treats the agencies as simply immune from scrutiny. The disagreement is over where that line sits, which is a considerably more sophisticated question than whether ESG belongs in credit analysis, and it is the question this essay takes as its subject.
Authority Exercised, Autonomy Conditioned
The existing scholarship on credit rating agencies has, with good reason, spent most of its energy establishing that the agencies possess extraordinary power over those they rate. Sinclair’s study of the agencies as exercising a form of epistemic authority, one that does not seek to persuade so much as to render binding judgement, explains why sovereigns and corporations alike organise their behaviour around anticipated rating outcomes (Sinclair, 2005). Bruner and Abdelal trace how that authority is compounded by regulatory reliance: because ratings are written directly into securities law and banking regulation across much of the world, the agencies function as gatekeepers to significant pools of investment capital, a status no purely private information provider could achieve through reputation alone (Bruner and Abdelal, 2005). Kruck extends this into a theory of path dependent power, showing that once regulators have delegated authority to an intermediary and market practice has entrenched around its judgements, disempowering that intermediary becomes costly even for a determined regulator, because the intermediary has by then accumulated sources of power that were never fully contingent on the original delegation (Kruck, 2017).
This literature describes a relationship running in one direction: evaluator to evaluated. What the 2026 letters expose is a second relationship, running the other way, that the same literature acknowledges without placing at the centre of analysis. Abbott, Levi-Faur, and Snidal’s regulator intermediary target model is useful here because it insists that the relationship between a regulator and an intermediary such as a credit rating agency is never a simple, one-way delegation. The arrows can run in more than one direction, and intermediaries with genuine capabilities of their own retain the ability to shape or resist the terms of the arrangement even after regulatory reliance has been established (Abbott, Levi-Faur, and Snidal, 2017). Kruck’s own framework implies the same point differently: agencies were never simply given power and left alone with it. They operate inside a relationship that regulators can, however imperfectly and at whatever cost, still reopen (Kruck, 2017). This angle is often overlooked when the position of the regulator is discussed in the field.
This is the distinction worth making precise. Evaluative authority is the capacity to render judgements that carry real consequences for other actors; judgements that issuers and investors treat as consequential regardless of whether they agree with the underlying reasoning. Evaluative autonomy is narrower and more fragile: the capacity to determine and apply the terms on which those judgements are reached, free from an external party substituting its own preferred substantive conclusion for the agency’s analytical one. Credit rating agencies possess the first in a form few private institutions can match. The April and August letters together show that the second is conditional, bounded by legal frameworks built precisely to allow governments to supervise how ratings are produced without allowing them to dictate what those ratings should say.
Precision matters here about what the claim does and does not assert. It does not assert that either letter changed a single rating, and nothing in the record suggests that it did. The correct term for what the letters demonstrate is not capture, which would require proof that external pressure altered analytical judgement, but vulnerability: the fact that the channels through which such pressure could in principle be exerted - statutory, regulatory and reputational - are open and have been used from opposing directions within the same year. An agency can be entirely unmoved by a given letter and still occupy a position in which its autonomy is conditional rather than absolute.
Governing the Evaluator Without Becoming the Evaluator
The statutory architecture surrounding credit rating agencies in the United States rests on a deliberate and, on inspection, quite delicate compromise. Congress wants the SEC to be able to police how agencies operate: whether they maintain adequate internal controls, whether they disclose conflicts of interest, whether they actually follow the methodologies they publish. Congress does not want the SEC or any state to be able to tell an agency what its rating should conclude. Section 15E(c)(2) of the Exchange Act states this directly: neither the Commission nor any state or political subdivision may regulate the substance of credit ratings or the procedures and methodologies by which a nationally recognised statistical rating organisation determines them. The same statute, in Section 15E(c)(3), requires every such organisation to establish, maintain, and document an effective internal control structure governing adherence to its own methodologies, and to submit an annual report attesting to that structure’s effectiveness.
The 2014 SEC rulemaking that operationalised this requirement worked carefully to keep procedural oversight from sliding into substantive review. The final rule specifies that management is not permitted to conclude its internal control structure was effective if there were one or more material weaknesses during the fiscal year, and defines a material weakness as a deficiency or combination of deficiencies creating a reasonable possibility that a failure to implement or adhere to the agency’s own stated methodology will not be prevented or detected on a timely basis (U.S. Securities and Exchange Commission, 2014). The test asks whether the agency followed the methodology it published, not whether the methodology reached the right conclusion.
A federal district court applied that same distinction in the SEC’s enforcement action against Morningstar Credit Ratings. The court held that failing to provide any criteria dictating how, why, or when analysts could make discretionary loan-level adjustments amounted to a failure to establish and enforce an effective internal control structure, since a rating adjusted for reasons unrelated to the underlying loan, such as nudging a model-generated result toward an expected outcome, would leave the agency’s own customers unable to understand its methodology (SEC v. Morningstar Credit Ratings, LLC, 578 F. Supp. 3d 563, 568 to 570, S.D.N.Y. 2022, quoted in Hilgers et al., 2026). The court took care to note that requiring some effective controls was not a judgement on the substance of Morningstar’s methodology (SEC v. Morningstar Credit Ratings, LLC, 578 F. Supp. 3d 563, 576). The finding concerned disclosure and process, not whether Morningstar’s ratings were correct.
This is the boundary the two 2026 letters are testing from opposite directions. The April letter’s request that the agencies provide a written explanation of the specific financial basis for each maintained downgrade sits squarely within the procedural space Section 15E permits, framed around methodological adherence rather than a demand for a particular grade. But several of its detailed questions - asking an agency to explain why it has not reversed a downgrade given that a specific prediction did not materialise - edge toward asking the agency to justify a substantive conclusion rather than a procedural failure. The August letter seizes on precisely this ambiguity, arguing that the politically charged framing of some of the questions demonstrates an intent to compel a particular outcome rather than neutrally to gather information (James and Bonta, 2026). Whether that characterisation is fair is a judgement this essay declines to make. The boundary between governing the evaluator and becoming the evaluator is not self-applying. It depends on how a demand is phrased, and on which side of the line a court or the Commission eventually decides a particular request falls.
The European Union has constructed an almost identical compromise, and it is worth setting the two side-by-side because the parallel is unusually exact. Article 23 of the 2009 Regulation on credit rating agencies provides that neither the competent authorities nor any other public authorities of a member state shall interfere with the content of credit ratings or methodologies, even as the same Regulation grants those authorities extensive powers to access documents, demand information, and carry out on-site inspections (European Union, 2009). When supervision of the industry was recentralised at the European Securities and Markets Authority in 2011, the equivalent prohibition was carried over and strengthened rather than diluted: the amended Article 23 states that ESMA, the Commission, or any public authorities of a member state shall not interfere with the content of credit ratings or methodologies, while ESMA simultaneously acquired powers to demand information, examine compliance with methodological back-testing obligations, and impose sanctions ranging from public notice to outright withdrawal of registration (European Union, 2011). The 2013 amendments went further still, requiring agencies to notify ESMA of intended material changes to their methodologies and to publish discovered errors in how those methodologies were applied (European Union, 2013). The European regime has built exactly the same wall the American one has - oversight of governance, disclosure, and methodological consistency running alongside an explicit statutory bar on dictating outcomes - using different institutional machinery to do it.
That two separate legal systems, developed independently and shaped by different political histories, converged on the same structural compromise is itself suggestive of a recurring regulatory problem rather than an accident specific to American securities law or the 2026 episode. It points to something close to a structural challenge for any regime that wants to supervise a private evaluator without becoming the evaluator itself, because the only available levers - disclosure requirements, internal control standards, and consistency between stated and applied methodology - are inherently procedural, and procedural levers can only ever partially determine a genuinely open-ended, forward-looking judgement. The 2026 letters found that gap rather than inventing it.
It is worth being precise about how far the comparison travels. The instruments used in the American dispute - state unfair and deceptive practices statutes, antitrust theory built on the earlier action against major asset managers, and referral to a Commission answerable to a specific administration - are features of the American regulatory landscape. The European regime relies instead on direct supervisory authority vested in a single pan-European body. The comparison demonstrates that functionally similar arrangements, procedural oversight paired with a protected sphere of substantive judgement, can emerge through different legal instruments in different jurisdictions. Functionally equivalent is not legally identical.
Two Letters, Two Registers of Evidence
Let us set the legal architecture aside and look at what the two letters are arguing about on the ground. Both concern the same underlying question: what evidence legitimately supports a forward-looking judgement about the credit risk fossil fuel companies and fossil fuel producing states will face over coming years and decades. Both claim to be defending fact-based analysis, yet they build their cases on almost entirely different kinds of evidence.
The April letter leans heavily on enacted policy and recently observed market behaviour. It points to the withdrawal of the United States from the Paris Agreement, the rescission of the Environmental Protection Agency’s endangerment finding, the collapse of net zero investor alliances following legal pressure from several of the letter’s own signatories, continuing outflows from ESG focused funds, and roughly seventy billion dollars in write-downs booked by automakers on electric vehicle investments (Hilgers et al., 2026). Its central evidentiary move treats the International Energy Agency’s 2025 World Energy Outlook as support for continued growth in oil and gas demand through 2050, citing the report’s Current Policies Scenario, which the IEA itself defines as a snapshot of policy settings already in place, assuming no further change even where governments have signalled an intention to act (Hilgers et al., 2026).
The August letter does not deny the individual facts so much as it disputes their sufficiency. It observes that the Outlook presents three scenarios explicitly described by the IEA as not forecasts, and that the Stated Policies Scenario, sitting alongside the one the April letter relies on almost exclusively, projects oil demand peaking by 2030 and coal demand falling well below 2024 levels by 2035 (James and Bonta, 2026). It supplements this with a different category of evidence, physical climate science drawn from the World Meteorological Organization and from reinsurers such as Munich Re on rising catastrophe losses, alongside continuing state and international policy commitments it says the April letter omits (James and Bonta, 2026). Where the earlier letter privileges enacted policy and near-term realised outcomes as the proper measure of forward-looking credit risk, the later one privileges long-horizon physical risk and international policy trajectory, treating a single administration’s current posture as an unreliable guide to conditions a thirty-year bond will actually face.
Neither register is inherently illegitimate, and this essay takes no position on which should prevail in a rating committee’s judgement. The pairing demonstrates something narrower: that an agency’s own procedural vocabulary does little work in resolving the disagreement between them. Fitch weighs environmental factors according to the level of certainty with which it can predict a driver will occur. Moody’s incorporates environmental considerations when it has visibility into relevant trends. S&P Global extends its forecasts only over the period for which it has a sufficiently clear view of an issuer’s future performance (Hilgers et al., 2026). Each formulation sounds like a self-applying rule. None specifies whether enacted policy should be weighted more heavily than physical risk projected decades out, or whether a scenario its own authors describe as not a forecast should count as sufficiently certain evidence at all. These are questions a rating committee resolves through discretion, and discretion is precisely the space into which a determined external actor, whichever direction it is arguing from, can attempt to insert its own preferred answer.
Two related but distinct sources of exposure are worth separating here. One is external: legal and political pressure applied through statutes capable in principle of triggering an investigation regardless of whether the underlying methodological choice was reasonable. The other sits inside the methodology itself: standards like ‘sufficient visibility’ or ‘reasonable certainty’ leave genuine interpretive latitude, latitude that exists whether or not any attorney general ever writes a letter about it. The 2026 episode matters because it shows the first kind of pressure locating and exploiting the second kind of latitude. The latitude was there before the letters arrived and will remain there after both coalitions move on to other issues.
When a Forecast Fails, What Exactly Has Failed
One thread running through the April letter illustrates this mechanism with particular clarity. It repeatedly argues that because a given prediction - that governments would tighten climate regulation, that hydrocarbon demand would peak, that renewables would displace fossil fuels - has not materialised as expected, the original judgement must have been methodologically defective and the downgrade should now be reversed (Hilgers et al., 2026). This move deserves scrutiny on its own terms, because it conflates two claims that are logically distinct.
The first possible claim is that an assumption was unreasonable at the moment it was made, given the information available at that time. The second is that an assumption, reasonable enough when made, was subsequently overtaken by events the agency could not have been expected to foresee with confidence. These are different failures with different implications. A methodology producing the second kind of miss is not thereby shown to be defective. Forward-looking judgement about a horizon extending years or decades into the future will diverge from realised outcomes some of the time, and an agency’s own stated standard, that ratings should be updated as new information becomes available, is fully consistent with having made a reasonable initial call that later needed revision rather than having been wrong from the outset (Moody’s Investors Service, 2014).
The inference from an unrealised forecast to a methodological defect only works if one has already decided that near-term realised policy and market behaviour are the correct validation criterion against which a forward-looking rating should be judged. That is not a neutral inference. It is the evidentiary register question in miniature, folded into the temporal framing of validation rather than argued directly. Academic work on how agencies apply discretion within their published scorecards supports the broader point. Lennkh and Moshammer’s analysis of Moody’s sovereign methodology finds that a meaningful share of any rating reflects committee-level judgement layered on top of quantitative inputs, applied inconsistently across regions, income levels, and the direction of the rating change, and best explained by variables such as bond yields and GDP growth rather than by the formal scorecard alone (Lennkh and Moshammer, 2018). A missed forecast tells an observer that the future diverged from an expectation. It does not by itself tell the observer whether the expectation was reasonable when formed, whether the underlying methodology was actually being followed, or whether some influence other than ordinary analytical judgement drove the original call. Establishing any of the latter requires different evidence than a chart of realised outcomes.
Scrutiny with a Longer History
Political and legal pressure on credit rating agencies did not begin with a letter dated 22 April 2026. The most consequential precedent sits in the aftermath of the 2008 financial crisis, though the character of that precedent needs to be stated carefully rather than flattened into a simple analogy with the present dispute.
In February 2015 the Department of Justice and nineteen states reached a $1.375 billion settlement with S&P over its ratings of residential mortgage-backed securities and collateralised debt obligations issued between 2004 and 2007. The agreed statement of facts records S&P’s admission that decisions about its rating models were affected by business concerns, and that the agency had, at points, declined to downgrade underperforming assets out of concern for client relationships, notwithstanding public representations that its ratings were independent (U.S. Department of Justice, 2015). Moody’s later reached a comparable settlement with the Department of Justice and a coalition of states involving a payment of $863,791,823 and a set of compliance commitments the agency agreed to maintain for five years (Moody’s Corporation Settlement Agreement, 2017). Separately, in January 2015, the SEC censured S&P and ordered it to pay disgorgement of $6.2 million, prejudgment interest of $800,000, and a civil penalty of $35 million over an undisclosed change in how it calculated a key credit metric used to rate commercial mortgage-backed securities, finding that S&P had altered its methodology without disclosing the change and without internal controls adequate to catch an internal complaint that had flagged the problem (U.S. Securities and Exchange Commission, 2015).
These episodes establish something specific: that credit rating agencies have long been legally exposed to scrutiny of their integrity, their disclosure practices and their adherence to stated methodology, and that regulators and state attorneys general have not hesitated to use that exposure when they believed an agency had misrepresented the independence of its own judgement. What they do not establish is that the 2026 dispute simply repeats that pattern with the political valence reversed. The post-crisis cases involved admitted misrepresentation and undisclosed methodological changes made under commercial pressure from issuer clients within the issuer-pays model. The 2026 letters involve a live disagreement over which evidentiary register should govern a forward-looking judgement about an unresolved policy question, argued by state officials rather than by federal prosecutors, with no admission of wrongdoing on either side and no settlement in sight. The earlier cases are best read as precedent for the proposition that the machinery of supervision, internal controls, disclosure obligations, methodological consistency, is capable of being activated against these agencies when there is reason to think it should be. They are not direct precedent for the particular kind of contest now underway, in which the underlying complaint concerns less what was concealed than which facts should count as decisive for a judgement that remains, on any reasonable view, genuinely uncertain.
Congressional oversight tells a related but distinct story. The House Financial Services Committee’s 2022 hearing on the credit rating industry, convened after S&P Global withdrew a proposed change to its insurer capital methodology following bipartisan objection, produced exchanges in which members from both parties pressed the agencies on transparency, competition, and the adequacy of their internal governance (U.S. House of Representatives Committee on Financial Services, 2022). That hearing shows that legislative scrutiny of how agencies reach their conclusions long predates the current dispute and has never tracked a single partisan direction. What distinguishes 2026 is not the existence of scrutiny but its simultaneity from two directions within a single calendar year, each invoking the same statutory categories in service of positions that are not simple mirror images of one another.
A Hierarchy Without a Fixed Apex
Credit rating agencies did not lose their evaluative authority in 2026. Fitch, Moody’s, and S&P Global continue to price risk for governments and corporations across the world, and nothing in either letter altered that basic fact. What the two letters together demonstrate is that this authority has never been unconditioned, and that the conditions attached to it are capable of being tested, from opposing directions and by different means, by actors equipped with genuine legal instruments rather than mere rhetorical objection. An institution can occupy a position of enormous consequence for the parties it evaluates while remaining answerable, in ways that are neither trivial nor fully determinate, to the legal and political systems within which it operates. The increasing strength of the nexus between financialisation and politics will make this more apparent in coming years rather than less.
This suggests a more precise way of reading the word hierarchy than a simple vertical chain with one institution permanently fixed at the top. Hierarchy, understood this way, describes the relational character of evaluative authority rather than a single fixed ordering. An institution may possess considerable authority over one set of actors while remaining subject to another institution’s capacity to govern the conditions under which that authority is exercised. A credit rating agency sits above the issuers and sovereigns it evaluates for the purpose of credit judgement. The same agency sits beneath the statutory and regulatory framework that determines how far its discretion in reaching that judgement extends, and beneath whatever body - an American court, a European supervisor, a state attorney general acting under consumer protection law - is empowered to test whether that discretion was properly exercised. Authority and subordination are not properties an institution simply has or lacks. They describe different relationships the same institution can occupy at once.
The regulatory architecture built around the rating industry, in the United States and in the European Union alike, tries to keep those relationships distinct by design, permitting oversight of process while forbidding dictation of substance. The 2026 letters show how difficult that separation is to maintain in practice, given that an agency’s own procedural vocabulary cannot fully specify which evidentiary register should govern a genuinely contested forward-looking judgement. That indeterminacy is not a flaw introduced by either coalition of attorney generals. It is close to an unavoidable feature of what it means to render an opinion about the future, and it is precisely the feature both letters have found a way to test, each from its own position within the hierarchy the statute was written to hold in place.
References
Abbott, K.W., Levi-Faur, D. and Snidal, D. (2017) Theorizing regulatory intermediaries: the RIT model. The ANNALS of the American Academy of Political and Social Science, 670(1), pp.14-35.
Bruner, C.M. and Abdelal, R. (2005) To judge Leviathan: sovereign credit ratings, national law, and the world economy. Journal of Public Policy, 25(2), pp.191-217.
European Union (2009) Regulation (EC) No 1060/2009 of the European Parliament and of the Council of 16 September 2009 on credit rating agencies. Official Journal of the European Union.
European Union (2011) Regulation (EU) No 513/2011 of the European Parliament and of the Council of 11 May 2011 amending Regulation (EC) No 1060/2009 on credit rating agencies. Official Journal of the European Union.
European Union (2013) Regulation (EU) No 462/2013 of the European Parliament and of the Council of 21 May 2013 amending Regulation (EC) No 1060/2009 on credit rating agencies. Official Journal of the European Union.
Hilgers, M., Cox, S., Uthmeier, J., Paxton, K. et al. (2026) Letter to Fitch Ratings, Inc., Moody’s Corporation, S&P Global Ratings and the Securities and Exchange Commission, 22 April.
International Energy Agency (2025) World Energy Outlook 2025. Paris: International Energy Agency.
James, L., Bonta, R. et al. (2026) Letter to the Honorable Paul S. Atkins, Chairman, and David Woodcock, Director, Division of Enforcement, Securities and Exchange Commission, 27 August.
Kruck, A. (2017) Asymmetry in empowering and disempowering private intermediaries: the case of credit rating agencies. The ANNALS of the American Academy of Political and Social Science, 670(1), pp.133-151.
Lennkh, R.A. and Moshammer, E. (2018) Sovereign ratings: an analysis of the degree, changes and source of Moody’s judgement. ESM Working Paper Series No. 27. Luxembourg: European Stability Mechanism.
Moody’s Corporation Settlement Agreement (2017) Settlement Agreement between the United States of America, certain States, and Moody’s Corporation, Moody’s Investors Service, Inc. and Moody’s Analytics, Inc.
Moody’s Investors Service (2014) Code of Professional Conduct.
SEC v. Morningstar Credit Ratings, LLC, 578 F. Supp. 3d 563 (S.D.N.Y. 2022).
Sinclair, T.J. (2005) The New Masters of Capital: American Bond Rating Agencies and the Politics of Creditworthiness. Ithaca, NY: Cornell University Press.
U.S. Department of Justice (2015) Justice Department and State Partners Secure $1.375 Billion Settlement with S&P for Defrauding Investors in the Lead Up to the Financial Crisis. Office of Public Affairs press release, 3 February.
U.S. House of Representatives, Committee on Financial Services (2022) Hearing on credit rating agencies, 117th Congress.
U.S. Securities and Exchange Commission (2014) Nationally Recognized Statistical Rating Organizations, Final Rule. Release No. 34-72936.
U.S. Securities and Exchange Commission (2015) In the Matter of Standard & Poor’s Ratings Services. Securities Act Release No. 9705, Exchange Act Release No. 74104, File No. 3-16348, 21 January.