Carl Icahn has been known to go on TV. Photographer: Scott Eells/Bloomberg

New Investing Strategy: Talk Your Book

Matt Levine is a Bloomberg View columnist. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz and a clerk for the U.S. Court of Appeals for the Third Circuit.
Read More.
a | A

Guys, I have to tell you about an exciting recent financial innovation that you can use to get super rich and successful, are you ready for it? I am excited, I think this recent financial innovation is going to be big. Here it is:

We describe a recent financial innovation that allows limits to arbitrage to be sidestepped, and overvaluation thereby to be corrected, even in settings characterized by extreme costs of information discovery and severe short-sale constraints. We report evidence of shallow-pocketed "arbitrageurs" expending considerable resources to identify overvalued companies and profitably correcting overpricing. The innovation that allows the arbitrageurs to sidestep limits to arbitrage involves credibly revealing their information to the market, in an effort to induce long investors to sell so that prices fall.

Wait, hang on, the recent financial innovation is ... talking your book? Like, you make an investment and then you tell people about it and convince them that you're right? That ... I feel like ... I feel like people have been doing that for a while?

Nonetheless, that is a real quote (emphasis added) from a real paper by Alexander Ljungqvist of the Stern School of Business at New York University and Wenlan Qian of the National University of Singapore Business School (dated January 2014, NBER version here). Once you get past the "hahahahaha come on" aspect of the paper, it's actually pretty interesting, at least for its empirical information.

The paper is written against the sort of generic background assumption that short sellers keep prices honest by finding out negative information and then shorting the stock until the price gets to the "correct" level. As the authors point out, that doesn't really work; in particular, many stocks are very hard to short, because, mainly, of the difficulty of finding stock borrow. So "no-arbitrage" conditions don't really obtain: A stock can be overvalued, and short sellers can't pile in and correct it, because it's too expensive to short more than a few shares.

But one or two short sellers can come in, and short the shares that are available, and then make a bunch of noise, and convince the long investors to sell. The paper has a sample of 17 small firms -- unhelpfully referred to as "arbs" -- "who research companies, take short positions, and initiate coverage on at least two listed target companies" between July 2006 and December 2011. (So it misses at least one famous short report.) Those "arbs" targeted 113 listed stocks; on average, a targeted stock fell by 8.5 percent on the day the short seller released its first report, and 18.6 percent over the 60 trading days following the report. These drops were driven mostly by long investors selling, not additional short sellers piling in. Reports that contain new information, rather than just re-interpretations of already public information, have a bigger impact, as do new short sales from arbs with a previous record of producing credible reports.

I dunno, I guess that's all pretty intuitive, but now it is intuitive and science! And the science shows that credibility matters: If your noisy short selling has made people money in the past, people will listen to you now; if it hasn't, they won't. Your credibility and research are directly monetizable assets: The more reputable you are, the more money you make from noisy short selling.

Now, I am much less convinced that book-talking is a recent innovation. But there might be something to that. I have in the past occasionally poked gentle fun at Sahm Adrangi, who runs Kerrisdale Capital, a small hedge fund (mentioned in that Ljungqvist & Qian paper) whose strategy includes shorting stocks and then talking about them, and whose longer term plan "might be to turn himself into a brand, go on CNBC, get some gravitas, and start picking fights."

Adrangi is so devoted a book-talker that he e-mailed me a couple of weeks ago to tell me about a press release Kerrisdale was issuing on JGWPT Holdings Inc., a company whose shares Kerrisdale is long. In a bit of meta-book-talking, he pointed out the novelty of that release. Hedge funds don't normally (ever?) release press releases just to say that they own a stock and think it's undervalued: "Funds have said this in combination with traditional activist measures and recommendations," and of course they tend to be noisy on the short side, "but not if they're purely passive investors." Adrangi thinks that most funds don't prioritize persuasiveness, but that that's going to change:

Over time, you're going to have more and more funds, especially younger emerging ones, start to more actively talk their book, and leverage both the high quality of their internal research and credibility they've built in the marketplace via a brand, to create catalysts for their names.

If we are at the dawn of a new golden age of talking your book, why now? One obvious answer is that the techno-media landscape has changed. In the olden days you might try to talk your book by convincing your hedge-fund buddies to join your trade, or convincing sell-side research analysts to write something supporting your position: methods that rely on size and connections. (Or you might go on television, which, sort of the same.) The internet is a leveler of stock research as well as everything else, and makes it easier for small firms to publicize their trades. (Kerrisdale has a full-time social media coordinator, bless them.) The effectiveness of the message is less dependent on who you know and how much money you have, and more dependent on how persuasive you are and how right you've been in the past.

Also, funds might be more comfortable publishing their own research because their lawyers have lightened up. There's an obvious reason for hedge-fund lawyers to lighten up recently: The JOBS Act has made it much easier for hedge funds to advertise their services. Now, a research report on an individual stock is clearly not a "general solicitation" for investments in the hedge fund that wrote it. But the JOBS Act has had a broader impact than allowing hedge fund ads (of which there haven't been many). Instead, the act has mainly "permitted investment managers to feel more comfortable about branding their business more generally," with funds that are not actually advertising more comfortable doing branding stuff in public. Publishing research is a pretty plausible branding activity, with the nice side benefit that it can make you money directly.

On the other hand, one thing for lawyers to be worried about is that lots of people confuse "talking your book" and "market manipulation." (They are different: Market manipulation is talking your book untruthfully, more or less, but what is truth, etc. etc. etc.) Short sellers' targets sometimes go after them for market manipulation, though this seems less common than it used to be. Herbalife, for instance, has talked a lot about Bill Ackman's market manipulation, and hired fancy lawyers, but has not actually sued. And long investors who talk their books are almost never sued for market manipulation, since there's rarely any constituency to complain about stock prices going up.

One could also tell a story of the decline of sell-side research contributing to the rise of published buy-side research. If you're going to publish long careful research reports about stocks, you should get paid for it somehow. In the olden days, you got paid indirectly by equity trading commissions and investment banking business. With commissions declining, and with using research to win investment banking business disfavored, the alternative model of "get paid by trading the stocks you research" seems more appealing. (And free of conflicts of interest! Whatever!)

I guess it also fits into some sort of story of market efficiency, or meta-efficiency. Finding new information is hard; proving markets wrong is hard. If you can do it, you might as well squeeze all the value you can out of it. Just being right makes you money, but being noisily right might make you more money, and build your reputation so you can make money next time too. In a hard business, every little bit of edge helps.

(Matt Levine writes about Wall Street and the financial world for Bloomberg View.)

  1. From the paper:

    How easy would it be to short these companies' stock? As Table 2 shows, it would be expensive: one month before the first report is released, the average target company has a shorting fee of 4 basis points a day (10.6% annualized), which is in the right tail of the CRSP distribution (76th percentile). This reflects an unusually tight supply of lendable stock: on average, only 4.85% of shares outstanding are available for borrowing (39th percentile).

    A 10.6 percent borrow cost is high: much higher than the cost of borrowing Herbalife, for instance, just to pick one high-profile short.

  2. Roughly half of which were Chinese; 2006 to 2011 was a boom time for shorting U.S.-listed Chinese stocks.

  3. Among the evidence for that: Expensive-to-borrow stocks fell by an average of 7.2 percent on the day a report was issued; cheap-to-borrow stocks fell by an average of 8.6 percent, not a statistically significant difference.

  4. How this works:

    [T]o determine whether a report is likely to be credible, we examine each arb's prior track record, on the assumption that arbs with a stronger track record are more readily believed when they target a stock. We measure an arb's track record at time t as the rolling mean of the three-month cumulative abnormal returns of all his previous reports (issued at least three months before time t, to avoid look-ahead bias) that are independently confirmed by a third party (such as the SEC, the DoJ, or an exchange). Using all 332 reports in our sample, we then code a report issued at time t as more credible if the arb's prior track record produced profits (a negative rolling mean CAR), and as less credible otherwise. This yields 167 reports coded as more credible and 36 reports coded as less credible. Note that an arb's track record evolves over time such that he can gain or lose credibility depending on how accurate his reports prove to be. ...

    When a more credible arb is the first to issue a report on a target, the target's share price falls by an average of 12.5% on the report day, net of market movements (p<0.001). This is a significantly larger than the -4.5% average price fall for less credible reports, which in turn is not significantly different from zero (p=0.15).
  5. And whose shares, I feel obligated to inform you, are up over 20 percent since Kerrisdale filed its 13G.

  6. From that DealBook article:

    Bridgewater Associates, the $150 billion hedge fund, posted a video on YouTube a day before the new advertising rules took effect, featuring its billionaire founder, Ray Dalio, in a 30-minute lecture called "How the Economic Machine Works." Other funds are adding more information to their websites, and fund managers appear more frequently on business networks like CNBC and Bloomberg Television, their advisers say.
  7. From the Ljungqvist and Qian paper:

    The arbs also face the risk of being sued by their targets. In the words of the Wall Street Journal (June 14, 2006), "What was already a difficult and sometimes gut-wrenching strategy has only gotten harder in recent years. Nowadays, more of the companies that short sellers target are fighting back with lawsuits. They also mount sophisticated public relations campaigns against shorts." On a more positive note, the very real risk of lawsuits will, to some extent, keep the arbs from making claims they cannot substantiate. This, in turn, will make it likelier that their reports will be believed.
  8. But not never! This is one of my favorite lawsuits of all time.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

To contact the author on this story:
Matthew S Levine at

To contact the editor on this story:
Toby Harshaw at