The great write-down debate is well under way. And one thing that is apparent is that there is a lot of confusion about how—or even if—the credit products held by some of our biggest financial institutions should be valued. Merrill’s estimates of its losses jumped from $5 billion to nearly $9 billion in the space of a few weeks, and Citigroup is now saying that downgrades in credit ratings of some of their holdings will result in a fourth-quarter write-down of between $5 billion to $7 billion. Some analysts are expecting write-downs all across the Street.
[More after the jump]
This morning’s Heard on the Street column in the Wall Street Journal takes a look the trouble firm’s are having with estimating the values of their assets. One reason the write-downs keep coming in waves is that the valuation models used by the banks are highly dependent on credit ratings. This makes them extremely vulnerable to downgrades by the ratings agencies.
“Ratings play a big role in valuation models used by many banks, investment funds and insurance companies. Meanwhile, the market for securities linked to subprime loans has deteriorated in recent weeks as defaults have confirmed some of analysts' most dire forecasts, increasing the likelihood of further ratings downgrades,” Carrick Mollenkamp and David Reilly write.
So more credit downgrades could trigger more losses, and there is little sign that the turmoil in the credit markets has subsided. The ABX index, which is based on risk of underlying mortgage bonds, currently implies even deeper losses. The ratings agencies may still be behind the curve here.
And this points to an even deeper problem with the financial models used to value these assets—namely, it’s not clear why anyone should be relying on the ratings agencies at this point. Investors don’t seem to be, since they are shying away from even highly rated credit products. So why are the banks still relying on these ratings? Haven’t the numerous downgrades demonstrated that the ratings agencies are not reliable on this stuff?
It’s hard to escape the conclusion that the banks are relying on the ratings agencies because (a) despite what they say about sophisticated mathematical financial models, they don’t know any other way to value their assets and (b) they believe they get some CYA value by relying on a third-party’s point on view. Larry Ribstein made this point today.
“So why didn't the firms re-price on their own to reflect the rise in defaults as soon as they learned of that rise? Could it be the risk of liability for departing from industry standards?” he asks.
The Financial Times this morning also takes a crack at looking at the difficulty with valuing credit instruments that the banks are holding. One new reason they come up with is that government is creating uncertainty. There’s clearly a couple of different “reforms” afoot on Capitol Hill to address the excesses of the mortgage loan market and they point in different directions for the banks. If the US government compels lender lenience towards borrowers, then the losses in the subprime and other sectors could be as much as $100 billion. Or several times that number. On the other hand, if the government bails out the mortgage holders directly—by paying off part or all of their mortgages, or pushes interest rates even further down—then the losses could be far less. There’s no model for predicting what the government will do here.
The uncomfortable reality seems to be that no-one really knows the extent of the losses and that any numbers we're hearing are guesses at best.
Why Citi Struggles to Tally Losses [Wall Street Journal]
What's the damage? [Financial Times]
Citi and Merrill [Ideoblog]