Credit Rating Agencies

Giulia Mennillo

• Verlag: Agenda Publishing
• Jahr: 2022
• ISBN:
  9781788215404
• Thema: Economics
• Thema: Banks & Banking

Credit rating agencies (CRAs) assess the creditworthiness of debt issuers on financial markets. They are private companies and the ratings they issue are judgements about the prospect of repayment of debt by an issuer in time and in full. A rating is not an investment recommendation, but it is an opinion about the creditworthiness of a financial product or an issuing bank, government, supra- or sub-national institution.

In recent years CRAs have gained an authority in bond markets that far surpasses their original design. The financial crisis of 2008 thrust the CRAs into the spotlight as their highly rated financial products turned out to be toxic assets. CRAs were blamed not only for their excessively optimistic ratings, but also for their complicity in creating them.

This short book introduces and explores the complex world of the credit rating industry: how it works, how it has evolved, the role it played in the financial crisis, and how it is regulated. Giulia Mennillo shows, as constitutive actors of global financial capitalism, CRAs have a social and political relevance that reaches well beyond finance into areas of transport, infrastructure, education and health and their impact is emblematic of the increasing financialization of our world.

1

The “what” and the “who” about credit rating

What is a rating and what is it not?

Ratings are conventionally, and erroneously, regarded as metrics about the probability of default of a debt issuer. Ratings are not the product of calculation, however. In fact, rating is a judgement on a debt issuer’s prospect of repayment of the liability on time and in full. Legally speaking, a rating is not an investment recommendation. In line with the rating industry’s original business idea, a rating offers an opinion about the creditworthiness of a financial product. Investors may consider this opinion before taking their investment decisions, but a rating does not exempt investors from due diligence. This distinction is highly relevant, as it protects credit rating agencies from liability claims by virtue of the first amendment of the US constitution, which guarantees freedom of speech and of the press, among other things. According to Moody’s Investors Service (2013b: 4), credit ratings are “forward-looking opinions of the relative credit risks of financial obligations issued by non-financial corporates, financial institutions, structured finance vehicles, project finance vehicles, and public sector entities”. According to the website of S&P Global Ratings, a credit rating is best described as an “informed opinion”,¹ conveying the impression of an underlying well-thought-out process instead of an ad hoc, spontaneous whim:

Credit ratings are forward looking opinions about an issuer’s relative creditworthiness. They provide a common and transparent global language for investors to form a view on and compare the relative likelihood of whether an issuer may repay its debts on time and in full. Credit ratings are just one of many inputs that investors and other market participants can consider as part of their decision-making processes.

From these official definitions of credit rating, we can make a few observations: First, the two major CRAs concur in defining ratings as “forward-looking opinions”. The word “judgement” does not appear. There is an emphasis on ratings being relative in two ways: first, because of the ordinal scale, ratings are supposed to be relative measures of credit risk in comparison to each other; and, second, ratings should be used as one input among many to assess creditworthiness, which suggests that the CRAs reject claims of absoluteness. At the same time, CRAs claim authorship for the common language of credit risk. Therefore, the implication is that ratings are constitutive of creditworthiness; creditworthiness has come to mean “credit rating”. Finally – as if CRAs were aware of the tensions between defining rating as opinion and stylizing it as a technical product – they do not refer to credit ratings explicitly in terms of “probability” of default but, rather, of “likelihood”. Unlike the former, “likelihood” does not necessarily imply that the rating is a metric, but it does not explicitly exclude it either.²

To a certain extent, the easy-to-understand alphanumerical rating symbol conveys the impression of objectively calculated probabilities. Instead of using cumbersome text to reproduce what a rating actually is declared to be, namely an “informed opinion”, the conversion into the letter grade facilitates the appearance as a quasi-scientific relative measure of credit risk. The rating symbol helps ratings pretend to be something they are not, inducing market participants to believe that ratings are unambiguous figures. As if the rating users desire a reality that can be captured in a probability distribution, the symbol offers a pleasant fallacy of precision, assuming a law-like repeatability of history, while assuming away contingencies of social reality.

Why does this distinction between what a rating actually is in legal terms and what it is commonly thought (and wished) to be matter? It has something to do with the CRAs’ authority. CRAs benefit if ratings are perceived as probabilities of default. A “positivist” aura makes ratings appear objective.³ CRAs gain in credibility if ratings are regarded as metrics. Therefore, it is not surprising that CRAs do not explicitly reject the notion of ratings being a technical product. The perception of ratings as measured probabilities of default amounts to a farce as it suggests a positivist epistemology without entirely following it. Sinclair (2005: 46) maintains that “agencies assert that rating determinations are opinions but simultaneously seek to objectify and offer their views as facts”. Thus CRAs face a continuous dilemma that demands a delicate balancing act from them. In turn, to equate an opinion with a technical product takes courage, and confirms the CRAs’ self-conception of their authoritative status. Market participants’ tacit tolerance of this tension only testifies to this.

The authority of CRAs – and, related to it, the popularity of ratings – cannot be explained without taking into account firmly held beliefs about the possibilities of how we can know about the future. This epistemological aspect is crucial for understanding the agencies’ success. The very “existence of rating agencies necessarily assumes predictability” (Sinclair 1999: 164). The rating symbol invokes predictability, and, with that, the assumption that the future is knowable and can be anticipated. This is independent of the number of cases of default that are used as a basis for their assessments. In other words, the impression that the future is predictable is conveyed both with “small” and with “large” n (i.e. cases of default). The rating symbol does so without resorting to stochastic measurements, regardless of whether these calculations are even feasible.⁴ The question as to whether prediction is possible at all is irrelevant in the case of CRAs. What matters is that the rating format gives a pretence of predictability and computability, and market participants buy into it.

Paradoxically, the more ratings pretend to be something they are not, the more credible they become. By portraying their ratings as metrics and products of stochastic calculations, not only do CRAs become accomplices to scientistic approaches but their ratings gain legitimization. The more credibly the impression of stochastic calculations based on “hard facts” is conveyed, the more the judgemental, inconclusive character of the creditworthiness assessment is glossed over. This is why, within certain boundaries, it is not surprising that CRAs vow to stick to a scientistic methodology in their official communications. For example, Fitch states in the case of sovereign ratings (Fitch Ratings 2002: 3–4):

It is important, though, that investors realise the limitations of this exercise, which is necessarily far less certain than our ability to analyse either bank or corporate risks of default. The essential problem is that the world of sovereign borrowers is far smaller than the world of large banks or corporations, and that the number of instances of default in the modern period when we have reasonable national accounts is tinier still. [...] So the rating of sovereigns depends more on the art of political economy than on the science of econometrics.

The Fitch warning suggests that, if there were sufficient n, an indication of the probability of sovereign default would be based more on econometric calculations; the opinion, would become, so to speak, less opinion, and more scientific. Against the background that the proportion of the admixture of quantitative and qualitative data as well as interpretive elements remains unknown, the differentiation regarding whether the rating judgement is based on the art of political economy or the science of econometrics is actually obsolete, whether with small or large n. Further, the rating symbol conveys the impression of calculated predictability and comparability either way.

Fitch’s statement can be understood as a concession regarding the limits of the positivist stance that CRAs usually adopt to be credible and, at the same time, as an invocation thereof. The inadequacy of calculation in the special case of sovereign ratings is attributed to the “not enough n problem”, and not to its inherently judgemental nature. The irony in all this is that, even if there were enough cases of default, a rating would not amount to a measure of probability of default. If rating was a repeatable algorithm then there would not be a business case for having CRA ratings, as everyone could reproduce them. But ratings make a difference, as they “price in the priceless”.⁵

A common understanding limits the agencies’ role to one of an informational intermediary that decreases the information asymmetries between investors and borrowers. From such a perspective, a rating is a suitable problem-solving device that coordinates actions “between those having funds to invest and those seeking funds to use” (Sinclair 1999: 164). In comparison to when ratings were invented at the beginning of the twentieth century, and given the increase in information asymmetries, today’s market-based finance has increased the demand for information about credit risk and, thus, for ratings.⁶ Unsurprisingly, and in line with conventional wisdom, CRAs portray themselves in such a functionalist perspective in their self-representation. They see the usefulness of ratings as decreasing market inefficiencies, which is just an alternative expression for reducing information asymmetries: “Credit ratings help facilitate an efficient capital marketplace. They provide transparent third-party information that’s not only forward-looking, but standardized for consistency.”⁷

In a functionalist perspective, information about credit risk is seen as neutral, objective and measurable, and thus as unambiguous, with no room for interpretation. Even though this is seldom spelled out, this implies that the agencies’ work is assumed as replaceable and multipliable. CRA ratings are seen as independent, third-party opinions about credit risk. Since borrowers would be incentivized to downplay the riskiness of their projects, whereas those having funds – the investors – would have an incentive to overestimate these risks, the idea is that CRAs minimize the bias in reporting credit risks. In the ideal case, CRAs have the key role of balancing the distorted perceptions of credit risk between the two parties in order to provide a supposedly neutral assessment of the debt issuers’ creditworthiness.⁸

The acknowledgement of the authoritative standing of only a few agencies whose ratings matter challenges such a reading. Ratings are not reproducible technical metrics that can be issued by any entity. When issuing ratings, CRAs provide centralized interpretations of creditworthiness that are commonly shared between market participants. This is why questions of authority and rating are inextricably linked. Ratings provide orientation for a multitude of actors in disintermediated markets. An unlimited number of rating agencies intermediating between borrowers and lenders, as assumed in a functionalist perspective, would undermine this guiding function of rating.

The centralized, dominant interpretations of creditworthiness that CRAs provide do not emerge from nothing. Rating is based on knowledge that is intersubjectively shared and validated among the finance and business communities. A rating is an aggregated social judgement. For example, sovereign ratings shape and reproduce the common sense, the market orthodoxy, the norms and dominant ideas about fiscal and economic policy. From these observations we can draw two implications in terms of the relevance of CRAs that go beyond a conventional functionalist reading of rating. First, in order to establish whether CRAs matter, it is to no avail to frame the question as “Who follows whom?”: the market following the CRAs or the other way round. This is attributable to the inherently intersubjective nature of the rating business. Much scholarly effort went into analysing empirical evidence as to whether markets “really follow” the CRAs. Trivial though it sounds, whether ratings have an effect on the market’s creditworthiness perception of bond issuers – “interest rates”, as economists would say, or “yields”, as practitioners say – is difficult to grasp in econometric terms (this often holds true for econometrics, and causality in general). Because of this measurement problem, the claim that CRAs “matter” lacks robust evidence. Ratings are imputed to have no independent influence on interest rates, resulting in a tenaciously held mantra that ratings would “only follow the markets”, and, by extension, that CRAs would not matter. Taking intersubjectivity into account, we can decouple the discussion about the CRAs’ authority from the “Who follows whom?” question, and glean important insights into the very nature of the agencies’ role in today’s financialized capitalism.

Second, as a consequence of the agencies’ authority – more precisely, epistemic authority (CRAs define the relevant knowledge about creditworthiness) – ratings are even constitutive of creditworthiness. Creditworthiness is defined by ratings; “creditworthiness” has come to mean, and boils down to, “credit rating”. Illustrative evidence for this bold claim can be found, for example, in the Free Dictionary. The online dictionary and encyclopaedia paraphrases the adjective “creditworthy” as having an “acceptable” or “satisfactory” credit rating.⁹ CRAs have gained power of definition over creditworthiness: a bond issuer is deemed creditworthy because it has the top rating, rather than the top rating reflecting a preconceived notion of creditworthiness. As CRAs invented and can claim authorship of the common language of credit risk on financial markets, this legitimizes and testifies to their position as epistemic authority. At the same time, thanks to their position as epistemic authority, they enjoy the credibility and reputation to continue to provide the common language of risk – several crises notwithstanding.

CRAs as a case of structural power

The authority of CRAs can be conceptualized by, for example, drawing on Strange’s structural power (Mennillo 2016). Strange (1998: 24–5) defines structural power as “the power to shape and determine...

Inhaltsverzeichnis

• Cover
• Half Title
• Series Information
• Title Page
• Dedication
• Copyright Page
• Contents
• Acknowledgements
• List of figures and tables
• List of abbreviations
• Introduction
• 1 The “what” and the “who” about credit rating
• 2 What do credit rating agencies do?
• 3 The use of ratings
• 4 Credit rating agencies under criticism
• 5 Regulating the credit rating agencies
• 6 Credit rating in China
• Conclusion
• Notes
• References
• Index