The decision by Australian Federal Court Judge Jayne Jagot to find rating agency Standard & Poor’s liable for investor losses in connection with a transaction it rated during the structured credit bubble has shaken one of the rating agencies’ fundamental shibboleths – that they provide opinions rather than investment advice and therefore cannot be held liable for their determinations.
The ruling itself stems from a recent case (Bathurst Regional Council vs LGFS Pty) brought by 12 local councils in New South Wales against municipal adviser Local Government Financial Service (LGFS). In 2006 the councils invested A$18.5 million ($19.3 million) in a constant proportion debt obligation (CPDO) structure called Rembrandt 2006-3 that LGFS had bought from ABN Amro.
CPDOs sold unfunded CDS on corporate debt indices such as the iTraxx in Europe and the CDX index in the US. Because these indices are rolled over every six months, the CDS were closed out and new transactions entered into when the indices were rebalanced. At this point the difference in spread from the last roll – which will be multiplied by the extent to which the structure is levered – was applied to the structure. Thus, if corporate spreads fell – as they did – the impact on the structure was devastating. Once the markets imploded in 2007 it did not take very long for CPDO deals – some of which were levered as much as 15 times – to implode as well. The Australian councils incurred an A$16 million loss on their A$18.5 million investment.
Judge Jagot deemed that Standard & Poor's had engaged in “misleading and deceptive” conduct by rating the Rembrandt 2006-3 deal triple-A, concluding that a “reasonably competent” rating agency would never have given such a rating. S&P said in a statement: “We are disappointed with the court’s decision, we reject any suggestion our opinions were inappropriate and we will appeal the Australian ruling, which relates to a specific CPDO rating.”
However, this is an area where the rating industry has ample previous form.
In early 2007, Moody's discovered that the rating model it had been using to rate certain CPDO notes contained an inadvertent coding error, which resulted in the rating being set between 1.5 and 3.5 notches higher than it should have been. However, when the error came to light in 2007 the agency did not downgrade the $1 billion of CPDO notes in 11 separate deals that were incorrectly rated. A member of the European rating committee emailed at the time: “To be fully honest this latter issue is so important that I would feel inclined at this stage to minimize ratings impact and accept unstressed parameters that are within possible ranges rather than even allow for the possibility of a hint that the model has a bug.” The SEC investigated the issue in 2010 but did not pursue a fraud enforcement action against the agency.
Judge Jagot found against S&P in the November 5 judgement largely on the basis that S&P’s rating methodology was flawed. The agency does seem to have relied heavily on ABN Amro’s own model and ABN Amro’s inputs to determine the rating – using higher spreads and lower volatility assumptions than it should have.
In its CPDO rating methodology S&P assumed that spreads would revert towards 40 basis points over the first year and 80bp for the subsequent nine years of a 10-year CPDO deal. It assumed a volatility of spread movement of 25%. As the court ruling points out, the fact that the CPDO performed better when the long-term average spread increased was counter-intuitive because increasing spreads are a reflection of increased risks in markets. At least one person within S&P considered that ABN Amro, whether intentionally or not, had effectively “gamed” the model it knew S&P would apply to rate the CPDO. It is informative to note that the ABN Amro executives primarily responsible for liaising with S&P over the rating - Juan-Carlos Martorell and Michael Drexler - were both former employees of the rating agency.
In protracted exchanges in the testimony it emerges that S&P became aware that the initial portfolio spread of 32bp would be insufficient to generate a triple-A rating, something that certainly seems to have caused internal dispute at the agency. In a colourful but ill-tempered email exchange between Derek Ding, a quantitative analyst at S&P jointly responsible for the modelling of the CPDO for rating by S&P and Sebastian Venus a quantitative analyst, primarily responsible for the day-to-day development of S&P’s own model for rating CPDOs, the controversial nature of the rating methodology is very apparent.
When Venus declared: “I am done with the whole CPDO – wish I was never involved in that whole mess that was made,” Ding retorts “What a wuss.” Venus then replies “I am not responsible, I just helped build an internal model. You are the wuss for bending over in front of bankers and taking it...You rate something AAA, when it is really A-? You proud of that little mistake?”
When this exchange was revealed in court, S&P stated: “Those email exchanges also evidence the fact that S&P did not simply rubber-stamp ABN Amro’s opinion. Conflicting views are to be expected in the context of any serious process of analysis and consideration.”
CPDOs were hands down the most contentious of the structured credit products to emerge from the boom and many in the market were sceptical about their triple-A status. CreditSights said at the time that “while the strategy is very simple and surprisingly robust when modelled...there are a number of things that give us grounds for unease.” The structures were developed by ABN Amro as a rated version of constant proportion portfolio insurance but were swiftly copied by many other banks. The fact that the deals could pay 200bp over mid-swaps for triple-A risk immediately drew suspicion; rating agency Fitch, which did not rate any CPDOs, correctly argued that there was insufficient historical performance data on the iTraxx and CDX to model deals for 10 years. However, even CPDOs' harshest critics said that the structure probably deserved to be investment grade – just not triple-A.
But they subsequently proved to be a graphic illustration of the risk amplification that staggering leverage embodied. In a paper published in 2010, Michael Gordy and Sren Willemann at the Federal Reserve described them thus: “In its complexity and its vulnerability to market volatility, the CPDO might be viewed as the poster child for the excesses of financial engineering in the credit market... For both S&P and Moody’s, the rating of CPDOs was driven by models that assigned effectively zero probability to reaching the levels of spreads actually realized in the CDS markets in late 2008. This spike in spreads would, to the best of our knowledge, have triggered default and large losses for every CPDO note that had managed to survive to that point. All the issues in question had been rated AAA at origination in late 2006 or early 2007.”
The decision by the Federal Court of Australia to find Standard & Poor’s (along with ABN Amro – since acquired by RBS – and LGFS) liable for the local council losses has been seized upon as a precedent for wider legal action against the rating agencies in connection with highly structured products. But investors expecting a flood of cases might be disappointed.
Free speech
Rating agencies have long been excoriated for their role in the credit crisis. But the fact that their ratings are credit opinions has always offered protection under the first amendment to the US constitution – the guarantee of free speech. Roughly 50 cases have been brought against the rating agencies since 2007 but the majority have been dismissed on this basis. Their defence is that investors buying on rating have failed to conduct their own due diligence in relying on an credit opinion for an investment decision.
The Australian judge’s ruling in this case was probably swayed by the complexity of the deal in question. Although local councils losing money on derivative transactions they do not understand has a long and rich history it would be tough to expect them to have fully understood the CPDO. It is therefore unlikely that the reasoning applied in this case can be taken and applied to more vanilla mortgage-backed deals that also turned out to have been rated on overly optimistic assumptions. Questions were raised over the rating methodologies of CPDOs from the moment they were devised, whereas even the worst-performing sub-prime MBS was based on securitization technology that had been tried and tested for decades. The length of the Australian ruling (1,500 pages) gives some idea as to how complex the deal was.
Another factor mitigating against many new legal actions is the length of time that has now elapsed since the collapse of the structured credit market. In the US the statute of limitations runs for a year after the date of breach but in cases alleging deliberate misrepresentation it can run for three years (see CDO litigation: Casino banking victims face last roll of the dice Euromoney February 2011).
In the case of CPDOs, however, of which around $5 billion were issued, there might be impetus for investors to follow the Australian councils’ lead. The firm that funded their litigation, IMF Australia, is believed to be examining the viability of further claims in Europe (CPDOs were largely arranged by European banks and sold in Europe). But as to this ruling opening the door to a flood of litigation against the rating agencies – much as many investors might like it to – this might be a conclusion too far.
 |
| Biannual index roll dates and their midpoints are marked on the time axis. Cash-out event marked with a circle. Source: Federal Reserve Board |
