Felix Salmon

Why analysts hate putting out sell ratings

Felix Salmon
Apr 25, 2011 23:32 UTC

Herb Greenberg kicked off an interesting discussion today when he said that it took “a lot of guts” for Wedbush Morgan analyst Michael Pachter to go on CNBC today with his bearish view on Netflix, ahead of its earnings announcement this afternoon. David Wilkerson has the details of Pachter’s analysis, complete with the context:

With the stock hovering in the $80 range in April 2010, Pachter downgraded the shares to underperform from perform, with a target price of $73. Even after seeing the stock zoom into the $240s, Pachter is sticking to his rating, and a target price of $80.

That’s a very aggressive price target: Pachter’s saying that Netflix is set to lose more than two-thirds of its value, despite having a subscriber base of some 23 million Americans. But is it really a bold move for Pachter? Does it take a lot of guts for him to say this, and if so, why?

Part of the answer to that question is buried in a market symmetry: the long bias of investors is matched by a bullish bias on the analyst side. Most investors are long-only, and even the ones who go short tend to follow a 120/20 or 130/30 strategy: they nearly always have many more long positions than they have short positions. As a result, the market as a whole is already biased against anybody with a “sell” recommendation.

And when it comes to screamingly-hot stocks like Netflix, that’s even more true. Such stocks tend not to be held as part of a long-term portfolio of diversified names; they’re held by momentum traders who want to buy high and sell higher. These people tend to be pretty emotionally invested in their trade and in the sentiment which is driving it, and they can be quite aggressive towards anybody who might damage that sentiment.

On top of that, Wall Street does have a habit of boiling everything down to a right/wrong duality: if you say that a stock will go down, you’re right if it does, and wrong if it doesn’t. The intelligence of your analysis, or the idea that all these things are probabilistic rather than certain, rarely even gets lip service. This is why you see so much technical analysis on Wall Street: it makes no intellectual sense at all, but it works just as well as — or even better than — fundamentals-based analysis. (Which, admittedly, isn’t saying very much.) And that’s all that matters.

There’s also an inability for anybody to appreciate the difference between “buy/sell” on the one hand, and “long/short” on the other. A “sell” rating is not the same as a recommendation to go short. Selling your NFLX at $250 is a risk-averse move: you’re taking a volatile and overvalued stock, taking what are probably enormous profits, and saying that you’d rather sell too early than too late. Shorting NFLX at $250, by contrast, is a highly risky move which can hurt you very badly indeed. Yet when an analyst says “sell”, everybody starts talking about his “short position”, and saying things like “how’s your Netflix short coming along, Mikey?”.

One thing that’s certain about a “sell” rating, of course, is that it’s going to annoy the management of the company in question. And this is where the distinction comes in between issuing a “sell” rating on a privately-circulated report, on the one hand, and loudly proclaiming your analysis on CNBC, on the other. The television audience isn’t just sophisticated investors: it’s a much broader public than that, and corporate management hates even thinking about the idea that their company is being trashed in front of a huge audience. So if you’re going to present a bearish case on TV, be prepared to lose much if not all of your access to management.

If you make a very public bearish case, on TV, for a very visible consumer stock with lots of name recognition, that’s about as far as you can stick your neck out if you’re an analyst. That’s what Greenberg was referring to: yes, on one level it’s Pachter’s job to have an opinion on Netflix and be willing to be called on it if CNBC calls him up. But it’s easy to see why most analysts try hard never to put themselves in that kind of situation. “Clients will always be pissed if you’re wrong,” Greenberg told me in a short conversation this afternoon. “A long guy against the crowd is a value investor. A short is taking his life in his hands.”

Finally, if Netflix does fall dramatically from here — if Pachter’s call turns out to have marked the very top of the Netflix market — he’s still not going to be a hero. He’s been wrong for long enough, now, that people can just say that he was sure to be right eventually. And they’ll probably credit their own perspicaciousness if they followed Pachter’s advice, rather than giving him the credit he deserves.

So well done to Pachter for sticking to his convictions. I hope he doesn’t expect to gain much from them.

Update: I just got a great email from someone who wishes to remain anonymous:

a. In Pachter’s specific case, he’s been down on Netflix for quite a while, so it’s not exactly a significant amount of courage to go out and keep reiterating it. Whatever bridges there are to burn with a company are already burned by now, and, he’s just at the point of re-re-re-reiterating or capitulating.
b. The incentives for a guy at a (relatively) smaller shop can be a bit different – you make a name for yourself by standing out more. Whereas the bigger-shop guys have more of an incentive to limit how crazy their calls get.
c. From my perspective, there is minimal need for a research analyst to actually get his calls right. The majority of his compensation is driven by how useful he is to institutional clients. Fidelity is not going to outsource their investment decisions to a bank (for the most part), so they don’t really care what your rating or target is. They just want you to know everything there is to know about a company when they call. Some analysts are good stock pickers and they end up at hedge funds eventually. Many successful analysts are not.
d. As an outgrowth of c, I think it can be a disservice to retail investors to put some analysts on tv. They’re better at setting the scene than telling someone who owns a couple hundred shares of a stock what to do with it (setting aside whether that person should be holding/trading individual stocks to begin with).
e. Also, an analyst has to have a rating on every stock he covers. But he might only have a strong opinion on one, or a small handful. Good luck getting that context if you aren’t a paying client.


At the end of the day, the world is 100% net long with itself. Only the exchange rate versus cash changes at the margin.

Posted by DavidMerkel | Report as abusive

Wall Street analysts deserve a “sell” rating

Felix Salmon
Jan 21, 2011 17:37 UTC

Jennifer Saba makes an important point about milquetoast tech analysts today: faced with huge news about shake-ups in the CEO position at both Apple and Google, precisely zero analysts asked any questions on that subject during the companies’ earnings calls.

In the case of Google, the company went out of its way to open itself up to such questions: it started its call with comments from Page, Schmidt, and Brin, all of whom talked explicitly about the changes, and then invited questions for them specifically before returning to old-fashioned earnings-call questions. The transcript, when you read it, is almost hilarious:

>>Patrick Pichette

Jay, if you don’t mind, what we will do is we will set up two sets of questions for this afternoon. We will just take a few questions now because we have Larry, Sergey and Eric, and so we will take a few questions. And then after that, we will close that section of the call, let them run back — go back to work, guys — and then we will take the regular call as we usually proceed. So, Jay, if I can ask you to give us the instructions to take a couple of questions for Eric, Larry and Sergey.


(Operator Instructions). James Mitchell, Goldman Sachs.

>>James Mitchell

Congratulations on the movements, and congratulations on the results. I guess one question I had just stemming from the results rather than from the movements perhaps is when I look at the investments in 111 8th Avenue in New York City, do you feel that Google is now at a point where in order to continue facilitating the growth of the Internet that there will be a land grab for desirable physical locations?

It’s sad but true: faced with an open invitation to ask Google’s troika anything he wanted about the leadership of the company, Goldman’s star tech analyst instead simply said “congratulations on the movements,” and moved swiftly on to a question about Google’s decision to buy rather than rent its New York office.

There’s no good reason for this kind of behavior. Analysts are happy asking tough questions on calls and in their reports: it’s not like they are or should be scared of annoying senior management. What’s more, Google clearly wanted and expected questions about the C-suite changes.

But Wall Street analysts are much happier with spreadsheets than they are with anything human, and faced with the opportunity to ask flesh-and-blood questions, they get squeamish and retreat to their quanty ivory towers.

Which is one more reason, if reason be needed, to treat everything coming out of Wall Street’s research shops as fundamentally incomplete.


I might read an analyst for his industry evaluation; I would never read an analyst for his buy/sell recommendation. Other than general observations on broad issues, I don’t trust any of these guys with my money.

I’m not sure on the degree of market Efficiency, which is an enormously complex question possibly beyond answer until much later in this century, but I’m sure that by the time I read any recommendation the market has had time to digest the information. Bargains take dis-equilibrium on a substantial scale.

Posted by ARJTurgot2 | Report as abusive