What I’ve learned about short selling

The most disappointing thing that I’ve learned about short selling is that identifying frauds and scams is not as consistently profitable as it “should” be.  Identifying frauds actually isn’t the hard part.  The real issue is that fraud isn’t symmetrical.  While fraud reduces returns on the long side, actually making money by betting against fraud is much harder than it seems.  Here are some reasons:

1. The wealth isn’t exactly transferred from the longs to the shorts.  Short sellers have to pay money to borrow shares.   Much of the profit from short selling ends up in the hands of brokers, who earn large fees from lending shares to short sellers.
2. Short positions tend to be correlated because of the way they are marketed.  The parties lending out shares tend to be institutional investors who fall prey to the marketing tactics of small investment banks.  Being unable to ride out market conditions can cause short sellers to lose money.
3. Worthless companies often get taken over.  There is always some rich person or misguided CEO that will throw money at a fraud.

The seductiveness of short selling

Partly, some people are predisposed to short selling because they are “glass half empty” types.  These people are naturally pessimistic or cynical (e.g. permabears).  However, that’s not the full story as you can be a permabear without shorting stocks (e.g. by holding cash, long physical gold, etc.).  There are also optimistic people who short stocks as part of a market-neutral or long-short strategy.

For me personally, short selling has appeal because short selling is fairly objective.  Either something is a fraud or it isn’t.  There isn’t a lot of grey area in between.  Contrast this with identifying great companies.  Airbnb’s business model seems like a terrible idea because it involves renting out your home (your safe space) to strangers on the Internet.  Society teaches us not to trust strangers- more so if it’s strangers from the Internet.  Here’s what Paul Graham, airbnb’s seed investor, had to say about Airbnb’s business plan:

In fact, when we funded Airbnb, we thought it was too crazy. We couldn’t believe large numbers of people would want to stay in other people’s places.

And also:

We thought Airbnb was a bad idea. We funded it because we liked the founders.

In practice however, Airbnb turns out to have been a good business idea.  Finding great companies is not very clear-cut.  It’s also highly dependent on luck.  Paul Graham notes that “there is probably at most one company in each YC batch that will have a significant effect on our returns, and the rest are just a cost of doing business”.  Out of the 1900+ companies that Paul Graham’s Y Combinator has funded, there have only been a few successes: Dropbox, Airbnb, and to a lesser extent Reddit and Stripe.  (See Y Combinator’s website and this for a full list of Y Combinator companies not in stealth mode.)  The success rate has been fairly low.  In contrast, short selling is not so wishy-washy.  Brazen stock promotion (e.g. shill articles promoting a stock), money laundering, security law violations, and accounting fraud are all fairly objective occurrences that can be uncovered.  It’s also possible to figure out if the people involved in a stock mastermind pump and dumps as their career.  These are well-defined truths that can be uncovered.

Market inefficiencies are difficult to arbitrage via short selling

Spotting awful stocks is really easy.  If you simply look at stocks with very expensive borrows, then you will find plenty of awful stocks.  This is the main reason why my Motley Fool CAPS account (glenn12345) is ranked #21 out of 69,795.  (See my 2013 post on my CAPS strategy.)  However, shorting stocks with very expensive borrows is not a great investment strategy because you’d have to pay the 50-100%+ borrow fee on many stocks.

It’s also possible to find crappy stocks with cheap borrows.  For example, there are plenty of junior miners, reverse mergers, and Chinese VIEs with cheap borrows.  I believe that virtually all of them are good short positions.  However, it’s still difficult to arbitrage these inefficiencies.  For example, a reasonable strategy would be to short every junior miner.  A reasonable proxy for that strategy would be to look at the TSX Venture (it is almost a pure play on promotional mining garbage).  As I’ve pointed out previously, shorting the TSX Venture would have been a good trade.  The chart below plots the TSX Venture (in blue) versus a physical gold ETF (in red).

However, there have been multi-year stretches where the TSX Venture has gone up rather than down.  Those periods are painful.

Good shorts are correlated

It seems like it would be a good idea to diversify among different short selling strategies by betting against junior miners, development-stage biotech, junior / independent oil and gas, reverse mergers, Chinese VIEs, etc. etc.  The problem is that actionable shorts require institutional bagholders that are lending out their shares.  (Retail bagholders almost never lend out their shares so it’s too difficult to get a borrow to take the other side of their trades.)  My theory is that bagholders are attracted to scams and frauds because of investment banks actively promoting awful stocks.  (To be fair to the investment banks, it’s likely the case that they promote scams because the scams find them and not the other way around.)

Perhaps a better way to think of it is this: fraud is an industry.  People do it for a living and go to industry conferences.  There are websites with useful information on how to execute a pump and dump because some of the helpers/accomplices use content marketing to attract clients.  The majority of my short portfolio consists of betting against the fraud industry’s output.  So, I’m not really betting against metals, cleantech, china, marijuana, biotech, independent oil and gas, etc.  I’m betting against the institutional bagholders who get scammed by the fraud industry.  Not surprisingly, the stocks that institutional bagholders buy tend to be correlated.

What I’ve found is that there have been timeframes where crappy stocks outperform as a group.  In 2013, having short positions was quite punishing unless you only shorted junior miners.  The 2016-2018 timeframe also hasn’t been that great for my style of short selling- my short positions have gone up slightly more than the S&P 500 index.

So, unfortunately, diversifying among different sectors doesn’t solve the issue of having somewhat long periods of underperformance.

Stock performance only loosely tracks fundamentals

A core tenet of value investing is that stock prices will ultimately fall in line with fundamentals.  However, in the world of mining stocks, this definitely isn’t true.  In my post on how mineral exploration has been a disaster since 2000, I argue that dubious acquisitions by senior miners have been propping up the fortunes of junior mining stocks.  While junior exploration companies have been massive destroyers of value (despite a commodities bull market), their fundamentals have been propped up by dubious acquisitions.  Worthless deposits aren’t so worthless when a greater fool steps in and pays hundreds of millions of dollars for that deposit.  In that sense, bad stocks don’t underperform as much as they “should”.

Quirks of having to borrow stocks

Having to pay money to borrow shares is an obvious drag on performance.  It’s why I haven’t tried backtesting performance to test ideas.

Another subtlety of short selling is the buy-in: if your broker cannot locate shares to borrow, your broker will force you to cover your short position.  Other market participants often engineer buy-ins to profit from short sellers, buying a large amount of shares and then transferring the shares into a cash account where they can’t be lent out to short sellers.  Crooked brokers may help facilitate this process by leaking information about vulnerable short sellers to preferred hedge fund clients like Jim Cramer (who wrote about this strategy in his first book Confessions of a Street Addict).  This phenomenon doesn’t show up on backtests; I don’t even know where to get that data.  So, getting good data on short selling strategies is hard.

One of the quirks of short selling is that the borrow costs can fluctuate a lot.  Here are historical borrow costs for Restoration Hardware (RH) using Interactive Brokers’ rates (the excellent website iBorrowDesk is where I got the chart from):

It seems like there was a massive short squeeze in RH last year.  For whatever reason, short sellers piled into the stock and created a situation ripe for a short squeeze (clearly not because of this blog because I started posting about RH in April 2014).  Weird and crazy things happen in the stock market.  Borrow costs are sometimes unrelated to future stock performance.

A very small number of positions will determine returns

It’s easy to believe that short selling is about spotting stocks that will go to zero.  In reality, finding the zeroes is quite normal.  It’s avoiding the stocks that turn into multi-baggers that has the biggest effect on returns.

Suppose for example that you bet against every Stratton Oakmont / Jordan Belfort stock.  This would have been a fairly reasonable short selling strategy because every single one of them was a scam.  Stratton Oakmont used ‘boiler room’ sales tactics to push stocks onto unsuspecting investors.  (You can watch the Hollywood movie inspired by Jordan Belfort or the other Hollywood movie about boiler rooms.)  Many (though not all) of the people involved went to jail.  Despite all of the scumbaggery involved, your short portfolio would have included the shoe maker/retailer Steve Madden (SHOO), whose share price went up by more than 20X.  (Note: Steve Madden, the founder of his namesake company, went to jail because of his role in the pump and dumps.)  The existence of that one stock in your short portfolio would have a huge impact on your returns.

In theory, you could analyze potential shorts to avoid stocks where there is a real business of meaningful size.  However, this doesn’t fully avoid stocks that go up several-fold.  Short squeezes and buy-ins can cause short sellers to be forced to cover their position after a stock has gone up a few or several times.  This happened with China Medical, where the wine mogul Peter Deutsch decided to transfer all of his shares into a cash account.  While the stock ultimately went to zero, it’s likely that many short sellers were forced to cover their short position at a loss.  (*Disclosure: I happened to make a very very small profit on my China Medical short when my buy-in occurred at a slightly lower price than what I sold the shares at.)

At the end of the day, a diversified short portfolio will have a few stocks with share prices that go up dramatically.  It’s similar to what happens when going long stocks, except that scummy stocks are highly volatile and the stock price movements are extreme.  The outliers have a huge impact on performance.  Unfortunately, this phenomenon increases short sellers’ exposure to luck if their portfolio isn’t that diversified.  A short portfolio of “only” 30 stocks will still be strongly affected by any outliers.

Luck versus skill

I would like to believe that investing is about skill.  Life would be so easy if my real-world portfolio resembled my Motley Fool CAPS account (glenn12345) where I am ranked #21 out of 69,795.  Unfortunately, the real world isn’t as easy as gaming a stock picking simulation where borrowing stock doesn’t cost anything.  Even in my CAPS account, I’ve noticed that my stock picks did not do much from 2016-2018.

I simply haven’t figured out a way to consistently outperform via short selling.  Frustratingly, identifying frauds does not lead to consistent outperformance.