As further heads roll in the top ranks of America’s big banks, the magnitude of the subprime debacle is only now beginning to be revealed – as the UK is about to find out, says Tim Bennett
What’s happened?
The credit crunch is claiming new victims. This time it’s Chuck Prince, head of the world’s biggest bank, Citigroup, who walked the plank just days after the resignation of the Merrill Lynch CEO, Stan O’Neal. Both Citi and Merrill Lynch have been hit by massive subprime losses.
Worse, Swiss bank UBS has warned that more subprime writedowns are “possible” (banking speak for “certain”) and JP Morgan analysts speculate that total losses from mortgage-linked securities might reach $200bn.
Meanwhile, Bill Gross, CEO of Pacific Investment Management, hinted that even this is conservative, commenting “there are $1trn worth of subprimes and garbage loans”.
Why all the different numbers?
Scarily, even now no one seems sure how much subprime junk is out there, or who owns it all. And the world’s investment banks can’t even seem accurately to value what they do know about. Merrill’s recent $7.9bn writedown, for example, followed an estimate of just $5bn only a few weeks earlier.
Normally, if you want a price for something you simply check what it is selling for in the market. The trouble is that for many of the assets in question – often bonds backed by subprime mortgages – there is no “market” because, due to a breakdown in mutual trust following wave after wave of US home-loan defaults, no one is buying.
So banks, forced by regulators and accountants to “mark to market” for reporting purposes, when there is no market to mark to, are making up their own valuations based on what Jeremy Warner in The Independent describes as, “startlingly different attitudes” about how mortgage-backed securities should be priced.
How are the banks doing this?
By using “mark to model”, a hugely subjective process whereby the owner of a bond values it by adding up all the future cash receipts (a string of interest receipts plus a final repayment) and compares the figure generated, say $80m, to the value already “in the books” the last time the exercise was done – say, $100m. The $20m difference is knocked off the value of the bond on the bank’s balance sheet and reduces reported profits.
The fact that a bond can be worth $100m one month and only $80m the next reflects the current need to reduce (or “discount”) the bond’s expected cash flows to take account of a host of risks, especially of non-payment by the underlying borrower.
A mortgage-backed bond might be underpinned by loans to hundreds of individual mortgagees adding up to the original $100m, say 30% of whom are either defaulting or thought to be likely to default. But that number could be 25%, or even 40%. No one knows, but small changes in this type of assumption can swing the value of the bond.
So there is no set valuation method?
No. However, as the FT points out, there may be a way of telling how far out some of the valuations could be. There may not be a market price for many bonds at the moment, but there is one for the insurance policies – a type of derivative called a credit default swap – bought by nervous bond holders and designed to payout in the event of a default.
An index, called the ABX-HE, captures the price of this protection across groups of 20 different bonds at a time – the greater the insurance premium, the less the insurers think the underlying bonds might actually be worth, and the lower the index. On this basis, say analysts at Merrill Lynch, debt being carried by UBS at 90% of its face value should be at 40%. This implies “another $8bn write-down”.
This isn’t good for the insurers either is it?
No. They may be in trouble too. BusinessWeek cites the example of ACA Capital. It has sold around $16bn of cover in the last few years, yet holds less than $350m in capital to fund payouts if the underlying bonds go bad.
Meanwhile, the largest bond insurer in the world, MBIA, announced a third-quarter loss of $36m, prompting S&P to announce that it is reviewing “their ability to withstand further subprime stress”.
How do these write-downs affect investors?
The falls in bank share prices aren’t good for investors. But there is a bigger danger out there – that banks hit by subprime losses fall foul of regulatory capital requirements. In the US and UK big banks have to maintain a minimum relationship between their own funds and the amount other counterparties, who might default, owe them. Each write-down reduces those funds and ups the odds of cut dividends, asset firesales, or even a bank closure.
No-one’s immune: CIBC analysts say even Citigroup will struggle to deal with “low capital ratios” and may need to raise “up to $30bn in equity”. Expect “pressure across all financial stocks” as a result.
What might be the affect on the UK?
“Devastating,” says Peter Toeman at HSBS. Although the spotlight has fallen recently on US investment banks, such as Citigroup and Merrill Lynch, the outlook for financial stocks in the UK is equally grim, with analysts continuing to downgrade the likes of Barclays, Royal Bank of Scotland and HBOS.
And it’s not just investors in banking shares who are getting burnt. Across the board, lenders are withdrawing risky mortgage products, making life increasingly difficult for homebuyers, while in the credit-card market Moneyfacts.co.uk described the situation as “chaos” – 125 separate rate and fee increases have been introduced in just two months. Far from being over it seems the real crunch has only just begun.