I'm with Felix on this one. When I became an analyst a few years ago (I'm no longer one), my DoR showed me a survey from II that said buy side managers rank analysts' picks something like 10th (out of 10) on a list of the things they value from analysts. Items like access to management, industry knowledge, etc. ranked at the top. So despite some (I'd argue a few) buy side managers relying on picks, I hardly think its de rigeur.

Besides which, using changes in consensus estimates as an indicator--especially on heavily researched names--is a different thing from relying on an individual analysts view, in a "wisdom of crowds" sort of way.

One thing that hasn't been discussed here is the practical aspects (aside from regulatory): How *much* stock should an analyst own? A thousand dollars? What about one share? Or would you advocate a percentage of an analyst's investable assets? Makes a huge difference, no?

I still maintain that a better--and certainly much more practical--thing would be to publicize the impact of an analyst's recommendations, relative to an appropriate benchmark. How a hypothetical $1000 in each of 10 different calls compares over time to $10,000 invested in the relevant index, say. If you want to know how good a stock picker an analyst is--assuming that's relevant to you--that seems like the best way.

Dead on, performance should be comped to the most likely alternative investment baseline. When I was an analyst we always compared against the R2000 or SP600 for small/mid caps. But then your result should come as no surprise, since we all know it isn't venture (or mutual) *funds* where the big money is made, it's in fund *management*.

Where is my yacht, anyway? :-)

The issue is really dishonesty vs. incompetence, isn't it? Ownership is irrelevant to the former, as things can be gamed either way (e.g. talking your book). But trading restrictions create a different profitability situation between analysts and investors, so competence could be difficult to discern as well based solely on ownership.

Real-time measurement of performance vs. calls would seem to be the best solution, and frankly wouldn't be difficult to do--though the regulatory issues associated with verification could prove problematic

Would *you* be willing to invest in a stock if your shares were locked up whenever anything market-moving happened? Analysts aren't able to trade for days either before or after ratings changes. And even then must clear trades in advance and in most cases hold the account at their employer. Hard to make real money with those restrictions.

Hate to look like a luddite here, but I'm with tt and Erik. Skimming has it's value, but so does diving. Barely have time for either, frankly. I said 6 years ago that the primary survival skill in the 21st century is to be a good filter (or to know one, or have one), and it seems to be playing out. Let me coin a term: the "Balkanization of Attention". Where does it lead us?

This whole deal looks just like the house for sale at the end of my street. It's clearly a tear down, and buyers are accounting for demolition costs when valuing it based on the land only. But the sellers are clinging to the belief the house is worth something (since they've been living there) and so refuse to lower their price.

Yahoo's problem is that nobody else (including the market) values its business the way it does. Ballmer's problem is that there aren't any other suitable vacant lots.