Thursday, May 15, 2008

Stock Ratings from Lake Wobegon

Today’s New York Times reports that Merrill Lynch unveiled a new rating system that requires their equity analysts to assign “Underperform” ratings to 20 percent of the stocks they cover.  As a lead-in to the article, the authors refer to the Lake Wobegon quality of stock ratings – most stocks are rated above average.  (For those of you who are not familiar with Lake Wobegon, it is the fictional town in Minnesota featured in Garrison Keillor’s Prairie Home Companion radio show.  Each week Garrison Keillor reports the “News from Lake Wobegon” and finishes with the tag line, “That’s the news from Lake Wobegon, where all the women are strong, the men are good looking and all of the children are above average”.)  The article goes on to cite Bloomberg data to the effect that only about 5 percent of all stock recommendations on Wall Street fall into the “Sell/Underperform” category.

 

Now, I could have a lot of fun with this, as clearly, far more than 5 percent of all stocks are going to underperform.  Even if you throw in a “Neutral” category, a normalized distribution would require far more than 5 percent “Underperforms”, even adjusting for the market’s tendency to rise over time.  But that would be too easy.  So what I want to do is examine the reasons “Outperform/Buy” so dominate the world of stock ratings and “Underperform/Sell” is so rarely seen.  After all, it’s not like all these MBAs suddenly become Pollyannas when they get assigned to equity research.  This is Wall Street we’re talking about here, so cynicism is not in short supply.  Clearly, a different set of incentives is at work.

 

Quite simply, if an analyst puts a sell on a company stock, the company hates him, the long only investors that currently own the stock hate him, the long only investors that don’t own the stock aren’t going to buy it, and the firm’s investment bankers get really mad at him.  Only the short sellers will like the sell rating.

 

Understandably, if a company receives a sell rating from an analyst, they don’t like it.  Ratings changes can occur for a variety of reasons – valuation, fundamentals, sector or industry changes; but companies usually don’t care.  A common reaction by companies is to cut analysts off from corporate access.  Much as we like to think that analysts are engaged in independent research, much of their information surrounding formal disclosure comes from talking to the investor relations contact at the company.  Cut off this flow of information and deciphering the balance sheet and cash flow statements becomes much more difficult.  The analyst will also find that the probability of their being able to ask a question early in a company conference call is remote.  Companies control the order of the question queue and they are not going to feature an analyst with a sell rating as the first questioner on the call.

 

Further, much of what the Buy Side values from the Sell Side these days is access to management.  Just try arranging a trip to company headquarters for your clients if you have a sell on the stock.  If you’re an investor relations officer with more requests for meetings than you can handle, think about trying to convince your management to take a meeting for someone who has a sell on the stock – the same person whose name your CEO mentions only in combination with an expletive.  There are better ways to spend your day. 

 

As to the Buy Side of the equation, if a firm owns a stock, they’ve made a commitment to it and have often invested a lot of research in it (not to mention that they often become emotionally attached to it as well).  Then along comes a Sell Side analyst and tells them they’re wrong.  The Sell rating may have already caused the stock to decline, plus he’s making either the Buy Side analyst, the portfolio manager, or both, look bad.  This is a situation that can cause conflict and strife.  Plus the commission flow from the Buy Side firm is likely to decline because they're mad at the analyst, which will antagonize the Sell Side firm's sales force and trading desk.

 

Finally, investment bankers really don’t like it when they’ve invested weeks, months or years trying to build a relationship with a company, only to have an analyst rain their parade by putting a sell on the company stock.  There may be a theoretical Chinese wall between investment banking and research, but don’t think that company managements give a hoot about that.  A sell rating stands a good chance of tainting the whole firm, thus jeopardizing all those juicy M&A and underwriting fees.

 

Is it any wonder there are so few sell ratings?  The incentives are all going the other way.  It’s just too painful to put out a sell rating, which is why you usually see them only after the stock has already collapsed or on companies that are small and don’t command much respect when it comes to fee generating abilities.

 

I’ve been thinking about alternatives to the way stocks are rated, but I’m going to save it for another post.  I’ve got to go – my broker just called with a great buy rated stock

3 comments:

Robert Boyd said...

Given this pressure on analysts, it seems like a quota of "underperform" might give analysts at least some of the incentive they need to be honest. At least they can go back to the irate IR managers or irked investment bankers and say, "Hey, what could I do? It's company policy that we give out X% underperform ratings."

Gary Vineberg said...

To Robert Boyd: I doubt that many managements will take comfort in being part of the 20% of stocks rated as underperformers. I've seen studies showing that sell-rated stocks often outperformed others. Why? Because analysts frequently downgrade stocks to sells after they've already blown up and their managements are contrite. Also, many analysts reserve their sell ratings for relatively insignificant companies that they haven't bothered to analyze properly. It is the fear of offending that governs analyst behavior in this regard, and putting a quota on sell ratings will do little to change that.

Robert Boyd said...

I agree with you. I admit I was putting the best possible face on it--it occurred to me that the situations you describe might indeed happen. Given this, how can we give analysts incentives to not engage in grade inflation? Or is it always going to be an unsatisfying compromise between access to the covered company and honestly about said company?