There's a mix of opinions when it comes to the value of stock research. Some still see it as a force in the market; others think analysts are dinosaurs. After looking at 44,000 plus changes in recommendations, Oya Altinkiliç of the University of Pittsburgh and Robert Hansen of Tulane University have concluded that analysts really don't move stocks much--about 0.03 percent up or down.

According to the Financial Times, previous studies found that stocks move up or down about 4 percent in tune with changes in recommendations. This study looked at share prices 20 minutes before and after the issuance of a recommendation, aiming to control movement linked to other news events. It concludes that analysts' revisions are "typically information-free." Ouch. You could quibble with the study. 

This is something of a challenge for the industry. Frankly, the business model for research is still a bit iffy. The sell-side has decoupled research from investment banking, and that has really undermined the business case for research. Independent stock analysis never really took off and stands to take another hit now that the "global settlement" is expiring. Top firms will no longer be required to provide independent research after July. 

Frankly so much research seems commoditized now. It will be interesting to see how Bloomberg deploys the equity analysts they intend to hire. You can imagine them creating custom data cuts that might be interesting in real time. That raises the bar for all analysts. 

It's fair to say that the collective cachet has been reduced now that they're not a P&L force as they were before the "global settlement." At the same time, they can serve marketing purposes. One way to see them: As PR mechanisms, driving brand recognition via TV appearances and the like. That certainly has value. We may also see more specialized boutiques built around personalities, like Meredith Whitney's new firm. 

Some people are convinced that more really good analysts will migrate to the buy-side, where they'll perhaps benefit from a compensation model that rewards their ideas.  

All that said, I do think there will be some forms of research that will prove so valuable that clients will be willing to pay for it. It will have to be of a real-time nature that utilizes data far beyond the stock market data and really more on industry and other data. Goldman Sachs and Merrill Lynch have moved in this direction with their independent platforms. We'll see how they fare. - Jim