In other words, fredoiul cannot profit from this any longer, as people with deeper pockets than him and earlier access to information already have ;-)

Can the last politician to go out the revolving door please turn the lights off?

FSA sounds alarm on subprime lending (4 July)

The Financial Services Authority has started disciplinary action against five mortgage brokers for weak "responsible lending" practices in the subprime mortgage market.

The UK financial regulator on Wednesday issued a damning report on the sector, which lends to borrowers with spotty credit histories. Arrears among subprime borrowers are currently running at 20 times those of mainstream mortgage holders, sparking concerns firms have taken on excessive risk in ramping up their lending in this area.

SEC examines subprime market (27 June)

The Securities and Exchange Commission yesterday said it had initiated a broad-based investigation into the troubled subprime mortgage market.

Christopher Cox, chairman of the SEC, told a congressional panel that the regulator was investigating a dozen subprime mortgage issues, including collaterallised debt obligations (CDOs), which are repackaged pools of debt sold to investors.

The SEC is also looking into the secondary market for these instruments.

Asked what had prompted the scrutiny, Mr Cox said "the climate and the environment". He added: "The attention that is being paid generally to problems in this area causes us to be alert to the potential for violations of the laws and regulations that we enforce."

Actually, it is not relevant to the "subscriber of the insurance": most of the time, he wants a hedge. If the rate of payments for the hedge doesn't chew up a too big proportion of the coupons, he'll be happy. He only wants a big name to sell the protection (you don't want your insurance to go belly up, do you ?)

The model is only pertinent to the book value of the contingent claim in the bank's account. Pretty much like the valuation of a portfolio of drivers at an insurer. The problem is that driver insurers have models based on big old historical records where Gauss works very well.

The banks do have a problem: they sold lots of protections for too cheap fees, valuing them a derisory price coming from a fancy model (ACME will default in a credit crunch ? nononon, that can't be). And they thought they were faster than Insurance companies, because models enabled them to roll out new products faster and without waiting for the accumulation of long historical records, etc...

Now the models will either be tweaked or they will make it apparent that lots of cash will be due shortly, and that the hedges of the banks are grossly inadequate. Note that it is only because the swing in the inputs is terrible (from bubble to crunch) that the model can do nothing but confirm the disaster (that common sense could have predicted). We do not know for real how these models would perform in a non-bubble, non-crunch period, because they have never been back tested on such a period.

I take it they are rather poor in that they did not incorporate more indicators that would have raised alarms as early as two years ago in the real estate market.

Also, I think the theoricians will soon discover such realities as "friction" and "inefficiency" in the market. There's plenty of this in an illiquid real estate market (legal costs, janitor costs, realtor fees...)

Pierre