1. Introduction

The importance of the financial system for achieving sustainable development goals (SDGs) can never be overstated.¹ By playing an allocative role, the financial system reorients capital supply towards SDG-oriented activities,² which then steers firm-level sustainable behaviours³ and impels the transition of the entire economic system.⁴

The contemplated status that the financial system forges and maintains long-term financing support for economic, social and environmental sustainability is ‘sustainable finance’.⁵ On a functional basis, it describes the process of integrating environmental, social and governance (ESG) considerations into investment decision-making.⁶ Driven by policy initiatives, the European financial markets are committed to mainstreaming sustainable finance.⁷ The EU’s Sustainable Finance Disclosure Regulation requires market participants and advisers to report their policies on the integration of sustainability risks and the impact of their investments on sustainability.⁸ In the UK, asset managers and owners shall make disclosures in accordance with the Taskforce on Climate-related Financial Disclosures (TCFD).⁹ Investors have also acted proactively: from 2016 to 2021, the number of signatories to the UN Principles for Responsible Investment (PRI)¹⁰ has surged by 250 per cent.¹¹ A survey carried out by the FTSE Russell in 2021 showed that 84 per cent of institutional asset owners have been implementing sustainability in their portfolios.¹²

What is equally important to the growth of sustainable finance is its effectiveness in capital allocation. The effectiveness is premised on investment decision-making being comprehensively informed by the ESG characteristics of business entities, of which access to their ESG information is the starting point and groundwork.¹³ To facilitate the provision of ESG information, standard-setting organizations have set up guiding frameworks and standards for corporate ESG reporting,¹⁴ parallel with which are the mandate disclosures initiated by national regulators more recently.¹⁵ Businesses are provided with a plethora of frameworks and standards based on differentiated concepts and techniques,¹⁶ and their ESG disclosures are resultantly in disparate patterns.¹⁷ The disparity in the breadth, depth, nature and format of information, which thickens the information thicket, is a material challenge for effective sustainable investing.¹⁸ In contrast to the oversupply of disclosure guidance is investors’ dissatisfaction with the inadequacy of ESG information.¹⁹ On the one hand, under voluntary reporting frameworks, businesses discretionally decide what kind of information to communicate to investors,²⁰ and on the other, mandatory regimes in the domain of public regulation are still under construction, resulting in disclosures being made in a piecemeal and incomplete manner.

As a spontaneous market response to the fact that sustainable finance is constrained by information unavailability, ESG information services burgeon.²¹ This industry is currently led by large providers operating as the ESG subsidiaries of well-known, leading financial services firms, such as MSCI, Bloomberg and Standard & Poor (S&P).²² It also continues to attract an increasing number of new entrants,²³ many of which have a line of business exclusive to ESG information.²⁴

In general, services provided by these firms range from the provision of raw ESG data to more sophisticated rating services.²⁵ To provide raw data, they aggregate and organize ESG information from piecemeal disclosures of business and private sources, such as questionnaires.²⁶ This work underpins rating services. ESG ratings are professional assessments of the ESG risks or opportunities a business entity or an investment instrument is exposed to and the impact of its operation on the sustainable development of broader society and the environment.²⁷ The format of ESG ratings is analogous to that of credit ratings, which are professional, independent opinions about the credit risk of a business. According to the assessment outcomes, entities or instruments are categorized into a hierarchy of grades (eg from ‘AAA’ to ‘CCC’) to indicate their ESG performance and impact relative to their peers. ESG scores and rankings are essentially the same services using different formats.²⁸

ESG ratings convey to investors implicitly a backward-looking verification of the original information about an entity or instrument, as well as explicitly a forward-looking, predictive assessment of its ESG performance and impact. The results of both processes are important for investors. When investors are not able to access, verify or analyse ESG information themselves, relying on ESG ratings is a simpler and quicker solution. As reported, asset managers and owners, such as investment firms and pension funds, as well as other financial institutions, such as banks, are the main users of ESG ratings.²⁹ They refer to ESG ratings to design financial products, assess sustainability-related risks in portfolio assets, and guide their stewardship and engagement strategies.³⁰ Nowadays, ESG ratings become the most widely and frequently used information services by investors,³¹ as well as the core business of service providers. For this reason, this article focuses on ESG rating services and applies the term ‘ESG rating firms’ to denote these providers.

It seems that the relatively competitive and dynamic industry environment of ESG ratings provides investors with a wide range of options. However, given the profit-driven motive of most participants and the laissez-faire status quo, the quality of many ratings may not satisfy investment needs. Investors have observed the use of out-of-date information, unreasonable estimations and inexperienced analysts³² and have ongoing concerns about the divergent and inconsistent ratings provided by different firms and the lack of transparency in methodologies.³³ They also cast doubt on the integrity and independence of the rating process, as rating firms are providing ESG-related certification and consultation services to rated business entities.³⁴

Investors’ concerns relate to the accuracy and reliability of ESG ratings and their impact on investment outcomes. From a holistic perspective, ESG rating firms, with their widespread influence on investing behaviours, are one of the main forces driving and directing the process of sustainable finance. Their failures in providing correct and appropriate guidance will lead to system-wide allocative inefficiency and thereby retard the achievement of SDGs. With a precautious stance, this article explores how to propel ESG rating firms to improve the accuracy and reliability of their ratings.

International organizations and financial regulators are proposing ad hoc regulatory schemes to address these concerns. However, regulators have overlooked the nature of the role of ESG rating firms in capital markets. As a result, the effectuation of their proposals is questionable. This article argues that ESG rating firms perform a gatekeeping function in the arena of sustainable finance because their ratings, if accurately made, help keep the ‘gate’ toward capital markets closed to those businesses falling behind in sustainability. The commonly recognized gatekeepers in capital markets include underwriters, lawyers, auditors, securities analysts and credit rating agencies (CRAs).³⁵ Based on their historic failures, there have been well-established views and arguments on the limitations of the market mechanism in monitoring gatekeeping practices and the paths for regulations to step in.

By applying the generic analytical framework of gatekeepers to ESG rating firms, this article argues that: as a compounded effect of the complex, multi-dimensional and variable attributes of ESG issues and the predictive nature of ratings, the accuracy of ESG ratings is extremely difficult to verify. This inherent defect creates a hotbed of both intentional and unintentional poor performance in ratings, which should be the focal issue that regulators aim to solve. Based on this argument, the present article critically comments on the weakness of the proposed disclosure rules in dealing with two major problems in ESG ratings—divergence and the lack of transparency because they fail to take into consideration the difficulty in verification. Alternatively, this article proposes a disgorgement scheme which derives from the idea of strict civil liability but dodges its disadvantages in the meantime. This scheme will require ESG rating firms to have a stake in the negative consequences of their ratings and bear a share of allocative inefficiency. This article also discusses the potential for conflicts of interest arising in the business of ESG rating firms and proposes a ‘no-conflict rating’ principle, which is a more rigid separation measure than the organizational arrangements in proposed regulations.

The remainder of the article is structured as follows. Section 2 articulates the rationales for identifying ESG rating firms as the gatekeepers of sustainable finance. Section 3 elaborates on investors’ concerns about the current rating practices and the solutions proposed by regulators, based on which it focuses on the particular characteristics of ESG ratings to argue that the market mechanism and the proposed regulations will not be effective in promoting rating quality. Section 4 introduces the ‘no-conflict rating’ principle and the disgorgement scheme and sheds light on their superiority. Section 5 is the conclusion.

2. The gatekeeping role of ESG rating firms in sustainable finance

As mentioned in the Introduction, in terms of ESG information, the problems of overload and scarcity co-exist, which exacerbates the difficulty for investors to obtain, organize and verify useful information for sustainable investing.³⁶ The government’s endeavours to mandate and standardize ESG disclosures will help alleviate information asymmetry. Nevertheless, given the wide spectrum, complexity and multi-layers of ESG issues and their variance across markets and industries,³⁷ processing ample information and conducting precise assessments of ESG performance and impact remain difficult and costly,³⁸ even for sophisticated investors.

As noted by Ronald Coase, a firm will buy a product from the market if the costs of producing it internally exceed the market price.³⁹ In capital markets, instead of processing information themselves, investors may go to specialized information providers⁴⁰ who produce well-analysed information to facilitate investments with cost advantages arising from economies of scale and expertise.⁴¹ Furthermore, information is a special product as it is highly vulnerable to leakage⁴² and depreciation from the first use in trading.⁴³ This attribute makes free-riding a significant problem in the production of information and necessitates specialized information providers to overcome it.⁴⁴

Traditional financial information is processed and conveyed by a range of professional bodies, such as auditors, analysts and CRAs.⁴⁵ These professions either review and verify historical information that is subject to transparency rules or provide independent investment-related opinions based on information analysis. In addition to facilitating the exchange and utilization of original information, their verifications and analyses also embed new information which adds value to the entire set of information used for transactions.⁴⁶ These are the typical functions of information intermediaries.

The proliferation of ESG rating firms resembles the evolution of financial information intermediaries, which can also be regarded as the financial markets’ self-improvement for informational efficiency.⁴⁷ ESG rating firms play a large part in the ESG information ecosystem: they obtain, verify and convey original ESG information⁴⁸ and create new information by publishing their analyses and assessments. Although the substances of financial and ESG information are different, the economic function performed by ESG rating firms is the same, that is, they are also information intermediaries.

ESG rating firms also exercise market oversight. They aggregate information about a business entity or instrument from multiple sources, and then corroborate and communicate authentic information, screen out erroneous or fraudulent contents and prevent them from reaching investors.⁴⁹ Through their ratings, these firms convey an implicit verification of the authenticity of underlying information.

This implicit verification acts as an antifraud mechanism.⁵⁰ As noted by Professor Choi, the antifraud function of information intermediaries becomes more valuable when other antifraud mechanisms are weak,⁵¹ which particularly refers to two situations: first, due to easy market entry, there are numerous service providers, which increases the costs for consumers to verify the quality by themselves; second, the legal liability for fraud cannot be effectively enforced.⁵² Both situations are the status quo of sustainable finance. In recent years, the volume of ESG-linked financial instruments has increased significantly.⁵³ Although they are subject to traditional capital market rules, labelling them as ESG-linked is free from any mandatory examination.⁵⁴ In the absence of regulation, there are increasing concerns about issuers falsely propagandizing their ESG commitments and taking advantage of the ‘ESG’ label for profit or using the funds unsustainably.⁵⁵ These phenomena are roughly defined as ‘greenwashing’.⁵⁶ In the UK and the EU, regulators are implementing labelling schemes for sustainable investment funds to combat greenwashing.⁵⁷ However, they are not monitoring schemes, which do little to help with the detection and investigation of greenwashing.

Moreover, under the current legal framework of securities fraud, it is hard to make issuers liable for their greenwashing actions. Despite being used frequently, there lacks an unequivocal definition of greenwashing,⁵⁸ and its denotation ranges from mild vagueness to deceit.⁵⁹ Many forms of greenwashing may not constitute securities fraud in law. Taking UK law as an example, a misleading statement is committed if an issuer conceals material facts dishonestly.⁶⁰ Considering ESG disclosures largely remain voluntary, the concealed ESG contents which are not subject to mandatory disclosures may fail to be identified as ‘material’.⁶¹ In addition, ‘economic loss’ is a required element of the claim against securities fraud.⁶² However, greenwashing does not necessarily lead to economic losses; instead, investors may earn high returns as a consequence of it.⁶³ Due to these two barriers, the existing legal liability for securities fraud is ineffective at deterring greenwashing. When powerful legal or regulatory forces against greenwashing in the capital markets remain absent, investors need to rely on private forces to safeguard their interests.

The existence of ESG rating firms as verifiers is meaningful for deterrence, and through their verification practices, they are expected to monitor ESG-related misconduct. In the sense that gatekeepers are private parties in a market who bear a watchdog function,⁶⁴ as opposed to the government as the public watchdog,⁶⁵ ESG rating firms can be regarded as the gatekeepers of the arena of sustainable finance.

As far as capital markets are concerned, gatekeepers safeguard the ‘gate’ through which a business can access the markets.⁶⁶ They comply with business or legal rules to decide whether or not to open the gate for a specific business by making professional judgements on its eligibility. The role of some gatekeepers is determinant and indispensable as they keep the only ‘gate’,⁶⁷ such as underwriters and auditors. A key function for them to exercise when gatekeeping is the refusal of support or cooperation based on the finding of false information.⁶⁸ Since their opinions on financial statements are legally mandated, the effect of their gatekeeping is theoretically powerful.

ESG rating firms do not have such controls stemming from their rating services. They can detect false information but cannot veto the underlying misconduct. Nevertheless, ESG rating firms play a guiding role via their rating services whereby they directly influence investors’ decision-making. This guiding role is another approach to gatekeeping. Not all investors have the capacity to distinguish between accurate and inaccurate information to make correct investment decisions. This problem is exacerbated by the complexity and diversity of ESG issues. ESG rating firms integrate the results of verification into their assessments of ESG performance and predictions of ESG impact and convert them into simple ordinals, which are straightforward for investors to understand. In so doing, ESG rating firms signify misbehaving businesses with low ratings or delisting them from their rating systems. Guided by these signals, investors will stay away from businesses with poor ESG performance. In a perfect situation (though unrealistic) where all ratings are correct and all investors are rational, the economic effect of rating is equivalent to ‘keeping the gate toward capital markets closed’.

‘Performing a gatekeeping role’ and ‘being able to fulfil the needs of gatekeeping’ are two different questions. The above discussions are about the former only, which demonstrate the significant impact that ESG rating firms can exert on sustainable finance. In terms of the latter, to become competent and accountable gatekeepers, ESG rating firms need to have sufficient expertise and skills and adopt reasonable and impartial mechanisms for ESG-related analysis and assessment.⁶⁹

Currently, ESG rating firms are the unitary gatekeepers for sustainable finance. This is different from traditional gatekeeping practices regarding financial performance, which are divided among several information intermediaries.⁷⁰ In a system of multiple gatekeepers, one’s error may not lead to wrong decisions as investors have chances to obtain correct opinions from others, whereas a unitary gatekeeper is not constrained by any checks and balances. This implies the greater and broader impact of ESG rating firms, as well as the utmost importance of their competence and accountability. However, in reality, investors constantly express their dissatisfaction with ESG ratings.⁷¹ In Section 3, the flaws perceived by investors in the operation of ESG rating firms and their influence on sustainable investing will be examined.

3. Concerns about ESG ratings and regulatory proposals

This section discusses whether market discipline, namely, the reputational capital mechanism and the ad hoc measures being proposed by regulators will be effective to incentivize ESG rating firms to improve the accuracy and reliability of their ratings. It starts with a brief overview of the theoretical arguments on reputational capital and then elaborates on the flaws perceived by investors in current rating practices and how regulators plan to address them. Building on these, this section critically analyses the potential effects of reputational capital and the proposed regulations.

Reputational capital

A seller can build up a good reputation by continuously providing high-quality goods or services to consumers, which constitutes its valuable capital asset.⁷² Gatekeepers accumulate their reputational capital by providing informational and advisory services over many years.⁷³ As repeated players in capital markets, they are employed as ‘reputational intermediaries’ by securities issuers who themselves face more difficulty to establish trust with investors.⁷⁴ Since trustworthiness is the essential endowment of a gatekeeper, it has incentives to maintain a good reputation among investors by providing accurate information and impartial opinions.⁷⁵ A significant implication of reputational capital arguments is that a market where trust is essential for transactions is in and of itself an effective monitoring mechanism, therefore, public regulatory intervention is unnecessary.⁷⁶

However, the historic failures of auditors⁷⁷ and CRAs⁷⁸ suggest that this reputational capital mechanism may not work as well as it is assumed in theory. As pointed out in scholarly arguments, it is rational for gatekeepers to undertake reputation-depleting activities, such as aiding or abetting misstatements or making inaccurate ratings, as long as the losses in reputational capital are lower than the profits they earn from these activities.⁷⁹ In particular, the diminishment in reputational capital occurs over the long haul whereas profit-making decisions are made with short-term horizons, and the actual decision makers—the employees of a gatekeeper share a stake in short-term profits but do not have to bear the consequences of its long-term reputation damages.⁸⁰

The extent to which the reputational capital mechanism can police gatekeepers is also determined by the verifiability of their statements or opinions and the costs of verification.⁸¹ If verification is difficult, the probability of misbehaving gatekeepers suffering materialized losses in reputational capital is low.

Concerns about ESG ratings

Divergence in ratings

It is unanimously proven by contemporaneous empirical studies that ESG ratings provided by different firms are highly divergent.⁸² Scholars identify that different categories and measurements of ESG factors applied by rating firms are the two primary sources of systematic disagreements.⁸³ The conceptual and methodological divergence in ESG ratings is inherently reflective of the multi-dimensional, complex and variable attributes of ESG issues.⁸⁴ For feasibility, rating firms need to select the key issues from a wide range of ESG factors, which vary across industries and markets.⁸⁵ This selection contains a certain level of subjectivity and discretion. The design of rating methodologies is the same: for an ESG issue with multi-faceted interpretations, firms may use different proxies and metrics;⁸⁶ the criteria applied to classify performance or impact as well as the weighting, calculation and aggregation methods may also be different.⁸⁷

It is argued that increasing and standardizing ESG disclosures is a solution.⁸⁸ However, empirical evidence reversely suggests that greater disclosures by rated companies lead to a higher level of divergence in ratings because more available information entails more subjective judgements.⁸⁹

The work of rating by its nature involves subjective judgements,⁹⁰ especially when some ESG information is qualitative or equivocal. Unlike credit ratings, the single object of which is creditworthiness, ESG ratings are integrated assessments of heterogeneous factors. In fact, credit ratings of complex financial instruments, such as collateralized debt obligations (CDOs) are more divergent than that of simple corporate bonds.⁹¹ The complexity of rated objects necessitates more sophisticated cognition and analysis,⁹² which then increases the probability of divergent outcomes. Therefore, rating divergence is unavoidable and will remain prominent in the field of ESG. By saying so, it is not to deny the regulatory efforts to harmonize ESG disclosures. Nevertheless, public regulation can only achieve the standardization of structure and framing, rather than details. Requiring all rating firms to adopt the same steps and methods will undermine the value of the rating industry. ESG ratings are based on proprietary research, and maintaining the autonomy of rating practices has two advantages: first, ratings focused on different ESG factors or measuring the same factor from different perspectives are complementary to each other so that they can meet the plural needs of investors; second, it promotes competition and encourages innovations and improvements in rating techniques.

Despite the benefits, divergence is a significant challenge to investors. Disparate and even opposite ratings regarding the same entity or instrument confuse investors. As a result, they exacerbate rather than mitigate information asymmetry,⁹³ leading to the dislocation of financial sources.⁹⁴ When investors cannot make comparisons to distinguish between accurate and inaccurate ratings,⁹⁵ they perceive strong uncertainty in the use of ratings and tend to cast doubt on their validity.⁹⁶ For rating firms, investors’ distrust impedes them from establishing a good reputation.

Lack of transparency

Due to the divergence in ESG ratings, investors may want to access more details of rating practices, such as the criteria for data selection, methodologies, assessment process and internal controls to help ascertain the quality of ratings.⁹⁷ However, as flagged up in empirical surveys, the extent to which the details of rating practices are publicly communicated varies largely among different rating firms.⁹⁸ Investors and regulators criticize that the transparency of rating practices will remain inadequate without regulatory mandates.⁹⁹ As will be discussed in the next subsection, establishing a disclosure scheme for rating firms is the primary focus of the current regulatory proposals.

For most investors, it is not a pragmatic option to estimate the quality of ratings by reviewing and analysing the information about rating practices. The emergence of ESG ratings is the market response to informational inefficiency. If it is very difficult and costly for an investor to assess an entity’s or an instrument’s ESG performance itself, then it will be equally, if not more, difficult and costly for it to assess how others’ assessments are made. For investors who lack analytical resources and ability, more disclosures of rating practices will not help them develop an insightful understanding of the quality of ratings. In addition, disclosures do nothing about mitigating the divergence. More disclosures of divergent data sources and methodologies would exacerbate the difficulty of review and analysis.

When consumers cannot ascertain the product’s quality, as a suboptimal solution, they try to ascertain the quality of disclosures related to the product.¹⁰⁰ It is easier for investors to observe how much information is disclosed. More disclosures reflect a rating firm’s self-confidence in its ratings, which then enhances investors’ confidence. From this perspective, the proposed disclosure scheme may have some merits. However, as will be discussed later, disclosures may be utilized by firms as impression management tools.

Conflicts of interest

A conflict of interest arises in a situation where ‘a party in a transaction can gain by taking actions detrimental to its counterparty’.¹⁰¹ Due to the adoption of an ‘issuer-pay’ model, which means that businesses that are certified or rated pay for the services, underwriters, auditors and CRAs are regarded as having a higher potential for conflicts of interest.¹⁰² Differently, most ESG rating firms charge investors for subscribing to their rating services, known as the ‘subscriber-pay’ model.¹⁰³ Accordingly, some deem that their practices are relatively less prone to conflicts of interest,¹⁰⁴ which is, however, over-optimistic.

In addition to investor-facing services, ESG rating firms also provide certification or consultation services to businesses regarding their ESG credentials or strategies.¹⁰⁵ Conflicts of interest may therefore arise if a firm’s clients use these services but are also rated by it. In the context that external certifications and second-party opinions (SPOs) are increasingly adopted in the issuance of sustainability-linked securities,¹⁰⁶ conflicts of interest may occur more frequently. Issuers prefer compliant and cooperative gatekeepers.¹⁰⁷ When searching providers of certifications or SPOs, an issuer may refer to the ESG ratings it has received from different rating firms and select the one that has given the highest rating. This is similar to the ‘rating shopping’ strategy applied widely in the market of credit ratings, which means that a business solicits ratings from multiple agencies and appoints the one that gives the most favourable rating.¹⁰⁸ Possibly, the real benefit businesses seek by subscribing to ESG rating firms’ certification or consultation services does not lie in these services per se; instead, they aim particularly at spurring rating firms to give them high ESG ratings. In this situation, the actual object of ‘shopping’ is still ratings whereas certification or consultation services are the means only. This rating shopping strategy works even if businesses do not pay for ESG ratings because of two incentives for ESG rating firms: first, for the revenues from certification or consultation services, firms provide higher ratings to retain their clients; second, by giving certified entities or consultees higher ratings, rating firms can use these two different types of services to reciprocally corroborate the validity of each other.

The positive response by ESG rating firms is ‘rating catering’, namely, they provide excessively high ratings to attract clients.¹⁰⁹ According to the lesson of credit ratings, rating shopping and catering will lead to rating inflation.¹¹⁰ At this moment, there is no direct empirical evidence of ESG-related rating shopping or catering, nevertheless, the above theoretical analysis demonstrates the potential for both activities to take place in ESG rating firms. ESG rating inflation implies that an entity or instrument is less sustainable than the rating suggests, which increases the opportunities for greenwashing but retards the detection of it. In this sense, rating firms play a role in aiding greenwashing, which is contrary to the expectation on them as gatekeepers.

Different from credit ratings where rating deflation is less worrying, in ESG ratings, deflation is as severe as inflation. This is primarily due to the unsolicited feature of ESG ratings.¹¹¹ ESG rating firms take the initiative to rate whereby they can inversely solicit businesses to use their certification or consultation services. To attract potential clients, ESG rating firms may rate higher and imply extra privileges to businesses, such as communicating ex ante the provisional ratings in the name of consultation services to allow businesses to influence their own ESG ratings. Alternatively, they may rate lower to threaten non-clients. Businesses that perceive the ‘threat’ of deflation would be more likely to subscribe to their certification or consultation services.¹¹²

The coexistence of rating inflation and deflation magnifies the mismatch between actual ESG performance and ratings. An increasing number of empirical studies find that businesses with higher ESG ratings show worse records than their lower-rated peers.¹¹³

Divergence in ESG ratings is also an aggravating factor. When ratings are substantially disparate, rating shopping becomes more valuable for businesses to either boost their ratings or obtain the most favourable certification or SPO.¹¹⁴ Similarly, rating firms have stronger incentives to undertake solicitation. Assuming a contrary situation where the ratings of the same entity provided by different rating firms are approximate, when one of the firms intentionally deflates or inflates its rating, it is unlikely that the rated entity or investors will be convinced of the accuracy of this abnormal rating.

Conflicts of interest may also arise from the ownership of a rating firm. Recent research demonstrates that ESG rating firms tend to give inflated ratings to sister firms owned by the same large shareholders.¹¹⁵ Recent years have observed a trend of consolidation in this industry, amidst which large firms in financial information services have become the leading players in ESG information by acquiring smaller, specialized firms.¹¹⁶ If consolidation continues and the number of connected companies increases, rating inflation driven by ownership connection may become more frequent.

Divergence in ratings, inadequate transparency of rating practices and the potential for conflicts of interest do not necessarily mean that all ESG ratings are poorly made. However, ESG ratings in this context are more prone to be unreliable and misleading. Where there are conflicts of interest, rating firms have incentives to misrate; shielded with rating divergence and opaqueness, they misrate at low costs. Even if rating firms do not intend to misrate, when it is difficult for the market to ascertain the quality of their ratings, they are inadequately incentivized to improve their rating expertise and skills.

Regulatory proposals

At the global level, the International Organization of Securities Commissions (IOSCO) has put forward general recommendations¹¹⁷ for improving rating practices. Following its recommendations, the Financial Conduct Authority (FCA) in the UK is working on a voluntary code of conduct.¹¹⁸ Being advised by member states and independent investigators, the EU is also preparing a mandatory regulatory scheme.¹¹⁹

Regulators deny the justification for interfering with methodologies.¹²⁰ This implies the importance of maintaining rating firms’ autonomy in proprietary research and their innovative capacity.¹²¹ In the context that divergence in ratings will persist, the solution proposed by regulators against the negative impact of divergence is a comprehensive set of transparency requirements. Specifically, disclosures of data should include the sources of underlying data and the criteria of data estimation where original data are missing. ¹²² Disclosures of methodologies should include the process of data collection, the scope and objective of ratings, the categories, metrics, measurement criteria and weightings of ESG factors, the methods of assessment aggregation, the application of any public or third-party ESG standards or frameworks, and the rules and procedures of regular review and modification.¹²³

Moreover, with an emphasis on avoiding and managing conflicts of interest, rating firms are required to implement internal control mechanisms to ensure the integrity and independence of rating practices, such as a ‘Chinese Wall’¹²⁴ to separate teams and staff undertaking rating-related and other services. Rating firms should then make regular or ad hoc disclosures of the operation of these organizational arrangements and any existing or potential conflicts of interest.¹²⁵

These proposals take a disclosure-based regulatory approach, which, in theory, aims to make investors better informed when selecting and using ESG ratings and thereby discipline rating firms.¹²⁶ However, as mentioned, the effect of disclosures in facilitating investors, especially less sophisticated ones to better examine or estimate the quality of ratings is limited. Furthermore, the disclosure-based approach relies only on imposing external pressures to change rating firms’ conduct rather than restructuring their incentives. As will be argued in the next subsection, disclosures do not help increase the verifiability of the quality of ratings, which cannot sufficiently incentivize rating firms to improve their practices.

Limitations of reputational capital and regulatory proposals

Reputational capital is unworkable

In the absence of regulatory intervention, reputational capital provides the only constraint on misconduct in ESG ratings. However, two defects in the ESG rating industry at this stage frustrate this mechanism.

First, the entire industry of ESG ratings has not yet accumulated enough reputation. ESG information services are not brand new. Several firms were ushered into this field around 15 years ago.¹²⁷ However, before the mainstreaming of sustainable finance, these services were used narrowly for socially or ethically responsible investments.¹²⁸ For the majority of institutional investors, ESG ratings are new services. Since they embed predictions on investment risks and opportunities, investors cannot figure out the degree of rating accuracy or the credibility of rating firms in the short run. It takes time for ESG rating firms to build trust among investors.

This industry has many new entrants. Those who are not ‘repeated players’ do not have a reputation yet. Firms with an entrenched leading position in traditional financial information services and now extending to ESG ratings are also ‘greenhands’. Given that ESG ratings are different and disconnected from financial information services, the spillover effects of their previously accumulated reputations, either good or poor, are limited.¹²⁹ Furthermore, the objects of ESG ratings are largely novel: many businesses are in the transition towards a sustainable business model, and the assessments on them are based on their new ESG achievements or attributes; there are also an increasing number of ESG-labelled investment instruments being created and marketed every year,¹³⁰ in which rating firms have little experience.

Reputational capital operating as an incentive mechanism for accurate ratings and a deterrence mechanism against misconduct is premised essentially on the existence of an established reputation.¹³¹ As time goes on, firms with a long-term tenure in this field can gradually accumulate a good reputation. However, for some firms driven by short-term profit, when there is no reputation to deplete, the opportunity costs for them to behave dishonestly or indolently are low. If rating firms ‘have nothing to lose’, misbehaviours easily take place even if the potential benefits are small. As a corollary, even if an inexperienced firm is fully aware of its incompetence in making ESG ratings, it will still enter this industry to take a share of the profit.¹³²

The second defect is the extremely low verifiability of the quality of ESG ratings. If consumers are unable to verify the quality of a product after purchase or even use, the reputational capital mechanism is ineffective. ESG ratings are predictions by examining complex, multi-dimensional and variable ESG issues. Due to these essential attributes, the verification of their accuracy is unmanageable. Different from the first defect, the negative impact of this issue on the functioning of reputational capital will not ease over time.

Among traditional gatekeepers, the performance of CRAs is considered more difficult to verify as credit ratings are predictions on the probability of default (PD), which can hardly be zero. An AAA rating only suggests that the rated company is less likely to default. The occurrence of a single default event is not enough to prove that the rating is inaccurate.¹³³ A rigorous assessment of the accuracy of credit ratings should be based on the performance of all rated objects over a certain period.¹³⁴ Accounting and finance scholars systematically examine the ratings for all defaulting and non-defaulting obligors to evaluate the accuracy of a rating system.¹³⁵ Correspondingly, the greenwashing scandal of a company with a high ESG rating cannot prove the inaccuracy of this rating. It is also necessary to examine the ESG performance of all rated entities or instruments over a certain period to ascertain whether those with higher ratings have been performing proportionately better than their lower-rated counterparts. However, a systematic and holistic ex post assessment of ESG performance is hardly feasible.

ESG rating firms rate thousands of entities and instruments. Collecting and processing the information about their performance entails a heavy task load and significant costs. An investor who can accomplish these tasks can also directly estimate the ESG risks and opportunities associated with their investees, in which context ESG ratings are unnecessary.¹³⁶ Conversely, the high demand for ESG ratings in reality demonstrates that such a systematic and holistic assessment is beyond the capacity of most investors.

In the case of credit ratings, the inputs are complex, which integrate different indicators of credit risk,¹³⁷ whereas the output is simply the PD (or loss given default¹³⁸). For investors, assessing the accuracy of credit ratings is easier than assessing credit risk. ESG ratings synthesize the assessments on multiple non-financial dimensions. Unlike default events, the intricate outputs of ESG ratings are difficult to observe and measure, and very often financial markets do not have a consensus on how to measure ESG issues, which may lead to equivocal or contested opinions on the accuracy of ESG ratings.

It might be argued that investors do not need to stick to the very precise and rigorous verification approach. As consumers, they can ‘vote by foot’ if they intuitively notice frequent mismatches between the actual ESG performance of rated entities or instruments and their ratings. This argument assumes that investors are wise and wide awake. However, still taking credit ratings as an example, a large number of naive investors took the inflated ratings of CDOs at face value until the financial crisis was triggered.¹³⁹ Nevertheless, naive investors could still realize that credit ratings were problematic after they started to suffer financial losses. Differently, poor ESG performance may not necessarily lead to financial losses. Evidence of the negative or insignificant correlation between ESG and financial performance¹⁴⁰ indicates that financial markets may not always react to ESG controversies correctly or promptly. It is also ironic that greenwashing is a gimmick for profit-making,¹⁴¹ which may bring investors higher financial returns. Investors are only sensitive to changes in their financial returns, whereas most non-financial damages caused by unsustainable activities are in the first place borne by certain stakeholders such as the environment, local community and employees, without a direct influence on investors. These damages will remain invisible to investors until they are ‘financialised’.

In addition, without a systematic and holistic assessment, rating deflation is undetectable. Investors use ESG ratings to filter out underperforming entities or instruments, which means that they completely ignore those below their thresholds. If an entity is underrated, its actual contributions to sustainable development and the inaccurate rating will unlikely be detected.

Disclosure-based regulation is weak

Regulatory intervention should be able to minimize the potential benefits of misrating (the left) and maximize the verifiability of the quality of ratings (the right). As mentioned, regulators are considering a disclosure-based regulatory scheme. Intuitively, more transparency of rating practices will improve verifiability, which is, however, not practical.

It is presumed that the disclosures of ESG data, rating methodologies and procedures allow investors to deduce whether or not the ratings are accurate and reliable. However, investors’ views drawn in this way are estimations, rather than ex post verifications. The ex post verifiability of the quality of ratings depends on the extent to which the actual ESG performance of rated entities or instruments can be systematically observed and measured. The proposed disclosures are irrelevant to improving the ex post observability or measurability of ESG performance.

The underlying rationale for regulators to rely on disclosures is that investors are expected to give more trust to those rating firms making good-quality disclosures¹⁴² so that a ‘modified’ reputational capital mechanism based on the verification of the quality of disclosures will work. If a rating firm’s disclosures are inadequate and ambiguous or expose its flawed rating practice, it will suffer reputation erosion.

Regulators mistakenly equate and misguide investors to equate good-quality disclosures with good-quality ratings. The two may be disparate in two situations. The first is that disclosures are authentic and of good quality. The disclosed contents such as robust data sources and methodologies and an impartial and full-fledged rating process are the necessary conditions for accurate ratings; however, they are not sufficient conditions. Factors beyond these contents, especially those related to discretionary judgements and subjective initiatives, also play a critical role, whereas they cannot be perfectly communicated or visualized. As a result, even if disclosures are thorough and authentic, investors’ evaluations are incompletely informed. The second situation is that disclosures are opportunistically exploited by rating firms only for impression management, which is the tactic to control and manipulate the impressions conveyed to customers.¹⁴³ Impression management has been pervasively observed in corporate accounting and corporate social responsibility (CSR) reporting,¹⁴⁴ and there is no reason to assume that the ESG rating industry will be an exception. Some ESG rating firms may focus primarily on window dressing their rating practices with bombastic disclosure tones, instead of taking actual actions to improve rating expertise and skills. In particular, regulation creates a standardized framework and format, which provides rating firms with clearer and more specific directions about how to milk their disclosures. In nature, this hypocritical communication is equivalent to ‘greenwashing’.

These two situations reflect the inherent imperfections of any disclosure scheme.¹⁴⁵ The first situation is unavoidable as there is always some information staying private, and the second situation—impression management is also ubiquitous. Admittedly, verifying the quality of disclosures is easier than verifying the quality of ESG ratings. Assuming that investors are able to verify the quality of disclosures, with either of the above situations exists, the extent to which they are better informed to estimate the quality of ESG ratings is still limited. The consequence of the second situation is more severe as rating firms are incentivized to focus on catering to disclosure requirements rather than improving their rating practices.

Naive investors who cannot analyse the ESG information of investees also lack the capacity to assess the quality of disclosures by rating firms or see through the disclosures to make estimations. Nevertheless, when a uniform and mandatory disclosure scheme is in place, investors tend to be more confident in the ESG rating industry as the regulator is believed to play a monitoring role and look after their interests. As stated by the FCA, ‘improving trust’ is the primary purpose of the proposed code of conduct.¹⁴⁶ When market confidence increases, there will be more users of rating services, bringing higher revenues to rating firms. However, it is also possible that after perceiving regulatory ‘protection’, some investors become less vigilant. That is, disclosures may lead to the diminished willingness of investors to verify. This consequence will be counterproductive: on the one hand, when the number of optimistic investors increases, the benefits of rating, including those obtained by misrating (the left) are larger; whereas on the other hand, since the ex post verifiability remains low and investors’ verification willingness becomes lower, the costs of misrating, which are reified as the materialized reputation damages (the right), are smaller.

While this article does not intend to totally repudiate the role of disclosure in financial regulation, as far as ESG ratings are concerned, a regulatory framework solely made up of disclosure rules will not function effectively as either an incentive or deterrence mechanism. Considering that the reputational capital mechanism cannot effectively discipline rating practices either, regulatory intervention further than a disclosure scheme is needed.

4. An envisagement for regulatory intervention

Inaccurate ESG ratings may result from an objective factor that a rating firm’s resources and capacity are constrained or a subjective factor that the firm makes intentional or careless rating mistakes.¹⁴⁷ Ideally, regulatory intervention should be able to give rating firms more incentives to break through objective constraints and mitigate their incentives to misrate intentionally or carelessly. Section 3 demonstrates that neither the market-based reputational capital mechanism nor the proposed disclosure-based regulation can sufficiently achieve these two goals. This section envisages more radical regulatory initiatives to engender remarkable changes to the incentive structure of ESG rating firms. It starts with a rigid separation measure against conflicts of interest, which aims to reduce the benefits of misrating. To find a solution to the more difficult issue—increasing the costs of misrating by improving the ex post verifiability of the quality of ratings, the remainder of this section refines the solution proposed for credit ratings—a strict civil liability and designs additional measures; building on these, it proposes a disgorgement scheme for regulating ESG rating firms.

The ‘No-conflict Rating’ principle

A major source of the potential benefits of misrating is the revenues generated from the consultation or certification services provided to rated entities. The solution included in the current regulatory proposals is the installation of organizational separation arrangements and the regular reporting of their operation. ‘Chinese Wall’ mechanisms have been extensively adopted in the securities industry for decades.¹⁴⁸ However, as well addressed in the literature, they are not always successful.¹⁴⁹ In an integrated business structure, the motivation to make use of one business division to benefit another always exists because the benefits are ultimately enjoyed by the firm whereas the downsides are borne by clients. To eliminate this motivation, the firm should be made to internalize the downsides, for which physical separations are insufficient.

A radical solution to conflicts of interest is to decouple the services for counterparties at the firm level. In the case of ESG information services, it means that ratings and certification or consultation services shall not be housed in the same firm. However, with an integrated business structure, a rating firm can apply its experience and skills related to ESG to serve a broader range of market players. Thoroughly decoupling these services will lead to economically negative consequences.

To avoid fully compromising the economies of scale and expertise, this article proposes a ‘no-conflict rating’ principle, which can promote the independence of rating practices within an integrated business structure. Specifically, a rating firm shall be required not to rate the entities and their instruments if they are currently using the firm’s certification or consultation services. This idea is drawn from the EU CRA Regulation, according to which CRAs are not allowed to ‘provide consultancy or advisory services to the rated entity or a related third party … ’.¹⁵⁰ This method creates a trade-off between different services for rating firms to balance: a firm’s expansion in certification or consultation services will result in a narrower coverage of its ESG ratings. In this way, the downsides are internalized. This method can also mitigate the motivation of rated entities for ‘rating shopping’. To maintain the highest ESG ratings visible to investors, they will avoid using the certification or consultation services provided by the same rating firms. Accordingly, rating firms have fewer incentives to undertake ‘rating catering’.

What is not addressed by the prohibitive rule in the CRA Regulation is the conflicts of interest arising from ownership connections.¹⁵¹ The proposed organizational separation arrangements and disclosure requirements have also overlooked this issue. This article suggests that the ‘non-conflict rating’ principle shall also be applied to entities if they are: the significant shareholders of the rating firm, which as per common regulatory practice are defined as those holding 5 per cent or more in ownership, or controlled by the significant shareholders of the rating firm. In the context that financial firms in other areas are flooding into the ESG rating industry, this principle can also help contain the trend of consolidation and maintain a certain level of competition, which will give ESG rating firms the impetus to improve the accuracy of ratings.¹⁵²

Civil liabilities

‘Li’ represents the economic value of compensation should a civil liability be charged, while ‘En’ denotes enforceability,¹⁵⁶ which depends on the specific type of liability adopted in national securities law. A significant difference among different types of civil liabilities is the standard of the mental element. From liability for fraud, negligence liability to strict liability, the standard eases in sequence, from ‘knowingly or recklessly’,¹⁵⁷ ‘lack of reasonable care’ to nothing. For the highest enforceability, strict liability is prioritized. This option has also gained increasing scholarly support in response to the failures of auditors and CRAs.¹⁵⁸

A significant advantage of strict liability is that it relieves courts from ascertaining the intention or recklessness in the liability for fraud or defining the ‘reasonable care’ standard in relation to negligence liability.¹⁵⁹ In any circumstances, gatekeepers are liable for the damages to investors caused by the false or misleading information created by either issuers or themselves.¹⁶⁰ Strict liability is alleged to have the strongest deterrent effect against intentional misbehaviours; therefore, it provides gatekeepers with the optimal level of incentives to exercise precaution and improve their services.¹⁶¹

Nevertheless, opponents are concerned about its over-deterrence effect. The damages to investors may be too significant to allow the liable gatekeeper to survive. A much higher risk of bankruptcy will frustrate gatekeepers and in extreme cases, may lead to the collapse of the entire market.¹⁶² Even if there are gatekeepers staying in the market, they will substantially increase their service fees to compensate for their expected liabilities.¹⁶³ In the case of ESG ratings, the costs will be passed directly onto investors, and non-sophisticated investors, who are most in need of ESG ratings, will be worse off. It is also argued that strict liability is ill-suited for rating services¹⁶⁴ as they embed predictive analysis that can never be perfectly accurate.¹⁶⁵ As an unavoidable consequence, rating firms bear the failures caused by factors out of their control.

Views in favour of charging a liability for gatekeeper failures all assume that their failures are detectable, and the contestation among them focuses on how to figure out a proper level of enforceability. However, according to Formulation (b), the value of ‘Ve’ is another determinant of materialized losses. In terms of ESG ratings, the deterrent effect of any liability scheme, which is subject to the actual risk of being held liable, is diluted by the low verifiability of inaccurate ratings. To an extreme extent, if inaccurate ratings are not detectable at all, even strict liability will be ineffective. In this sense, civil liability is not so different from reputational capital as the functioning of both is dependent on an adequate level of verifiability.¹⁶⁶ Therefore, the focal point of regulatory intervention should be how to improve it.

The enforcement of civil liability, including strict liability, is also premised on several other elements, notably loss causation. In the history of US case law, CRAs have escaped liability because investors could not prove justifiable reliance on ratings.¹⁶⁷ ESG rating firms may also mimic the way CRAs have defended themselves by alleging that the ratings are not recommendations on investment and investors’ decisions are made upon their own knowledge and various information sources.¹⁶⁸ Given that investors usually subscribe to more than one rating service,¹⁶⁹ it is even harder to prove the reliance. Another essential component of loss causation is the losses. As discussed earlier, they must be economic losses, whereas the negative externalities of a business on stakeholders may not immediately lead to a decline in investment returns.

In addition, enforcing civil liability through the judicial system is costly for investors. In a jurisdiction where institutional and market obstacles to securities class actions exist, investors are confronted with the collective action problem, which is another fatal defect of the application of civil liability.

Compared to traditional gatekeepers, applying strict liability that is framed by existing law to ESG rating firms faces more obstacles. That said, an inquiry into the scholarly views on the options for gatekeeper liability is still inspiring. On the one hand, strict civil liability overwhelms the other two liabilities as it most broadly applies to all types of inaccurate ratings with the least requirements of evidence; however, on the other hand, resorting to the court may be impractical for ESG ratings. Regulators may take measures that assemble the mechanics of strict liability so that its advantages are maintained and the obstacles to lawsuits are dodged.

A disgorgement scheme

Two significant tasks are left to regulators: improving the ex post verifiability of the quality of ratings and designing an alternative mechanism of ‘skin in the game’ to strict liability. This subsection depicts the detailed design of the disgorgement scheme, which integrates the recommendations for regulators to achieve both tasks.

Compared to the market, regulators do not have any technical or expertise advantage to ex post verify the accuracy of ESG ratings. The only solution is operationalization, namely, using appropriate factors which are much easier to observe and measure as the proxies of ‘inaccurate ESG ratings’. For credit ratings, Pacces and Romano have proposed to use the default of any rated entity or instrument as the proxy.¹⁷⁰ In their proposal for a (modified) strict liability scheme, CRAs are liable for every single default. Thanks to the sole output of credit ratings, the operationalization is straightforward. In contrast, ESG ratings integrate a variety of dimensions, and different rating systems use different combinations of these dimensions. As discussed, this article deems that a reasonable level of divergence is conducive to the competitiveness of the ESG rating industry. For this reason, regulators coercively defining ‘one-size-fits-all’ proxies is unreasonable. Instead, as will be addressed below, proxies can be defined by rating firms at their discretion.

Inspired by the idea of ‘self-tailored liability’ proposed by Professor Choi, which means that gatekeepers are allowed to select the type and degree of liability by which they are bounded,¹⁷¹ this article envisages a contract-based disgorgement scheme for ESG rating firms. The general principle of this scheme is that rating firms are required to have a stake in the (negative) consequences of their ratings. Specifically, a rating firm must disgorge a portion of its revenues if certain trigger conditions materialize. The two essential components of disgorgement—the specific portion of disgorged revenues and the trigger conditions are pre-determined by the rating firm on a ‘self-tailored’ basis.

The trigger conditions are just the proxies of ‘inaccurate ESG ratings’. Akin to the proxy of inaccurate credit ratings, these trigger conditions may be reified as the adverse ESG events taking place in any rated entity or instrument. Rating firms shall be required to specify unequivocally what the adverse ESG events are. For instance, if a firm’s ESG ratings include the dimension of ‘carbon emissions’, it may define the proxy as the rated entity’s failure to achieve its net-zero targets without delay or its failure to comply with mandatory legal or regulatory requirements on emissions; if there is a ‘health and safety’ dimension, the proxy can be the occurrence of any fatal injury in the workplace during a pre-designated period or any fine due to the breach of health and safety law and regulation. As rating firms often use specific metrics to assess ESG risks and opportunities, the proxy of a dimension may be reified as a certain level of the corresponding metric.

Self-defined proxies also reflect the degree of rating accuracy a firm expects to achieve. Usually, if a proxy event has a higher probability of occurrence, it indicates that the firm is more confident in the accuracy of its ratings. Following the example of health and safety, in some industries, fatal injuries are uncommon whereas fines for the breach of law and regulation are relatively more frequent. In this context, a rating firm selecting the latter as the proxy conveys more confidence.

ESG ratings include several heterogeneous dimensions, and firms may specify a proxy for each of them. All proxies need to be made available to investors prior to subscription and incorporated in the service contract. In addition to proxies, rating firms shall also specify the portion of disgorged revenues. Revenues can be simply represented by total annual subscription fees. This article suggests that firms can freely designate how much portion they are willing to disgorge (ie disgorgement ratio), which literally ranges from 0 to 100 per cent.¹⁷² Choosing zero means that the firm does not want to bear any liability for misrating, implying that it has no confidence at all to assure investors of the quality of ratings.

As an incentive and deterrence mechanism, this disgorgement scheme has a number of advantages over other options. First and foremost, it provides an alternative, feasible solution to the problem of very low verifiability. By using proxies, it makes verification much easier and less costly. Although proxy events are not the real evidence of inaccurate ratings, they represent the poor ESG performance of some rated entities or instruments and to some degree reveal allocative inefficiency. In the field of sustainable finance, inefficient allocation of capital engenders not only economic but also social and ethical costs. ESG rating firms making a disgorgement out of their revenues is not a liability for the mistakes of third parties (rated companies or issuers), but a share in the compensation for these extra costs.

Second, the scheme ensures that rating firms have ‘skin in the game’ so that they are both held back from misrating intentionally and encouraged to improve the accuracy of their ratings as much as possible. These two objectives are achieved through the competition-enhancing function of the scheme. Firms have the freedom to determine proxies and their disgorgement ratios. While in theory a firm can define an event that is unlikely to occur as the proxy and set the ratio at zero, in practice a reasonable investor will not choose to use such ratings. Therefore, a ‘race-to-the-bottom’ is not a concern. In contrast, to convey sufficient self-confidence in their ratings to investors, firms have incentives to compete with each other by applying proxy events with a higher probability of occurrence and setting the disgorgement ratio at a higher level. Nevertheless, the competition-enhancing function is only effective when there is competition in this industry. In an oligopoly, rating firms can quickly ally to fix the probability of occurrence and the ratio at low levels. Currently, the ESG rating industry is not an oligopoly,¹⁷³ which suggests that the disgorgement scheme is suitable.

Third, in terms of economic function, this scheme assembles an insurance contract in some ways. Rating firms offer different coverages of insured risks and insured amounts to meet different preferences of investors. Compared to examining the disclosures of methodologies and the assessment process, understanding proxy events and disgorgement ratios is quicker and easier for investors. Of course, this advantage is premised on rating firms providing unequivocal and precise definitions and interpretations. As will be discussed later, regulators may play a part to ensure the preciseness of relevant provisions.

Fourth, different from civil liability, this scheme is contractual. Disgorgement simply becomes enforceable upon the occurrence of a proxy event. There is no need for investors to prove their actual losses or loss causation. Investors do not need to file lawsuits, which avoids the collective action problem and high costs, as well as reducing the workload for courts.

Despite these advantages, a disgorgement scheme may be distorted in practice, to prevent which regulators should step in to guide its implementation.

Rating firms do not have the motivation to proactively initiate this disgorgement scheme, otherwise, we would have observed one in reality. Regulators need to mandate the application and specification of a disgorgement scheme as a prerequisite for undertaking ESG ratings. As stated, competition in this industry is important for implementing the disgorgement scheme. When there is a range of different rating firms and market entry is not barred, new entrants and smaller participants will tend to apply a more radical disgorgement scheme to enhance their competitiveness. Regulators should avoid measures that may accelerate consolidation or jeopardize competition, such as enhancing entry barriers.

Rating firms have information and expertise advantages over investors. To mitigate the chance that they manoeuvre the definitions and descriptions of proxy events with paradoxical and puzzling language, regulators shall step in to specify the generic key elements for defining a proxy, set the sample models and formats, and make orders of correction to any equivocal statement.

So far, the envisaged proxies are focused only on adverse ESG events. That is, the disgorgement only deals with rating inflation. It is almost impossible to find reasonable proxies of inaccurately deflated ratings. Opposed to adverse ESG events are ESG outperformance, such as realizing the net-zero targets in advance or the absence of any breach of law and regulation. However, it does not make sense to use them as the proxies of deflated ratings. In addition, investors tend to rely primarily on adverse ESG events to select rating services. They are more vigilant about adverse ESG events which signify potential economic or non-economic losses. On the contrary, the ESG outperformance of an entity or instrument may be regarded as the signal of correct investment decisions by existing investors or promising opportunities by others.

With only adverse ESG events as proxies, rating firms will become conservative in ratings to evade disgorgement. They may exclude entities or instruments with poorer ESG performance or higher ESG risks from their ratings to reduce the occurrence frequency of proxy events. Of course, in so doing, firms also take a risk as ratings with a narrower coverage are less competitive. Moreover, if the disgorgement scheme is purely contractual, rating firms can confine the application of disgorgement to upper grades, for example, BBB and above, in which context they will be inclined to an overall rating deflation. To deal with this problem, regulators may impose an ad hoc prohibition on partial disgorgement.

To ensure that the disgorgement scheme also applies to rating deflation, this article suggests a different approach to defining the proxies of deflated ratings. In fact, rated entities are the ‘victims’ of deflated ratings, who have strong enough incentives to act against them. If rated entities have reasonable grounds to believe that their ratings are lower than they should have been, they should be given a chance to voice this. Regulators may therefore operate a panel of independent adjudicators with expertise in this industry to deal with any disagreements between rated entities and rating firms. A deflated rating confirmed by the panel shall be defined as the proxy of inaccurate ratings and therefore would trigger disgorgement.

Accordingly, the trigger conditions of disgorgement are made up of self-tailored proxies of rating inflation and the regulator-adjudicated proxy of rating deflation. In either circumstance, the self-tailored disgorgement ratio applies. It is worth noting that the panel model does not work efficiently for rating inflation. Applying the panel model to rating inflation means that investors need to file their disagreements. Unlike rated entities who only need to care about their own ratings, investors are not able to monitor thousands of ratings, and they are still subject to the collective action constraint.

Last but not least, another key issue is who should be entitled to disgorged revenues. The answer is not investors. The primary purpose of the disgorgement scheme, as stated, is to deter rating firms from misrating and give them sufficient incentives to improve the accuracy of ratings. Although investors do benefit from more accurate ratings, the most important purpose is to facilitate the efficient allocation of capital to sustainable activities. In this sense, by issuing inaccurate ratings, rating firms put sustainable finance at risk. The disgorged revenues, therefore, should be used to compensate for this risk. Regulators, as the public authority, can collect disgorgements and reasonably use the money to support sustainable development.

5. Conclusion

This article makes three contributions to the scholarly discussions and future policy making on ESG ratings and their providers.

First, it identifies ESG rating firms as the gatekeepers of sustainable finance and justifies this identity by analysing the verification role and guiding role embedded in ESG ratings. This theoretical perspective confirms their significance in sustainable finance and lays the foundation for understanding the market failures they are subject to.

Second, this article highlights that carrying out ex post verifications of the accuracy of ESG ratings is very difficult, which creates more incentives for ESG rating firms to misrate. For this reason, market monitoring through the reputational capital mechanism will fail. The article also contends that disclosure-based regulation will not work because it does not help with improving the ex post verifiability of the quality of ratings. Instead, it may distortedly incentivize rating firms to focus only on impression management through disclosures. These views alert regulators to think further about the purpose of regulatory intervention in ESG rating practices and how to intervene.

Third, the most significant contribution of the article is the disgorgement scheme it introduces. As part of the scheme, the article suggests an alternative solution to the low ex post verifiability, which is to use appropriate events as the proxies of inaccurate ratings. At the core of the disgorgement scheme is the idea to require rating firms to make compensations for the negative consequences of their ratings and bear a share of allocative inefficiency upon the occurrence of proxy events. In so doing, it is aimed at facilitating regulators to prevent rating firms from misrating and incentivize them to improve the quality of ratings as much as possible.

Acknowledgement

I am extremely grateful to the anonymous reviewers for their insightful comments. I would like to thank Prof. David Cabrelli for his helpful suggestions on an earlier version of this article, as well as Dr Simone Lamont-Black and Prof. Christopher K. Odinet for useful discussions. All errors are my own.

© The Author(s) (2024). Published by Oxford University Press.