Credit rating agencies have been cast as villains in the recent financial downturn. They are blamed for misjudging the creditworthiness of US sub-prime mortgages and collateralised debt obligations, thus helping to precipitate the crisis. “There were products where the raters appear to have got their analysis very wrong. And some people were using ratings in the wrong way, not understanding what those ratings meant,” says John Deacon, CEO of Giltspur Capital, a structured finance asset management and advisory firm. The blame game is understandable, because rating agencies play a crucial role in the capital markets. Investors, borrowers, issuers and governments all use their ratings to help make decisions on debt investments and financings.
The credit rating industry is dominated by three multinationals: Fitch Ratings, Moody’s and Standard & Poor’s. Many institutions, such as insurance companies, are not allowed to buy debt that is not top-rated; for banks, ratings are central to the calculations of their capital adequacy ratios (see panel, p16). Rating agencies say their job is to provide an independent opinion about the credit quality of issuers and debt securities. That is, they judge the risk of a default. “A credit rating is an opinion on whether, at the bond’s maturity, you will be repaid or not,” says Deacon. “It doesn’t tell you about what happens to the price of that security over its lifetime. So if a security drops in value by 30%, but pays off at par at maturity, the raters have got the rating right.” Nor do rating agencies assess an investment’s liquidity, price volatility, or indeed, whether it is good value.
Rating agencies admit, albeit very cagily, that they have not done well in the past few years. “Structured finance is one of the sectors most adversely affected by the financial crisis and the ratings in some asset categories have performed below expectations,” says Ian Linnell, global head of structured finance at Fitch Ratings. “That said, structured finance ratings have exhibited varied performance; many have performed very well in this period, such as European structured finance and US asset-backed securities.”
Three main flaws
There are three main criticisms of rating agencies: first, that they are beset by lack of competition and conflicts of interest; second, that they are slow to downgrade securities; and third, that their methodology for analysing some types of structured finance – such as residential mortgage-backed securities (RMBS) and collateralised debt obligations (CDOs) – was flawed.
The proposed solutions to these problems range from encouraging more competition among rating agencies or having an independent agency funded by a market levy, to creating a single, state-run agency. But in the US, Congress has backed away from legislating any major changes. The European Union has taken a more robust approach. It is setting up a pan- European authority to regulate and supervise rating agencies operating in the EU, which must now be registered and provide quarterly data on their securitisations; the Committee of European Securities Regulators (CESR) has been put in temporary charge of this.
The CESR will also set up and maintain a database on the past performance of rating agencies operating in the EU, allowing users of rating services, such as investors, to check their track records. That said, rating agencies have done much soul-searching and are overhauling themselves. Their initiatives include a legal and operational separation of rating and non-rating activities; setting up structures to review their performance; and disclosing much more, including additional information on structured finance ratings.
They are worried about the uncoordinated initiatives that countries are taking to regulate them and want to see a ‘level playing field’ agreed internationally. The financial crisis spotlighted an acute conflict of interest in that rating agencies are paid by issuers of securities – the very organisations whose products and performance they are judging. Thus they are open to the suspicion that their opinions about a deal’s creditworthiness are not impartial.
Moreover, rating agencies also provide advisory services. On structured financings such as commercial mortgage-backed securities issues, an agency might advise an issuer on the construction of a security, for a fee, then later rate that security, earning another fee. At the height of the market, agencies were earning big up-front fees of up to $1m a time from securitisations.
“There is an inherent conflict when the party receiving the rating pays the fees,” says Conor Downey, finance partner at law firm Paul Hastings. “Every professional service provider is very conscious of their income source. Anybody in that position would be motivated not to upset their client.” Another danger of the ‘issuer pays’ arrangement is that issuers will favour the agency that awards their securities the best ratings. ‘Ratings shopping’ by issuing banks was notable with US CDOs, and elsewhere.
Moody’s, for example, rated mainly the most senior, AAA European CMBS tranches, while the other two agencies went further down the capital stack. “The reason was simply that most of our ratings for the more junior classes would have been below the desired ratings, so we were not asked to rate these classes,” says Christian Aufsatz, senior vice-president, CMBS ratings, at Moody’s. There is no obvious solution to these conflicts of interests, but new EU regulations attempt to curb some of them. For example, they stipulate that rating agencies should not provide consultancy or advisory services and “should not make proposals or recommendations regarding the design of a structured finance instrument”.
But there is a caveat: “Rating agencies should be able to provide ancillary services where this does not create potential conflicts of interest with the issuing of credit ratings.” The question of who should pay for ratings is still being debated. Some want investors to pay for credit analysis, as they did in the 1970s. But rating agencies point out that substituting one set of interests for another does not remove conflicts of interests.
The ‘issuer-pays’ model does have one big advantage: ratings are released publicly, for free. The ‘investor-pays’ solution would reduce transparency, as investors might not want to share their information with others. “I find rating agencies useful because they have access to information that is far better than that released to investors,” says Graham Page, head of credit at RZB Bank in London. “They divulge it in presale reports and are happy to chat with you.”
The methodology used to rate structured finance products has also been lambasted, sometimes unfairly. It is widely agreed – even if the agencies won’t publicly admit it – that they fell down on their analysis of US subprime RMBS issues and CDOs. “Prime RMBS has performed very well, and while CMBS has cratered in Europe, that is less to do with rating agencies than with the valuations that underpinned the securities,” says Page.
Missing the big picture
“On an issue-by-issue rating of CMBS in Europe, they didn’t do badly,” says Deacon. “But I’d criticise them for not looking at the systemic picture. I don’t think they looked at real estate finance across Europe and how much leverage was put to work. Isn’t it likely or possible that there could be a systemic risk where everything will be hit simultaneously if there’s a withdrawal of liquidity?” Aware that they need to restore confidence in their ratings, agencies are addressing this and other criticisms.
Standard & Poor’s now publishes scenarios outlining different stress levels that rated products could survive without defaulting – for AAA ratings, that is anything up to a ‘Great Depression’ scenario. Fitch’s Linnell says his firm has increased its emphasis “on qualitative and determin- istic analysis. Early in 2008, we introduced regular stress-testing of structured finance ratings, which focus on their vulnerability to a broad range of increasingly severe macro-economic factors.”
Agencies have also been criticised both for taking too long to downgrade positions and for flip-flopping too quickly – in one case last year, some US CMBS bonds were steeply downgraded to junk status, only to be reassigned AAA status the following week. It is a fact that ratings tend to lag behind the market. “Most of that is to be expected because raters want to see clear credit trends before deciding to upgrade or downgrade a position,” notes Deacon. “The market will have a far more knee-jerk reaction: this is getting worse – let’s mark the price down.” Moody’s uses a bottom-up approach in its ratings of European CMBS, assessing the default risk of loans both over their term, and of repayment at maturity, on a loan-by-loan and property-by-property basis.
Scenarios including expected trends in property values, occupational markets, defaults and the availability of finance are also incorporated. Last year, Moody’s and others downgraded about 40% of all CMBS transactions. “We had to take into account that property values fell quite dramatically and we do not expect them to increase substantially over the short to medium term,” says Aufsatz. “The downturn was faster and more pronounced than ever before, and after the collapse of Lehman Brothers it was hard for investors to get funding, as the market froze.
“Over 2009 we went through all the European CMBS issues we rated, incorporating our assumptions on property value, how it would recover over time and the refinancing prospects of every single loan, given that the implied loan-to-value ratios were – and still are – very high and that the lending markets were not functioning, and have only partially recovered since then.” Rating agencies cannot force valuations of assets underlying CMBS loans. “We favour loans where annual valuations are incorporated in the loan agreement and the borrower would have to provide updated, third-party valuations for surveillance,” says Aufsatz. “But only a handful of loan agreements have this provision, meaning that the power to call for a new valuation lies with the servicer.
Fear of valuations
“At first, market participants were hesitant to call for updated valuations, assuming that drops in market values are only temporary and a valuation update then would have resulted in unnecessary LTV covenant breaches and potential loan defaults. “With hindsight, however, it might have been better to test LTV covenants at the end of 2007 or start of 2008, because then servicers could have put more pressure on the borrowers earlier in the downturn.” Rating agencies are also addressing the volatility of ratings, devising ways to measure credit stability. Moody’s, for example, has introduced V-Scores, which highlight the types of securities that are more exposed to data limitations and modelling assumptions; and parameter sensitivities, which measure how a security’s initial rating might have differed if key parameters had been different.
To educate investors and restore trust in ratings, the agencies are also releasing far more information about the methodologies, models and key assumptions they use in the rating process. This disclosure is now a legal requirement under new regulations introduced by the European Union in 2009. These regulations also require structured finance instruments to carry a symbol to distinguish them from other ratings. But most investors, issuers and raters consider this branding pointless or even confusing. “If anything, it detracts from value for investors, because an AA-rated corporate instrument and AA-rated structured finance instrument should have the same probability of default,” says Giltspur Capital’s Deacon.