Banking Crisis: reforming corporate governance and pay in the City - Treasury Contents

5  Credit Rating Agencies

The role of credit rating agencies in the banking crisis

180. In our previous report on Banking Crisis: dealing with the failure of UK banks we described how the use of securitisation, a measure intended to mitigate risk, actually compounded it. There are parallels here with the role of credit agencies. Credit ratings issued by credit rating agencies (CRAs) appeared to offer an objective measurement of a financial institution or financial product. Over-reliance on that apparently objective measure played a key role in the banking crisis.

181. The credit ratings market is dominated by three main players: Standard & Poors, Fitch Ratings and Moody's Investor Services. CRAs formulate and issue credit ratings on both institutions and individual debt instruments. Purchasers of debt or debt instruments, as well as other participants in the market, may then rely on these ratings as indicators of the credit risk of investment. For any responsibility for the banking crisis to be laid at the feet of the CRAs, two premises must be proven. First that market participants relied upon the ratings issued by the CRAs and second that the ratings issued by the CRAs were in some way deficient. It is these two premises which we will consider in this section before moving on to assess the prospects for reform.

182. Stephen Hester, Group Chief Executive of RBS, told us that "The world needs some shorthand of credit analysis because many market participants who use financial markets do not have the resources and time and expertise to do the work themselves".[284] What Mr Hester's comment perhaps reflects is the reality that many people place at least some reliance on credit ratings: if there was no such reliance then there would be no need for ratings and the CRAs would wither away. The arguments therefore focus on the degree of reliance. Lord Turner was one of those who told us that there had been an over-reliance on ratings:

there was also a problem of people relying on [credit ratings] to tell them something about the liquidity or the value rather than the default. One should never have assumed that a credit rating told you anything about the market value of the instrument, it was only meant to tell you about the probability of default. [285]

183. This is a view echoed by the CRAs themselves. Stephen Joynt, President and Chief Executive Officer of Fitch Ratings, accepted that some "investors may have relied on ratings to be all-encompassing, not reflecting as carefully on the risks that they did not cover".[286] The IMA were amongst those who were clear about the limited scope of a rating, asserting that ratings "do not seek to tell the investor anything about the value or liquidity of an instrument, but simply the probability of default and the potential loss in the event of default".[287]

184. Given the apparent consensus that there has been over-reliance, the question becomes who has been exercising it? Mr Joynt suggested that it was not large institutional investors. Such investors were, he argued, "sophisticated" and used the CRAs' ratings to complement their own analysis, relying upon the CRAs "only to a limited degree".[288] This view was echoed by others, including the IMA who informed us that ratings were just one of the opinions their members would assess.[289] Stephen Hester, Group Chief Executive of RBS, provided the banks' perspective as a user of ratings, confirming that banks had their own resources which were capable of "second-guessing rating agencies".[290]

185. Under the Basel II agreement, credit ratings from approved CRAs, known as External Credit Assessment Institutes, can be used when calculating net capital reserve regulatory requirements. The principle is that the more a financial institution is heavily invested in highly liquid and low risk securities, the less it needs to hold as capital in reserves. While credit ratings therefore play a part in the regulatory system, witnesses have assured us that regulators are not over-reliant on ratings. Mr Joynt offered us some reassurance, telling us that the "regulators—bank regulators certainly—we meet with often are sophisticated users of ratings and understand their limitations".[291] This was verified by Barry Hancock, Managing Director of Standard & Poor's:

the UK FSA will spend a lot of time with us to get behind what the ratings are actually saying and, probably as important as that, to understand our thinking and our research on the various sectors that we are rating. I would think they are certainly expert users.[292]

186. While witnesses did not suggest that the FSA was unsophisticated in its use of ratings, they did raise questions about the role played by ratings in the regulatory system. Indeed the FSA's Chief Executive, Hector Sants, told us that credit ratings "have become very deeply embedded in the regulatory architecture, so when they change they have knock-on effects across the board in terms of the way companies can fund themselves".[293] In the Turner Report, the FSA made clear its conviction that regulatory change was required in order to improve the "governance and conduct of rating agencies and the management of conflict of interest".[294]

187. Given that CRAs accept that some may be over-reliant on their work, but deny that this is the case in respect of either institutional investors or regulators, it is reasonable to infer that they are suggesting that it is the smaller investors who are falling into this trap. Indeed, Mr Hester acknowledged that many people did not have the resources available to the larger investors which would enable them to dig deeper into the ratings.[295]

188. It is worrying that markets appear to have used credit ratings for more than they were designed to do. Their primary role should be to provide an outside opinion on the credit risk of a product, not as a potential guide to its overall market price or liquidity risk. This 'overuse' may have been more prevalent among smaller investors, though our suspicion is that they had become a convenient short-cut for more experienced market professionals as an alternative to their own due diligence.

Quality of ratings

189. The message communicated by the CRAs has consistently been that they are but one group of contributors to the vast pool of information available to market participants during the build up to this crisis. This image of the CRAs as simply one amongst equals has been disputed by others who argue that CRAs' ratings played a central role in exacerbating the crisis. The charge levelled against CRAs is that they misled market participants by failing to reflect the difficulties faced by the banks in their ratings early enough. Richard Lambert, Director General of the CBI, told us that "clearly credit ratings agencies were part of the process of turning … stone into gold or whatever the right cliché is - alchemy".[296] Professor Buiter, of the LSE, argued that credit rating agencies should bear some responsibility:

The credit ratings agencies of course and those who used them are culpable. The credit rating agencies got into a line of business that they did not understand. They were reasonable at rating sovereign risk and large corporates but not at rating complex structures, but they did it anyway.[297]

190. A number of the submissions we received appealed to this distinction between corporate credit ratings and ratings given to structured finance products. Corporate credit ratings are assigned to companies and other entities who issue debt. These ratings represent an assessment of the balance of business risk and financial risk of the entity. In formulating them, CRAs can draw on industry-specific knowledge and company information gained over a number of years. In contrast, structured finance credit ratings are assigned to pools of assets which have been grouped together. These ratings represent a company's judgement on the likely total level of defaults or losses within a given pool and incorporate assumptions regarding the risk diversification benefit offered by the particular combination of assets.

191. In its written submission, the IMA argued that ratings of structured financial instruments relied on mathematical modelling which "clearly proved to be flawed, in part because they had a limited historical performance record".[298] Lord Turner and the Governor of the Bank of England are among those who have argued that the credit ratings assigned to structured financial instruments were less reliable then those assigned to corporations. The Governor told us that the CRAs had moved "into areas where they did not have appropriate expertise"[299] while Lord Turner accepted that the CRAs had a reasonable record when it came to corporate rating but that their work on complex financial instruments was of a lower quality:

What happened over the last five years was that first of all they went way beyond that single company or single bank credit rating and they began to credit rate all of these complicated things that we were talking about earlier, like CDOs. It turned out that these things were much more tricky to work out what their default characteristics were, or their value was much more tricky, that they were subject to much more rapid change in ratings than the single company or single bank ratings.[300]

192. The CRAs we heard from argued that all of their ratings were robust. With regard to corporate ratings Michel Madelain, Executive Vice-President of Moody's, was adamant that the ratings issued had been proven to be accurate:

if I look at financial institutions, we rate about 2500 of those and if I look back over the last two years we had 18 defaults and 35 institutions that moved from investment grade to non-investment grade. I would like just to put those facts in front of you just to give some balance to the perception that may have been created by the sub-prime events in the United States.[301]

193. In respect of ratings of complex instruments Ian Bell, European Structured Finance Ratings Managing Director at Standard & Poors, "absolutely" rejected the suggestion that agencies were ill-equipped to rate complex financial instruments. He argued that the staff had "decades of experience" of complex products and denied that any instruments had been too complex to rate.[302] He did, however, acknowledge the "inherent uncertainty" involved in any assumption regarding the future course of events but forcefully refused to accept that the structures the CRAs were rating were inherently too complex.[303] The only criticism which Mr Bell was willing to accept was that some of the assumptions made by CRAs had turned out to be mistaken:

We have all kind of acknowledged—certainly at S&P and I think my colleagues from the other firms would not disagree—that when you look at the ratings of the US sub-prime residential mortgage-backed securities and some of the CDOs that were backed out of the sub-prime, undoubtedly the assumptions we made about how those would perform in the future turned out not to be correct. There is no doubt about that.[304]

194. Even if the CRAs were not as accurate a predictor as they might have been it is only fair to acknowledge the fact they were not alone as Mr Hester of RBS told us "we [the banks], along with many others, made the same errors the rating agencies did in respect of these securities".[305]

195. We cannot accept the credit ratings agencies' argument that they were well-equipped to rate the complex financial products that marked the recent period of exuberant market expansion. History has proved otherwise. Ratings agencies, whether they like it or not, are significant market participants, and their decision to rate a product is an important step to ensuring a market develops in a product. Credit ratings agencies should ensure a greater lead time before rating new products, so that the default characteristics of such products can more assuredly be measured, and therefore commented upon.

Prospects for reform of credit rating agencies

196. The evidence we received suggested that there is an appetite for increased regulation of CRAs. Hector Sants, for the FSA, told us that "there is no question: we support extra regulation of credit rating agencies".[306] The Chancellor also expressed his support for further regulation:

The recognition is that [CRAs] do need to be regulated, they are pretty important … I have always believed that a board of a bank should rely on a credit rating agency for an assessment but it should not substitute the credit rating agency's judgment for its own, and I rather think that happened on too many occasions. There are other issues like conflict of interest with credit rating agencies … this is an area where we need to change the way in which we have done things.[307]

Some were more cautious. Andrew Crockett argued it was hard to see how regulation could "materially improve the performance of rating agencies" though he conceded that reforms "could reduce reliance on fallible judgments". [308]

197. There are already moves under way at the European level to introduce further regulation of rating agencies. In November 2008 the European Commission adopted a proposal to regulate CRAs which is likely to be considered by the European Parliament in April 2009. Key aspects of these regulations include requirements that CRAs:

·  may not provide consultancy or advisory services;

·  must disclose the models, methodologies and key assumptions on which they base their ratings and must publish an annual transparency report;

·  must differentiate ratings of structured finance instruments and corporate debt, either by using different rating categories or by publishing an additional report alongside the rating;

·  must create an internal function to review the quality of their ratings;

·  should have at least three independent directors on their boards whose remuneration cannot depend on the business performance of the rating agency. At least one of them should be an expert in securitization and structured finance; and

·  must be registered via the Committee of European Securities Regulators (CESR), although responsibility for registration and supervision will be performed by the national regulator.[309]

198. The FSA has made public its support of this proposal but notes the importance of such action being replicated at a global level.[310] Our consideration of the prospect of regulatory reform of CRAs will focus on the issues of conflicts of interest, over-reliance and global coordination.

Conflicts of interest

199. In its submission to us the ABI suggested that the relationship between CRAs and their clients can appear "symbiotic".[311] As this word suggests, the dependencies between CRAs and their clients are two-way. Self-evidently CRAs rely on their clients for income but the debt-issuers also rely on the CRAs. Sir Fred Goodwin, the former Chief Executive of RBS, told us that without a rating, or multiple ratings, institutions would find the sale of significant volumes of debt very difficult indeed:

[Banks] could sit and produce this securitised product until we were blue in the face and attach our own ratings to it, but we would not sell any of it to investors because investors wanted the independent rating and the ratings agencies were paid to rate these.[312]

200. CRAs now operate an 'issuer pays' business model. This means that the issuer of debt or a financial instrument, as opposed to the user of the rating, pays the CRA for producing the ratings. The conflict of interest generated by this model can be characterised in a number of ways. Undeniably CRAs have an interest in maintaining the income stream of fees from the issuers. Arguably, rating an issuer's product increases the likelihood of an issuance being successful and therefore of the issuer continuing to thrive and therefore of future issuances with their associated fee payments taking place in the future. A more short-term characterisation of the conflict might be that clients may be more inclined to commission the services of a CRA whose reports appear flattering.

201. There are those who argue that the 'issuer pays' model should simply be abandoned because of the scale of these conflicts.[313] Whenever we have taken evidence from CRAs they have always acknowledged the existence of a conflict. Their defence, articulated by Michel Madelain of Moody's, is that any payment model for credit ratings generates conflicts of interest and therefore attention should be focused on controlling the existing conflict rather than changing the model:

if the issue is about conflict there is conflict with every model. I am sure you are fully aware of the conflict inherent to an investor-pay model: people who short bonds or short stock have a keen interest in seeing a very aggressive downgrade; people who hold bonds have a very key stake in keeping the ratings stable, so nobody is out of conflict … The key to us is that we believe that the current model is the best model but it needs to be managed effectively.[314]

202. As Mr Madelain intimated, the most frequently quoted alternative to the issuer-pays model is the investor-pays model. Aside from the conflicts in this model noted by Mr Madelain, Ian Bell, for Standard and Poors, argued that the model presented a major obstacle to transparency:

the circle that will not square with the investor-pays model is the transparency … it is difficult to see how that model works when anybody can get for free what another investor has to pay for.[315]

Continuing with this theme, Mr Bell asserted that a change in model would mean a loss of one of the key benefits of the issuer pays model, namely that CRAs can "put the ratings and the explanations and the analysis out for free for everybody to see—potential investors, putative investors, regulators, policy-makers".[316]

203. While the CRAs oppose a change in the model they fully accept that conflicts exist. Mr Madelain's solution was that "regulation must take the place of any prospective change in payment model: regulation must offer the "extra level of comfort" regarding the independence of credit ratings".[317] He cautiously welcomed further regulation, suggesting that the regulation being contemplated by the European Commission would "effectively provide further assurance to the investors and to the other stakeholders that we are managing properly the conflict [of interest]".[318]

204. We remain deeply concerned by the conflict of interests faced by credit rating agencies, and have seen little evidence of the industry tackling the problems highlighted in our report on Financial Stability and Transparency with any sense of urgency. We do, however, recognise that there are conflicts inherent in every payment model. It is our view that transparency offers the best available defence against conflicts of interest and we recommend that CRAs publicise more widely the safeguards they have in place to mitigate the risks posed by conflicting interests. It remains that case that, with the major rating agencies being US based, global coordination of regulatory efforts in this area is required.

Combating over-reliance; the role of transparency

205. We have already acknowledged the problem of over-reliance on credit ratings. Reducing this reliance means changing behaviour. We will now assess the prospects for changing this particular behaviour.

206. The reduction of reliance by major institutions may result from explicit internal policy decisions made in the light of this crisis. For example, the Governor of the Bank of England told us that the Bank was going to reassess its reliance on ratings:

in the Bank I think we are going to think very hard about the extent to which we and other central banks rely on credit ratings as an indicator of access to our operations. We do not allow anyone with a given credit rating automatic access, we always do a second check, but there is a good deal to be said for downplaying the role of credit ratings in its entirety.[319]

Lord Turner argued that the problem of over-reliance would naturally dissipate:

80% or 90% of what will now happen to make us correctly use credit ratings is more likely to come because the market will just realise it used them in an inappropriate fashion and a relatively small amount may come from the regulation. That does not mean that the regulation is not important, it just means it is important for us to understand where the big impact is.[320]

207. Alongside Lord Turner's view that there would be a natural reduction in reliance on ratings, there are those who call for more proactive measures. One possible way to decrease over-reliance is to increase investors' knowledge about credit ratings. More information about the scope of ratings, how they are formulated and what information they use may better equip individuals to make their own judgments about the quality of a rating. Michel Madelain of Moody's, recognised the need for greater transparency:

we view it as essential that the level of transparency that we have today on the corporate ratings side is available on the structured finance side, which means that we need to have a level of information available to all market participants which is equivalent to what we see when we rate corporate or when we rate local authority or government or financial institutions. [321]

He suggested that information currently available only to CRAs and "a select group of market participants" should be more widely available.[322] This call was supported by the ABI, who argued that the information available to CRAs should be available to all.[323]

208. One potential downside of increased transparency is the increased risk of 'gaming' of ratings whereby institutions design their instruments around the published rating criteria. Lord Turner suggested that, even under the current arrangements, "there was a process by which people were, as it were, designing these structures to get a particular credit rating".[324]

209. As well as transparency, others have discussed the role of the regulator as a trend-setter. As Mr Sants told us, ratings are "very deeply embedded in the regulatory architecture".[325] Professor Buiter argued that ratings agencies should be stripped of their "quasi-official regulatory role" while the ABI warned that regulation itself could contribute to the problem of over-reliance if it is seen as giving a "seal of quality" to a particular agency.[326] This was a view supported by the IMA, who argued that the adoption of ratings by regulators into some of their rules had the effect of conferring "excessive legitimacy" upon the agencies.[327] In response Mr Sants said that the FSA supported change in this area.[328]

210. We believe the issues of over-reliance and the quality of ratings are interconnected: if the flaws and limitations of ratings were more widely recognised over-reliance would naturally decline. To some extent many of the problems engendered by CRAs will disappear as a consequence of market forces. There is unlikely to be a great deal of appetite in the near future for complex securitised instruments which have been most poorly measured by CRAs. We support the European efforts to throw light on the methods and methodologies of rating agencies and we call upon the regulators and the CRAs to ensure that new information reaches all users of their service.

Global coordination and assigning responsibilities

211. The CRAs we heard from were broadly welcoming of new regulation. Stephen Joynt, President of Fitch Rating, confirmed that the agencies were all feeding into the European process and noted that the "the issues that are being presented there in a way are constructive for the industry". He told us that the agencies "do not agree with every facet of it and some of the implementation will be costly and difficult, but we feel pretty good about it".[329] Mr Drevon and Mr Bell considered the regulatory proposals important for bringing back confidence in the system and re-establishing the credibility of the credit rating industry.[330]

212. All of the CRAs and many of the other submissions we received called upon the authorities to ensure that global consistency in regulation was achieved. For example, Mr Bell argued that credit ratings were, by their nature, global and required globally consistent regulation:

one of the benefits of ratings is that they create a global benchmark and therefore if we had a European rating style, a Japanese rating style and a US rating style all operating in different ways with different rules and different procedures I think that would be very damaging for the industry. Some kind of global consistency, however it is achieved, is therefore very important we think.[331]

213. The IMA also argued that "any actions taken in relation to the originate-to-distribute model should be coordinated at a global level; otherwise they will be ineffective and result in competitive disadvantage for those who implement them".[332]

214. When discussing the potential for increased regulation of CRAs, Mr Sants pointed out that the FSA "did not have a national remit to take [the regulation of CRAs] forward locally because they are not based here".[333] The Chancellor also spoke in favour of a global solution suggesting that a national solution would be inadequate:

Given the nature of what we are dealing with, a solution in one particular country, especially when institutions are trading across several, I do not think will be adequate. You need to have a solution to that certainly at a European-wide level, but also much wider than that, internationally. For us to have a solution that is purely directed to the UK would not be enough.[334]

He went on to assert that "if you left America out of it that would be rather a glaring hole in things given how important American institutions are".[335]

215. We note that the major credit rating agencies are global organisations. We call upon the Government to take forward efforts with other governments so as to ensure that a globally consistent approach to regulating credit rating agencies is achieved.

2 284  84 Q 2107 Back

2 285  85 Q 80 Back

2 286  86 Q 1205 Back

2 287  87 Ev 226 Back

2 288  88 Q 1205 Back

2 289  89 Ev 226 Back

2 290  90 Q 2108 Back

2 291  91 Q 1206 Back

2 292  92 Ibid. Back

2 293  93 Q 2218  Back

2 294  94 FSA, The Turner Review, A regulatory response to the global banking crisis, March 2009, p78 Back

2 295  95 Q 2108 Back

2 296  96 Q 743 Back

2 297  97 Q 708 Back

2 298  98 Ev 226 Back

2 299  99 Q 2352 Back

3 300  00 Q 77 Back

3 301  01 Q 1201 Back

3 302  02 Q 1202 Back

3 303  03 Ibid. Back

3 304  04 Q 1201 Back

3 305  05 Q 2108 Back

3 306  06 Q 2217 Back

3 307  07 Q 65 Back

3 308  08 Ev 295  Back

3 309  09 European Commission, Proposal for a regulation of the European parliament and of the council on credit rating agencies, 2008/0217 (COD), 12 November 2008 Back

3 310  10 FSA, The Turner Review, A regulatory response to the global banking crisis, p 78 Back

3 311  11 Ev 153  Back

3 312  12 Q 1882 Back

3 313  13 Ev 334 [The Financial Inclusion Centre] Back

3 314  14 Q 1267 Back

3 315  15 Q 1268 Back

3 316  16 Ibid. Back

3 317  17 Q 1269 Back

3 318  18 Q 1267 Back

3 319  19 Q 2352 Back

3 320  20 Q 81 Back

3 321  21 Q 1271 Back

3 322  22 Q 1271 Back

3 323  23 Ev 153  Back

3 324  24 Q 77 Back

3 325  25 Q 2218 Back

3 326  26 Ev 153  Back

3 327  27 Ev 226  Back

3 328  28 Q 2218 Back

3 329  29 Q 1217 Back

3 330  30 Q 1262-63 Back

3 331  31 Q 1263 Back

3 332  32 Ev 229  Back

3 333  33 Q 2218 Back

3 334  34 Q 67 Back

3 335  35 Q 70 Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2009
Prepared 15 May 2009