Standard & Poor’s and other ratings agencies need to re-examine their methods of mathematical modelling following the global financial crisis, says an international scholar of quantitative finance.
Professor Erik Schlögl, who served as an expert witness in the first case brought against S&P by twelve local councils in New South Wales, is Director of the Quantitative Finance Research Centre at the University of Technology, Sydney. He recently spoke at the University of Cape Town at a special seminar hosted by the African Institute of Financial Markets and Risk Management (AIFMRM). This seminal court case revolved around ratings issued for highly complex structured credit products known as constant proportion debt obligations (CPDOs).
Schlögl’s remarks came as fresh news broke that S&P would have to pay US$1.5 billion to resolve a collection of lawsuits over its ratings on mortgage securities that soured in the run-up to the 2008 financial crisis, concluding one of the US government’s most ambitious cases tied to the housing collapse.
That settlement was only achieved after more than two years of litigation, as S&P tried to work against allegations that it issued overly optimistic ratings in order to keep business booming.
Schlögl said the global financial crisis had prompted investors to reassess their expectations of rating agencies.
In the original Australian dispute, the civil case between Standard & Poor’s and twelve local councils in New South Wales set a precedent the world over – not in a strict legal sense, but, although laws differ from country to country, cases of this type do tend to be influential across borders, said Schlögl.
In essence, Standard & Poor’s appeared in court regarding the ratings it issued. The court found that there were no reasonable grounds for these ratings, nor had reasonable care been taken by Standard & Poor’s in arriving at these ratings. The rated products eventually failed, resulting in financial losses of over 90%, for which the agency (jointly with the issuing investment bank, ABN AMRO) was found liable.
Since the abovementioned case came to court, similar cases have appeared across the globe, with a concentration in Europe and the US.
The core issue in cases of this nature, believes Professor Schlögl, is to establish whether there was negligence involved. “In the most generous interpretation [in the Australian case] there was incompetence,” he said. “If one postulates competence, then there was something more sinister going on.”
Schlögl adds that S&P used a flawed model to determine the ratings, and this model must be re-examined – although it was not only the modelling that played a role in the erroneous rating. “One could say the product was only rated triple A according to S&P’s own criteria or model if you made these overly optimistic assumptions,” he says. “But the product was actually so highly unstable that if you made more conservative assumptions on any of the inputs, the risk that you would get out of the model from S&P is greater than what would have merited the rating of Triple A under S&P’s own criteria.”
He added, “The model was flawed but that never played a real role in the litigation. It was the fact that S&P used these unrealistic inputs into their model.”
According to the judgement, in 2006, ABN AMRO Bank created a new financial product known as the constant proportion debt obligation or CPDO. ABN AMRO retained S&P to rate the CPDO and sought a rating of AAA for the product – which S&P’s duly gave.
Local Government Financial Services (LGFS), got its client StateCover, a workers’ compensation insurer, to purchase a total of A$10,000,000 of the CPDOs (a series of which was known as Rembrandt 2006-2 notes). LGFS then purchased $40,000,000 (later increased by another $5,000,000) of Rembrandt 2006-3 notes for the purpose of on-sale to councils in New South Wales. Between November 2006 and July 2007 LGFS sold $17,000,000 of the Rembrandt 2006-3 notes to 15 councils. In October 2008, the Australian CPDOs cashed out because their net asset value fell below 10% of the par price for which they had been acquired. StateCover suffered an alleged loss of $9,244,000 and the councils suffered alleged losses totalling nearly $16,000,000.
“Cases like this are a relatively new phenomenon – appearing after the global financial crisis,” said Professor David Taylor, Director of AIFMRM. “It will be interesting to see what direction they take. In the US, it may be possible that ratings agencies are protected by the First Amendment, in saying that they merely gave their opinion.”
US banking and mortgage banking expert witness Don Coker disagrees. Speaking of US cases where the First Amendment was invoked, he said, “In my opinion as an experienced financial services industry professional, the rating agencies failed to exhibit good faith, fair dealing, ordinary care, honesty in fact, and reasonable commercial standards in their analysis of some of these mortgage-backed investments and the ratings that they decided to assign to them.”