http://ftalphaville.ft.com/2012/11/06/1246821/working-together-on-the-cpdo-dream/
and.....
http://market-ticker.org/akcs-www?post=213583
In the prodigiously long and detailed judgement from Australia’s Federal Court on the CPDO affair, there is a story jumping out from behind the complexity of the product itself. It’s about the key role played by former Standard & Poor’s employees who had left the rating agency to go to banks to help structure these products.
The judge found that S&P itself — rather than the individuals involved — had misled 12 local councils in Australia by awarding triple-A ratings to toxic structured debt. But as someone who has written about curious structured products, from CPDOs to rated CDO equity tranches, I have met a lot of the people whose names cropped up in this case. I was curious about what Judge Jagot would make of the raters who had effectively become in-house advisors.
Juan-Carlos Martorell and Michael Drexler were of this ilk, introduced in paragraph 73 of the judgement:
Mr Drexler and Mr Martorell, both former employees of S&P, were primarily responsible for managing ABN Amro’s dealings with S&P in respect of the rating of the CPDO.
This description of their role was supported by input from Jamie Cole, also an ABN Amro employee and colleague of the pair.
Additional characters in the cast inlude: Cian Chandler, an S&P analyst responsible for liaising between S&P quant teams, ABN Amro and other senior S&P peeps; and Perry Inglis, Cian Chandler’s boss.
I met all these guys apart from Drexler and found them — and their peers at other banks and agencies at the time — to be both intellectually honest and endlessly curious.
The judgement paints a picture of a hugely long and drawn out negotiation process between ABN Amro and S&P over the CPDO that’s the subject of the suit. By paragraph 128, the relationship dynamics take centre stage, with S&P struggling to make the leap from the last product the two sides worked on to the CPDO version of it. The models of the two parties were, at least initially, spitting out different answers.
The ridiculousness of it can perhaps be summarised by noting that a similar structured product paid an income of 50 basis points over Libor and attracted a single-A rating. Somehow that product had been rejigged into something that paid 200 basis points over Libor with a triple-A rating(!) — the CPDO.
Credit default swap markets were only trending towards lower and lower yields during 2006 when all this was going on. The chase for yield without comprising on the rating was in full-swing and CPDOs were positioned into place.
As for our characters, here’s Chandler writing to Inglis and other senior S&P bosses to update them on the discussions he had had with Drexler over at ABN:
He’s been insinuating that he was going to go over Sriram and Derek’s [two more junior S&P analysts] heads, so I wouldn’t be surprised if you get a call soon.The length and detail in the back and forth between the rating agency and ABN is astounding — and is part of why the judgement runs to so many thousands of words. The discussions covered the assumptions that must be baked in order to get the outcome that was desired: a specific pay-out with a specific rating. Historic market volatility, what the long-term spread on the credit indices could be expected to look like, and how quickly spreads would revert to an assumed mean after diverting from it, were all part of this.Clearly, a close working knowledge of how rating agency models and processes worked was a big advantage, especially when it came to getting a new product over the line. There were concerns in the months following this deal about the way in which such ratings had been reached.But it’s later on, after the CPDO was already being marketed, that an unforeseen problem crops up from the questions posed by a potential investor. And it’s at that point in time that the bank appears to really lean on its former S&P guys.The problem was around the roll over of the credit derivatives indices that happened every six months. CDS buys you protection against default of a bond for typically five years — to maintain a near constant five-year price for credit default risk, the indices had to be rebooted regularly. The index compilers did it every six months. This process also ensured that the credit exposures that had become sub-investment grade could be removed.Switching your own exposure from the old (or “off-the-run”) index into the new, “on-the-run” index, involved transaction costs. The CPDO modelling up to this point had not properly accounted for this.Paul Silcox at ABN’s Exotic Credit Derivatives Group emailed his team (paragraph 217):Potential issue I can see is that we mention it to S&P and they get all hung up on the roll cost and the modelling of it. However, if we don’t mention it then a client may ask about it on a call with S&P at which point things may be come a little ropey.To which Cole responded (paragraph 218):Yes I fully agree as discussed last week. JC/Mike best placed to know how S&P think – I had asked JC to think how best to approach them…The next few paras may make for instructive reading. Martorell (aka “JC”) initially notes in a response (paragraph 219) that:We should avoid S&P to overthink and open a can of worms.A few days later Martorell gives a fuller analysis of the pros and cons of telling S&P about the issue, concluding that it is probably better to say nothing (paragraph 222, emphasis mine):I think if asked by clients S&P wouldn’t want to look stupid that they haven’t thought about it. The diplomatic/ institutional S&P answer by an analyst will go typically as follows: “the model directly or indirectly captures the risks associated with CPDO” (ie stressed defaults, spread vol, low mean reversion, bid/ask, and structural considerations). All these mitigants are consistent with AAA standards.Useful to have someone around who knows that.
and.....
http://market-ticker.org/akcs-www?post=213583
The First Amendment Defense Fails (In Australia)
No comments:
Post a Comment