Thursday, December 20, 2007

When negative earnings don't matter

(Felix Salmon) When it comes to earnings, Wall Street wants guidance. Here's the cri de c??ur from Societe Generale's Bill Cavalier, during the notorious conference call with Sallie Mae's Al Lord yesterday:

We're trying to put together projections here, Al. We're trying to figure out what your stock is going to be worth, and you have got to give us some guidance, you've got to give us some numbers... Can you give us some handle on what your stock is worth?

Lord, of course, told Cavalier to go elsewhere for his guidance; investors voted with their feet, and bid the stock down more than 20%. Most companies deal with such questions with a bit more panache, and their earnings generally arrive more or less in line with the expectations they themselves are largely responsible for creating. Investment banks have a harder time of it - their earnings are necessarily more volatile than those of the average widget manufacturer. And when they take losses, it seems that guidance and expectations are utterly useless.

Yesterday, Morgan Stanley's quarterly loss was 926% of the expected size; today, Bear Stearns's quarterly loss of $854 million is 380% of the amount that Wall Street analysts expected.

Remember, Wall Street knew there would be losses. It's not like no one told them there was a massive cock-up on the Morgan Stanley trading desk. As long ago as early November, Morgan Stanley was making noises about "negative convexity" and saying that its trade-gone-wrong had cost $3.7 billion. It's just that $3.7 billion looks positively elfin compared to the losses which were finally realized yesterday.

And yet the stock went up, of course. It's a bit weird, but when a firm's profits miss expectations by a fairly narrow margin, the effect on the stock price can be large. When a firm's losses miss expectations by an enormous margin, by contrast, Wall Street is more likely to shrug than panic. For some reason, it seems, the minute that earnings fall below $0, they no longer seem to be treated as a highly-calibrated indicator of a company's value.

No comments: