Credit: Graded ‘could do better’
Martin Steward talks to Standard & Poor's about what exactly it means to rate a security AAA
"Tramp stamp and hooker heels do not begin to describe the sordid, nonsensical role that the rating services performed in perpetrating and perpetuating the subprime craze."
Ouch. PIMCO supremo Bill Gross doesn't exactly mince his words. Even Warren Buffett, who supposedly sided with ratings agencies once he was forced to speak before the US Congress's Financial Crisis Inquiry Commission, and who is counted as the largest shareholder in one of them, declined to shower them with praise: their ratings of mortgage-backed securities (MBS) merely reflected the fact that they got caught up in the same "mass delusion" as everybody else.
Simon Collingridge, managing director in Standard & Poor's structured finance ratings division, acknowledges the disappointing performance of US mortgage-related securities but points out that the vast majority of S&P's ratings have performed broadly as anticipated during the crisis. "What does our AAA rating mean?" he asks. "It means ‘things have got to get very bad before the bond defaults and you lose any of your money', and ‘this is how we define losing your money'."
To be precise, a security can be graded AAA if the agency is confident that it will not default - that is, fail to deliver full and timely payment of interest and full and ultimate payment of principal - even under the levels of economic and financial stress associated with the Great Depression. "If you are one dollar short for one day, as far as we are concerned you have defaulted."
Figures 1 and 2 illustrate that these ratings were pretty reliable even through the stresses of 2007-09 - even the downgrade rate of 8.8% looks pretty robust, let alone the default rate of 0.3%. That is remarkable given the extent to which structured credit expanded the world of AAA (and indeed AA) from what was a couple of multinationals and a handful of sovereign entities. Within European structured credit, 82% of all issuance got the top grade.
That was another aspect of structured credit that has attracted derision - but that derision has not always been thought through. For example, corporate and sovereign issuers will one day have to refinance their outstanding debt, and any impairment in the liquidity of their bonds can feed back into their creditworthiness, and vice versa in a nasty vicious cycle. By contrast, the issuer of a structured credit security is normally simply passing through to debtholders the cash flows from a ring-fenced set of assets, and the problem of refinancing technically never comes up.
Secondly, ABS can be made more robust through credit enhancement steps like over-collateralisation. So not only is it easier to judge the likelihood of loss from a portfolio of mortgages than from a corporation (because of all the latter's moving parts, like refinancing risk), once you have identified the number of defaults the portfolio will suffer for a given level of economic stress, it is relatively simple to define the level of credit enhancement required to ensure that the AAA-rated tranche will pay every cent of its interest and principal. "The question around US sub-prime wasn't so much: ‘Why were there so many AAAs?' but rather: ‘What were the levels of credit enhancement in the deals?'" as Andrew South, a senior director in structured finance ratings services, puts it.
Still, investors struggling for liquidity or using ABS for collateral in 2008 faced pricing that many would argue should not be characteristic of AAA-rated securities. Others who saw that pricing could be forgiven for bailing-out for fear that it reflected a much greater likelihood of default - even though the creditworthiness of the securities might have been unaffected and the ratings are not meant to reflect any opinion about their liquidity.
This relates to a broader question about the stability one might expect from an AAA-rated security. The rating says nothing about price stability, but before the crisis it did not technically say anything about rating stability either. Collingridge concedes that the market presumed that an AAA-rated security would not only not default, but remain stable, and that S&P's willingness to refer to their historical stability in transition studies made that presumption "not unreasonable".
In response to this, S&P formally introduced a stability element into its AA and AAA ratings in 2009. Now, if stress tests on a candidate for a AAA rating reveal that its volatility could cause it to be downgraded to AA within one year or A within three, it can no longer be given the top grade - even though its ultimate risk of default is just as remote. "In structured finance that probably means that you need more credit enhancement," says Collingridge. "There are going to be certain things that can't be AAA any more, even though we are comfortable that they will still pay out under the most extreme circumstances."
One sector where things look uncertain at the moment is commercial mortgage backed securities (CMBS): loans that are performing perfectly well are beginning to come up for refinancing in an environment in which real estate values have halved and credit conditions have tightened significantly. A property worth $100,000 backed by a bank happy to advance 85% might now be worth $50,000 with a bank only willing to loan 65%. That does not mean that AAA-rated CMBS are about to default - even though commercial property values have slumped even further than they did during the Great Depression. "But does the likelihood of default still justify a AAA rating?" asks Collingridge. "Probably not."
So far the only downgrades in Europe from senior ratings under this new approach have been on some CDO and CMBS issues that look weaker under the new stability criteria, although with CMBS these have equally been driven by real economic factors (like the kind of thing described). But now, when the environment does begin to become stressed, in theory AAA-rated securities should be downgraded much sooner than they would have been before, preventing the shock of paper going from AAA to junk or default overnight - or the suspicion that it might be heading that way despite the persistence of a senior rating.
As Collingridge observes: "If the single driver of a bond's default potential is property value, for example, and we decide that a AAA is at risk of default if values go down by 50%, we are not going to downgrade after a 1% fall, but we're not going to wait until they've fallen 49%, either." The main benefit for investors may be to remove the risk of bonds hanging on to senior ratings until the very last moment.