Abstract
Have EU regulations implemented after the Global Financial Crisis (GFC) on credit rating agencies (CRAs) been effective in curbing rating inflation? We answer this question by analysing a large dataset of asset-backed securities (ABS) issued between 1998-2018. We find that new regulations reduced rating catering. However, rating shopping remains a concern, especially for investors of premium ABS. Overall, our findings show that the effectiveness of the regulatory changes has been limited.
he 2007-08 Global Financial Crisis (GFC) exposed shortcomings in credit ratings, particularly for complex financial instruments such as asset-backed securities (ABS). The intricate structure and opaque nature of ABS makes it difficult for investors to conduct thorough due diligence, making ratings a vital source of information. Prior to the GFC, ratings issued by credit rating agencies (CRAs) were the primary factor guiding investors’ ABS risk assessment. Investors heavily relied on ratings, and often without a thorough understanding of the underlying risk. However, CRAs failed to accurately assess ABS risks in the years leading up to the GFC. It is now well documented that during this period there was a progressive decline in rating quality and the issuance of overly optimistic ratings, i.e. rating inflation.
Rating inflation, a phenomenon in which average rating increases and rating variance across listings decreases, was commonly observed in the pre-GFC period. It threatens the informativeness of ratings. A main cause of rating inflation is the existence of conflict of interest between CRAs and their clients (i.e. issuers) due to close business cooperation. In addition, competition in structured financial markets inflicted pressure on CRAs to award inflated ratings in order to win over customers. However, such competition often diminishes rating quality in a so called ‘race to the bottom’. It is observed that CRAs are more likely to facilitate rating favours during financial market boom periods, when potential reputational damage is lower and when investors are generally more trusting. Such environment helps banks to obtain the desired ratings for their ABS issue.
Rating inflation often manifests itself in the form of rating shopping1 and rating catering2. Rating shopping behaviour is motivated by issuers’ desire to disclose ratings selectively. By approaching multiple CRAs, issuers can choose to publish the highest rating while suppressing the less favourable ones. For example, if an ABS tranche has only one rating, it is more likely that rating shopping may have occurred in comparison to a tranche with two or more ratings from different CRAs. Put differently, the more ratings available, the less chance there is for selective disclosure. Figure 1 shows the distribution of number of ratings over time in our sample.
Rating catering is as an extension of rating shopping. When an issuer shops for favourable ratings, CRAs may be willing to cater the demands of the issuer due to a desire for a larger market share. One way to identify potential rating catering is to look at the level of rating agreement, where multiple CRAs assign the same rating to a single instrument. Conversely, rating disagreement, where a tranche receives multiple non-identical ratings, could indicate the absence of it. Such disagreement, first, could indicate that issuers are not suppressing unfavourable ratings, signalling transparency in reporting of all ratings. Second, it could also demonstrate that CRAs did not intentionally loosen their standards to match competitors’ ratings. Empirical evidence shows that both behaviours were prevalent for ABS in the pre-GFC period.
Figure 1. Number of ratings per ABS tranche
Note: The figure shows the distribution of number of ratings attached per tranche issued between 1998 and 2018. We observe an increase in the dual-rated tranches after the introduction of new rules. For example, in 2018, 80% of all issues were dual rated.
In Karimov et al. (2024), we examine whether behaviours leading to rating inflation have changed in the post-GFC period when EU policy makers introduce regulations to tackle it. The new CRA regulation was introduced in three stages, with the first implemented in 2009, focusing on reducing conflicts of interest and improving rating methodologies. In 2011, in addition to the creation of European Securities and Markets Authority (ESMA), a regulatory and supervisory body for CRAs, was introduced. Finally, in 2013, amendments and an additional set of measures (i.e. CRA III) were introduced to improve transparency and reduce rating overreliance.
In particular, the new regulation required issuers of ABS to obtain credit ratings from at least two independent CRAs and make these public. This measure aims to reduce overreliance on a single rating, increase competition in the rating industry, and, overall, improve rating quality. More ratings should help reduce information asymmetries for investors, as each additional rating provides them with more data on the creditworthiness of ABS. New rules also puts more pressure on CRAs to improve the quality of their ratings process. Relatedly, it can help reduce rating shopping as issuers would have to disclose at least two ratings.
However, critics argued that increasing the minimum number of required ratings may not be effective as many issuances are already dual rated3. Furthermore, the additional cost of obtaining a second rating potentially burdens smaller issuers. Enforcement may also present a challenge as CRA III places the responsibility for enforcing these provisions on the relevant authorities within each member state. Therefore, the effectiveness of the measure hinges on individual countries ability in establishing clear penalties for non-compliance.
We capture the risks of rating inflation in the initial issue price of 12,469 ABS tranches. Previous studies show that the information content of the initial market spreads4 of ABS instruments are valuable. It is also argued that credit ratings are the single most determinant of these spreads. However, empirical evidence shows that, when evaluating ABS, investors also consider additional factors such as the external credit enhancement, collateral quality, the possibility of rating inflation (via rating shopping and catering), issue size and issuer reputation. Overall, ABS issue price is responsive to all available data. Given that initial yield spreads reflect all risks considered by investors, we utilise the existence of multiple ratings and rating agreements between CRAs to capture the potential existence of rating inflation.
We find that regulatory changes have been effective in tackling the conflict of interest between issuers and CRAs in the ABS market. Rating catering, which is a direct consequence of issuer and CRA collusion, seems to have disappeared. Investors who demanded higher spreads previously for rating agreements did not consider the effect of rating harmony as a risk in the post-GFC period.
Regarding rating shopping, we find that the effectiveness of the changes has been limited, potentially for two reasons. Firstly, rating shopping could also be an innate cause of rating processing and issuing procedure, as it is at issuers’ discretion to report or suppress ratings. Secondly, the newly introduced at least two ratings rule, a regulation that has a direct impact on rating shopping behaviour, may not be enough to reduce rating shopping, as ABS can also be rated by three or more CRAs. Additionally, we find that rating overreliance may still be an issue, especially for investors of high-quality (i.e. triple-A rate) ABS.
Overall, our findings suggest that investors’ perceptions of potential conflict of interest between CRAs and ABS issuers have changed in the post-GFC period after the introduction of new regulations. Also, we find that European initiatives have been effective in reducing rating inflation, as rating catering is not reflected in the prices of the ABS. As for rating shopping and rating overreliance, we propose that further set of measures may be needed as these undesired behaviours appear to be still present in the market.
Bolton, P., Freixas, X., & Shapiro, J. (2012). The credit ratings game. Journal of Finance, 67(1), 85-111.
Griffin, J.M., Nickerson, J., & Tang, D.Y. (2013). Rating shopping or catering? an examination of the response to competitive pressure for CDO credit ratings. Review of Financial Studies, 26(9), 2270-2310.
He, J.J., Qian, J.Q., & Strahan, P.E. (2012). Are All Ratings Created Equal? The Impact of Issuer Size on the Pricing of Mortgage-Backed Securities. The Journal of Finance, 67(6), 2097-2137.
Karimov, N., Kara, A., Downing, G., & Marques-Ibanez, D. (2024). The impact of regulatory changes on rating behaviour. European Central Bank, Working Paper Series, No. 2920.
Skreta, V., & Veldkamp, L. (2009). Ratings shopping and asset complexity: A theory of ratings inflation. Journal of Monetary Economics, 56(5), 678-695.
Skreta and Veldkamp (2009).
Bolton et al. (2012); He et al., (2012); Griffin et al. (2013). Rating catering is a broad term and it can involve rating shopping. Here, we restrict its meaning to cases where ratings reported for ABS tranches are identical.
This is because such requirement is already applied by the European Central Bank (ECB) in qualifying ABS securities for their use in the Eurosystem credit operations.
Initial market spread or spread is the compensation margin over relevant benchmark for the risks of the related structured security. The spread, set in basis points, is determined at an auction upon issuer’s release of the ratings obtained for the bonds.
