Tuesday, September 16, 2008

Analysts Continue To Bewilder

How can investors trust analysts who can't tell the difference between a company that's solvent and one that is worth virtually nothing to shareholders. We've seen other examples here where analysts appear fraudulent in their recommendations. The latest example? That of Lehman Brothers (LEH).

Yesterday, the company filed for bankruptcy, and the share price lost 95% of its value. However, rumours have persisted for months that this bank is due to fail. The company has been desperately trying to sell off assets to raise cash to remain solvent, but last week it appeared their last-ditch efforts had failed.

Yet on the Friday (Sept 12th) before the imminent bankruptcy, here is a summary of the recommendations of the 19 analysts that were covering LEH according to Thompson/First Call:



There is not a single 'Sell' rating on this chart! So are the analysts just guessing? One would think that if they were, then surely at least one analyst out of nineteen would rate this a sell. Were some of these ratings made months ago? Perhaps, but they are nevertheless still current (and not listed as suspended). In fact, 3 days before the would-be bankruptcy was announced, both Citigroup and Argus downgraded LEH from 'Buy' to 'Hold'. Hold?! Not 'Sell'? Does this industry have any credibility?

2 comments:

Anonymous said...

It's a game theoretic problem with the agency cost of the analysts.

If the analyst is right and the other analysts are right, the outcome for the analyst is good.

If the analyst is right and the other analysts are wrong, the outcome for the analyst is good.

If the analyst is wrong and the other analysts are wrong, the outcome for the analyst is good.

If the analyst is wrong and the other analysts are right, the outcome for the analyst is bad.

Here good means "you get to keep your job" and bad means "you're fired!". Consequently, following the herd results in job security. Being contrarian results in either instant fame or getting fired.

Corollary: More analysts are not more accurate.

Jae Jun said...

Basically, hold is just a very nice way of saying sell. After all, analysts don't want to jeopardize the possibility of losing business with their sell recommendations. It might come back to bite them in the rear.