FOOL'S EYE VIEW
Don't Buy Analyst Share Tips

By James Carlisle
August 1, 2002

Investment analysts have been getting a bad press recently, and for good reason. The evidence, particularly in the United States, is that they've been producing investment research designed to promote the interests of their own firms rather than the investors that read the reports. In other walks of life when people set out deliberately to deceive people, it gets called 'deceit'. Apparently brokers' analysts have been getting away with it because everyone expects as much from them.

Yet analysts, and the research they produce, also serve the important purpose of spreading information around the stock market, hopefully helping to create an efficient market. The trouble is that the research is liable to do more harm than good if it isn't independent and objective. So the Financial Services Authority has made it a priority to clean up the analysts' act. The publication yesterday of the discussion paper "Investment Research: Conflicts and Other Issues" is the first step in that direction.

Conflicts of interest

The problem that the analysts have is that they have to please all sorts of different people. Their bank's corporate finance arm might want them to talk up a particular company to facilitate a flotation or takeover (not legal, but it happens nonetheless) or at least they'll want the analyst to be nice about their clients. Even in the absence of a client relationship, it's in the analyst's interests to be nice, since companies are hardly likely to keep them well informed if they're always picking holes in the management's strategy.

Then there's the bank's stockbroking department that will want to generate dealing commission and therefore likes to see a preponderance of buy recommendations, which naturally have a wider audience. After that, you've got the big institutional investors each of whom would like to be the first to hear the news and gossip. And finally, lest we forget (and the analysts clearly do), there are the little guys like you and me.

Bias

All these conflicts combine to produce a significant bias in analyst recommendations. Where a bank has no client relationship with a company, their analysts currently produce about 45% buys, 38% holds and just 18% sells. That looks bad enough, but incredibly, this rises to 80% buys, 18% holds and 2% sells where there is a client relationship. Perhaps worst of all, the ratio of buys to sells peaked when the market did back in early 2000 (maybe it's a useful 'contrarian indicator').

It wouldn't matter so much if everyone gave these recommendations the respect they deserved (that is, basically, none), but unfortunately that doesn't seem to be the case. While institutional investors are apparently more interested in the detail of the reports and have the sense and experience to pay zero attention to the recommendation that goes with them, private investors are prone to taking it all rather seriously.

Suggested solutions

A system that allows or encourages analysts to present biased recommendations as if it's objective advice, on a sort of understanding that everyone knows it isn't, is clearly a load of nonsense. Apart from anything else, many private investors are totally ignorant of the bias and the circumstances that give rise to it. One suggested solution is to try to educate consumers so that they realise what nonsense it all is, but it must surely be better to remove the nonsense in the first place.

Another suggested solution is to continue with the current system of 'managing' conflicts of interest through the use of 'Chinese walls' (a nickname for the supposed separation between departments in the same firm), but it's this system that has created all the problems. Human greed will inevitably cause people to cross the divide.

It might be a straightforward 'can you put out a buy on this company' from Bill in the corporate finance department or it might be a lunch date and a gentle 'you know that we have xyz plc as a client -- I think it's an excellent company'. Most likely of all, nothing actually needs to be said. Analysts know who the firm's clients are, they know where their interests lie, they are manifestly not stupid and they are, as I understand it, human. At the very, very least, a system that might enable this sort of thing is never going to inspire confidence.

Calling a spade a spade

The only way to resolve the problem is to call a spade a spade. Whenever someone issues a report on a share, it should make clear where it's coming from and whose interests it serves. If a broker wants to send something out, then it should make it clear that it's 'marketing material' -- "Come to Bernie's Brokerage and we'll fix you up with one of these nice shares"

If a corporate finance outfit wants to sponsor a document plugging a forthcoming flotation, then it should be called an advertisement -- "Look out for the new WonderWidget plc, we reckon the shares will go up but we're obliged to point out that we're making stacks of cash for saying so and our opinion is therefore totally unreliable".

Only if an operation has no connection with these other activities should it be entitled to use words like 'research' or 'independent' to describe its reports. Rather like Which? magazine -- "In this month's independent research report, we put general retailers through their paces". Combine this with a level playing field for the spreading of information and we could be onto a winner.

Maybe we'd end up with a situation where brokers would live and die by the efficiency of their dealing (bringing down costs), the analysts would live and die by the quality of their research and the companies issuing shares in the primary market would live and die by the sense of the deals that they're doing.

Now it's your turn!

Right, I've got that off my chest, now it's your turn! The FSA has explicitly said that it's hoping to hear from private investors. So have a gander at the discussion paper and let them know what you think!