Saturday, November 17, 2007

Fog and fear obscure the reality behind subprime losses

Another week, another memorable encounter with a nervous financial beast. This time, however, the animal in question is Royal Bank of Scotland, the British banking group.

Last week, RBS raised eyebrows when it was widely reported that one of its highly respected credit analysts had predicted that subprime losses could eventually rise to between $250bn and $500bn - or twice previous estimates.

However, this fascinating research was not being released to ordinary investors, even though it had been in the press. Instead, circulation was restricted to RBS's favoured clients. The analyst was unavailable for comment; apparently he was on holiday.

On one level, this is simply standard operating procedure for big banks these days. (RBS has particular reason to feel edgy as it is likely soon to disclose its own subprime writedowns.) But on another level, the saga reveals a much bigger point: how much fog and fear still surrounds the whole issue of subprime losses, notwithstanding the recent shocking write-offs by major banks.

For behind the scenes - and occasionally in public view - the credit analyst community remains distinctly divided about just how big the final hit might be, let alone what write-offs to expect from banks. Thus while some observers project a $100bn hit, others talk about $500bn. Perhaps there are even bigger numbers in ultra-ultra private reports I have yet to see.

This feels wearily familiar. A decade ago, I covered the Japanese bank crisis and became embroiled in a bad-loan guessing game that continued for many years. The tally of Japanese bad loans was estimated to be about $100m at the start of the 1990s, but by 1999 had risen to 1,000 times that size. I am told that a similar game occurred during the Latin American debt crisis in the 1980s and the Savings and Loans crisis - or indeed in almost every other recent banking shock.

What could make the 2007 subprime shock a little unusual, I suspect, is the way the rot is coming to light. Back in the 1990s, Japanese banks were able to sit on their problems for years, because they were a clubby bunch - and corporate loans have very long shelf lives. Similarly, in the Latin American debt crisis, American banks effectively colluded to conceal the rot as they restructured loans.

However, these days, Wall Street bankers have little appetite - or ability - for collaboration. Whereas the bank controllers used to discuss with each other how to value illiquid derivatives instruments in the 1990s, this crisis is unfolding with minimal interaction between banks.

Worse still, the losses are emerging with unusual speed. That is partly because instruments such as collateralised debt obligations tend to have a shelf life of three years or so, which means that losses crystallise faster than on, say, a Japanese corporate loan. Moreover, accounting reforms are forcing many institutions to mark their assets to market, however difficult this might be.

But this mark-to-market process is not being applied in a uniform way: though parts of the financial world are using it, others are not. Thus, senior bankers currently find themselves in the worst of all worlds: investors have just enough knowledge about the losses to feel scared but not enough information to think the worst is past. In this situation, partial transparency can actually be worse than no transparency at all.

Of course, eventually this fog and fear will clear. After all, even Japan produced clarity in the end (and, for the record, though the losses there were more than $100m, they were also less than the gloomy $1,000bn-odd projections.)

The Financial Times

No comments: