Another subprime lawsuit currently slugging its way through the courts is Cassa di Risparmio della Repubblica di San Marino SpA (CRSM) vs Barclays Plc.
The case concerns some Barclays-structured and triple A-rated CDO Squareds that were sold to the Italian bank around 2004. Cassa di Risparmio says Barclays “deliberately structured the CDO²s in such a way that they had, on Barclays’ own calculations, a very substantial risk of default, vastly higher than CRSM was led to believe.” The British bank, CRSM alleges, stood to profit from the CDOs’ decay.
Barclays, as Bloomberg reports, denies the allegations that it sold CRSM structured notes that misrepresented the risks. “CRSM was perfectly well aware of the basis upon which my client would wish to do any trade,” Barclays’ lawyer said this week.
The idea of a”ratings loophole” — or what CRSM calls “credit ratings arbitrage” — is currently making headlines, however. In a nutshell, CRSM says, rating agencies estimated the default probabilities of the reference assets underlying the CDOs concerned based on historical data. Barclays, however, used credit spreads.
Here’s CRSM’s lawyer making his closing submission this week:
There is very little said in Barclays’ closing submissions about the credit ratings arbitrage, which we say simply reflects the fact that there isn’t an answer to the point. What they do seek to say is on page 11 at paragraphs 20 and 21. Essentially, their answer is that probabilities derived from spreads — this is the same point — don’t mean actual probabilities of default, and so if you put in names with higher spreads, it doesn’t mean the probability of default is higher.
The flaw in that answer is not only the flaw that credit spreads do give you information about actual default probabilities, but when it is argued in this context, it enters an even higher realm, we say, of unreality, because you can have an argument about whether credit spreads are a better measure than historical data when you want to estimate default risk, but one thing that is absolutely clear, we say, and the contrary was certainly not suggested by Dr Ellis, is that if you are using spreads as your measure and you are looking at two different names on the same day and they both have the same rating and one has a spread, say, twice that of the other, then no-one who is knowledgeable about these matters, and certainly not Dr Ellis, is going to suggest that the higher spread doesn’t signify a higher default risk.
If there are extrinsic factors which affect spreads and mean that over time they may fluctuate up and down for reasons which aren’t solely linked to expectations of default risk, then those are all factored out if you are looking — comparing two names on the same day with the same rating, and we say also that it is clear from the documents that we looked at, the contemporaneous documents that Barclays disclosed, that they themselves recognised and believed that choosing names with higher spreads increased the risk of default of the CDO squareds. See, for example, the email about “the more we arb this” that we looked at on numerous occasions.
So it is the arbitrage strategy which Barclays used in this case which explains how it is that they were able to structure an instrument with a far higher probability of default than the rating agencies recognised. In effect, what they were doing was exploiting the loophole in the rating agency methodology which treated the historical rate, default rate, as the estimate of the probability of default for all names in that bracket, with that rating, as we saw, without taking account of the fact that within any given rating, at any given time, some of those companies were much riskier than others.
The full CRSM and Barclays closing submissions are in the usual place.
Related link:
Barclays used credit ratings loophole, San Marino lawyer says - Bloomberg