When analysts say sell do they mean it?

A few years back, after the implosion of the tech bubble and the scandalous crash of companies like Enron, the investment banking analysis community found themselves well and truly in the dock.

Here were a group of fabulously rewarded opinion-formers, who had apparently led investors into an expensive trap by goading them into buying worthless companies, hiked to the rafters during the dot com boom.

The analysts screaming "Buy!" "Buy!" "Buy!" all the way up were actually conflicted. The more of a frenzy they could stir among gullible investors, the more fees their bank employers could earn floating companies in the technology, media and telecoms space.

It was a racket and it lead to a swinging regulatory crackdown, with Wall Street companies fined hundreds of millions of dollars in punishment.

I wonder now – in the context of the present crisis paralysing New York and London – whether we have gone as far as is possible the other way. It is as though the analytical community have shaken off any requirement, whether direct or indirect, to tow any sort of party line set by their investment bank.

It is as though research analysts, as they pump out note after note to their clients, can say just about whatever they want, applying whatever level of aggression suits them along the way.

It's all about the use of prose. Consider a research note published by Goldman Sachs last week on America International Group, the huge US insurance combine.

Now, Goldman has what it calls a "neutral" investment recommendation on AIG – saying in effect to its clients that it doesn't have a strong view on whether the shares are likely to outperform or underperform the market as a whole.

But look at the title to last week's note: "Don't buy AIG: potential downgrades, capital raise on the horizon."

If that's a 'neutral' investment view, this correspondent needs a new dictionary. Now, to be fair to the analyst who authored AIG note, Thomas Cholnoky, the Goldman research view is that the near-term outlook for this insurance giant is so uncertain that he is not in a realistic position to offer firm investment advice, other than warning investors to stay away from the situation.

But consider more of the prose – the language applied by Cholnoky to get his point across. Here are the subheadings from his note:

– "A dangerous balance sheet posing as an inexpensive entry point."

– "There's nothing to be feared except fear itself… and mortgages."

– "Between a rock and a hard place."

– "Raising capital: Ultimate number too difficult to quantify."

– "Calling all cash: Please report to AIGFP." And so the tone continues for about fifteen pages.

Again, to be fair to Cholnoky, the analyst has produced a very detailed analysis of AIG's potential exposure to problematic asset-backed securities.

And the sheer size of the losses mounting up across the banking and insurance sectors in the United States and Europe absolutely supports the Goldman position here.

But I still finding the sneering edge to the attack shocking – even if, as a journalist, it also entertains me.

Consider this prominent statement: "The central tenet of our 'Don't buy AIG' argument is simple: the intricacies of AIG's business are so complex that management may not even know the extent of the company's ultimate exposures, let alone losses… thus, if management cannot accurately assess its ultimate exposures or losses, then how can one expect the rating agencies to do so?" Don't forget: AIG is one of the largest insurance groups in the world and therefore a key player across large parts of the financial sector, while Goldman is a top grade brokerage house that makes its money dealing with firms like… AIG.

Goldman's top executives, in reviewing the views of their own analysts, must simply hope AIG executives have pretty thick skins – and understand that Cholnoky is simply doing his job, albeit enthusiastically. Now, there is a different way you can approach this new genre of aggressive, personalised investment research. Fawning, over-zealous and comically enthusiastic investment research during the dot com boom actually represented screaming advice to 'sell'.

The extreme nature of what analysts were saying turned out to be the clearest warning that they were wrong.

Could it be that the extreme nature of the research we are now seeing published is the signal to 'buy'?

If Goldman Sachs are saying that AIG are near ruin and led by management that can't see their own likely losses, is that such an aggressive statement that really we should be thinking that the rout in the financial sector is bottoming out?

Recommendation: neutral. This correspondent simply does not know.

-- Paul Murphy is an Associate Editor with the Financial Times