Mostly at Bronte we short frauds or promotes. Stuff where the management says X but X is not true. Sometimes X is a stretch (a stretch made by promotional management). Sometimes X is patently false.
Either way we can describe what we do. We find things where the market is deliberately misinformed and hence comes up with inaccurate prices. We think (on reasonable grounds) that we are smarter than a deliberately misinformed market - and we often are.
We can answer the question "what is it that we see that others do not?"
Sometimes (rarely) we do valuation shorts. [We have a few valuation shorts now.]
In a valuation short we are working on the same information as everyone else has. This makes me uncomfortable. There is an arrogance in suggesting we can analyse the information better than anyone else. We find it harder to answer the question of what we see when others don't and hence harder to justify the position at all.
Indeed often we have no idea what the others are thinking and when I do not know what someone is thinking it is hard to justify (or test) the view that they are wrong...
I think this is the explanation for the cliché in short-selling circles that shorting valuation is a poor game.
Anyway - I am uncomfortable because I can't describe my informational edge and I don't like thinking I am smarter than other people (as opposed to better informed than other people).
That is all.
John
16 comments:
Is greater knowledge predictive of greater returns? You're not saying this explicitly, but you seem to implicitly. And I don't find that to be empirically true.
No: Asymmetrical knowledge is indicative of greater returns.
The most obvious source of asymmetrical knowledge is insider information - and it works a treat until you wind up in prison. Just ask Raj.
But asymmetrical information can come from other sources too. For instance you can detect fraud (and hence know things others do not). Or you can know what is going on in China better than someone in say Wisconsin and hence know that a certain industry is about to be destroyed by Chinese competition.
I find fraud detection to be my best source of asymmetrical information (and unlike Raj/Galleon it is legal).
But there are others.
John
" Asymmetrical knowledge is indicative of greater returns"
I wonder what CMEDY, UNXL, LOTE, Autonomy, etc. shorts have to say about that?
Greater knowledge is necessary, but INSUFFICIENT (it would seem) for repeatable excess returns.
Detail, and logical precision matters...
John, how do you justify a "valuation long" then? Not every position has to have an "edge" although it helps.
The main reason I avoid pure valuation shorts is because an irrational valuation can persist and worsen, and unlike a long where capital return can start to make up the difference, there's not necessarily any correction mechanism.
The thing about longs - as long as the company makes and distributes profits the longs on average win. Whether they win big or not - who knows.
And the asymmetry is in your favor.
And you are not obviously betting against anyone - just with them.
But I generally find it harder to justify my long positions than my shorts.
I am right less often too.
J
Internet stocks were screaming valuation shorts for years, but everybody who tried lost his shirt. Only when Boo.com went bust in May 2000 everybody suddenly realised that the game was up. And remember that Bill Gates once said to Warren Buffed that Eastman Kodak was toast? It ultimtely was but to short it immediately would have been very expensive. So once again timing is everything!
Hi John,
I'm a bit confused as to the distinction you're drawing here. Are you not looking at the same information set as everyone else with your fraud and promote shorts also?
You've described fraud shorts in the past where you've called B/S by simply looking at the financials (also bringing to bear, of course, vast experience looking at different companies and industries that allow you to call B/S better than most).
How is the ability to detect fraud in this manner any different from an ability to detect overvaluation? Both rely on some combination of better smarts; better information; and a better ability to draw on past experience and see the bigger picture.
The point you make in your comments, about longs winning overall, is an entirely different issue and one I agree with.
LT
The corollary is also quite profitable: being able to identify legitimate Chinese companies, which other investors suspect of being frauds.
I would argue that this approach is more profitable than shorting frauds - especially now that most Chinese frauds have already been outed. Have you seen the number of Chinese stocks that have gone private over the past year or so?
Case in point: IDG still has never invested in a fraudulent company - and I'd be willing to bet that they never will.
That is the asymmetrical knowledge that one could exploit, if one were to purchase stocks such as ASIA, BSDGF, CBPO, CEDU, CYOU, DANG, HMIN, HTHT, JRJC, KONG, MEMS, RDA, RENN, SFUN, VNET, etc.
Surprising to see you think a short book of promotes/frauds and valuation will outperform a short book of only promotes/frauds provided you have a large number of frauds/promotes. Perhaps you are playing the China macro cycle?
Or being able to identify legitimate nutritional supplement companies, which other investors accuse of being pyramids.
In the past you've called yourself a value investor.
Does that not imply that you have a view on the value of a company? And if you're investing in it your view must necessarily be different than other investors? And all based on the same information?
One wonders if timing (over-valued shorts can stay over-valued for a long time) -as some others commented- is not really the bugbear for valuation shorts?
Taking this a step further, could it be that looking for that informational edge you find so valuable in shorts, is hindering your search for longs?
John, once you've correctly identified a fraud/promotion short and put the position on, how do you think about covering? It sounds like there are different shades of grey regarding fraudulent activity/misleading information from these companies. IS the aim always short to (close to) zero? Or how do you determine to cover beforehand?
The key question, pardon my arrogance in framing this, is why would one want to short? I am not aware of any "shorting circle" - of its existence in the first place or its wisdom or gossip - so this is more of a no longer young fundamental credit analyst view
Even when it is a fraud, why would you want to short at the "current" level? For debt, it is manageable but for equity, the stock's upmove can singe and scar one for a long time.
So setting aside the broader market risk, one has to focus on risks to the upside explicitly flowing from company's performance (with which one as a short seller disagrees), risks to the upside from more buyers getting into the game (even at the risk of ridicule, I would believe that technical analysis can help in timing a short) and possible idiosyncratic risks – like the Chinese founders taking the company private, or someone wanting to acquire the company.
I would like to go back to First Solar (FSLR) which you've discussed extensively a few years ago. From my understanding at that time and what I remember of it now - we could call it an economic short or valuation short. Reading up your thesis at that time firmed up my negative views (I could never articulate the way you did). I remember that you took off the short much before it played out.
So the 2nd question is when do you want to get out when it does not play out as you anticipate.
The risk in valuation shorts, at least in my mind, is that we do not formulate scenarios clearly enough for the thesis to play out. We also run the same risks in a fraud short. (Chapter 4, Shortselling Is Not for Sissies, Hedgehogging by Barton Biggs). Some would want to call fraud shorts also has valuation shorts.
Some of the comments made by Jim Chanos and David Einhorn are in my notes and are something that I keep referring to whenever I'm convinced enough to recommend a short. To quote DE some extracts "for the most part, we avoided the damage in the short portfolio by refusing to sell short anything just because its valuation appeared silly. ... We concentrated on selling short companies with high evaluations combined with misunderstood fundamentals and deteriorating prospects. As always, frauds were preferred." His comments on Chemdex is also worth reading. JC in few of his earlier interviews mentioned that he did not short the Internet bubble because there was no debt. To me existence of debt is very important.
Probably TSLA after much of the current gas/fluff in the market ebbs away would be a good time to short. But I would wait for an impending "negative" event.
In valuation shorts:
If you are short a great company with an really expensive stock... and the intrinsic value of that company is growing rapidly... you can get into big trouble.
Hi John,
Thanks for the article. We have a similar approach, and rarely, (if ever) do shorts that (we think) are strictly driven by valuation. In fact, when we do – admittedly – our hit-rate falls to 50/50, or worse. Our accuracy in “frauds and promotes” is much higher.
However, you got me thinking.
Sometimes it is a fuzzy line between a market that is “deliberately misinformed” and a market that is deliberately ignoring disconfirming information. In the former case you have a “fraud”, and in the latter it is (more often than not) a “promote”.
The “deliberately misinformed” can be an outright fraud (Lernout & Hauspie), or it can be more subtle. It can be the management team that conveniently switches from GAAP to non-GAAP reporting (e.g. Salesforce.com), or overhypes the potential for efficiency gains for thin film without considering the similar is possible for polysilicon (e.g. First Solar in 2008).
In these examples (the FSLR and CRM cases), I actually didn’t/don’t think there is fraud per se, but there is definitely enough management promotion to make the case that the market was/is being deliberately misinformed.
We aren’t short FSLR anymore, but we are short CRM today. This is partly because this ringleading promotion by Marc Benioff has not only redirected the herd toward the wrong information, but because it has created a circus-like atmosphere where the sell-side analysts (and Mr Market) are deliberately ignoring information which might disconfirm their long thesis.
I could write a book about how this manifests at CRM, but that is not the subject or your article or my comment. This behaviour does, however, result in a wild valuation (we think, at least) for the shares. CRM (and others like it, Tesla?, Ocado?) tend to have experienced periods of high revenue growth and are unprofitable. Thus they are subject to relatively subjective valuation metrics, and consequently are more affected by sentiment than others.
So here I am, short something where there really is no asymmetric information, where Graham Smith (CRM CFO) is extremely careful to disclose (but not emphasize) all of the option granting and insider sales and pro-forma to GAAP reconciliations, and where we just think there is information which matters whereas others don’t. Thus, in these “promote” cases, the information out there can symmetrical, but the interpretation of it is asymmetrical.
And isn’t that what gives rise to a “valuation short”? FSLR may have been a short at $275, but it sure as hell wasn’t a short at $12. Green Mountain may have been a short at $100, but it sure wasn’t at $20. Netflix may have been a short at $270, but it sure wasn’t at $60, and in fact may again be one at $220. CRM may be a short at $44 (we hope at least), but will it be at $15? Probably not.
So, unless we are talking about frauds and bankruptcy (and I don’t think we are in any of these mentioned), then – at the end of the day – aren’t all of the expected returns to being short here driven by valuation?
Genomma Lab
Post a Comment