That "Free" Research You Get From Your Broker May Be Worth What You Paid for It

Conflicts of interest can be hard to spot from the inside, while they seem obvious to those on the outside. That's, of course, one of the things that makes such conflicts so complicated.

I've been thinking about conflicts of interest, and how to structure organizations to reduce their frequency, since reading Saturday's New York Times column by James B. Stewart on analysts' bullish recommendations on Apple stock, long after it started its current slide (it is now down by about one-third since its September 2012 peak). The entire column can be found here.

As Stewart noted, we have been aware of conflicts of interest in the banking and brokerage arenas for some time. A decade ago, "There were some notorious examples of analysts who curried favor with investment banking clients and potential clients by producing favorable research, and then were paid huge bonuses out of investment banking fees."

Those lapses seem outrageous to outsiders, but probably didn't to those on the inside at the time. It was just "how the business works". And we know that humans are stunningly bad at knowing how easily they are manipulated, how quickly they respond to even small rewards (hello, "gimme" caps!). So if you had asked an analyst at the time whether the fact that her bonus was paid by investment banking fees influenced her recommendations, she would have been insulted at the very thought.

Since then, thanks to Congressional and state investigations and the passage of several laws, including Sarbanes-Oxley, "investment banking and research operations were segregated. Conflicts had to be disclosed, and research and analyst pay was detached from investment banking revenues...."

So far, so good.

So how could analysts still be saying, "Buy, Buy," about Apple, when the stock is slipping? What else is going on?

A key problem, according to Prof. Stuart Gilson of Harvard Business School, is that "research is funded through the trading desks."

In other words,

If you're an analyst and one way your report brings in revenue is through increased trading, a buy recommendation will do this more than a sell. For a sell, you have to already own the stock to generate a trade. But anybody can potentially buy a stock. That's one hypothesis about why you still see a disproportionate number of buy recommendations.

But fixing all the conflicts -- even if that were truly possible -- might not solve all the problems. Analysts, it appears, fall into many of the same traps we ordinary investors do, such as extrapolating past performance into the future.

Moreover, like many of us, analysts "have a tendency to tell their audience what it wants to hear."

How can an organization fix that problem?

Carlo Besenius, "the only analyst who even came close to calling the peak in Apple's stock", isn't employed by an investment bank or a brokerage firm. Instead, he runs his own firm, founded a decade ago after many years in brokerage-house research.

I'm paid based on performance.... I have to go to my clients and explain why they should pay for my research when they can get it for nothing from the firms were they pay their trading commissions.

You know what they say about advice? It's worth what you pay for it....