Florida Professors Experimented With a Little Naked Short Selling

Naked short selling does not appear to be very lucrative, unless these professors were doing it wrong.

"Naked short selling" is the sort of thing that people sometimes get very excited about for very nebulous reasons. It seems like the sort of thing a movie villain would do; it sounds terrible and scammy and also I guess sort of sexy? But it is not! It is none of those things. It is weird and boring and technical. But since we're talking about stock borrow anyway today, let's discuss this Securities and Exchange Commission naked short selling case from Friday and maybe figure out what naked short selling actually is.

Here is the settlement. There were two Florida State professors -- Gonul Colak, who teaches finance, and Milen Kostov, who teaches engineering 1 -- who each set up accounts at different brokerage firms and then sold a bunch of options to each other and other people. Then Kostov would exercise a deep-in-the-money call option written by Colak, resulting in Kostov being long the stock and Colak being short it, without Colak ever having arranged to borrow it. Colak's broker required him to rectify that by borrowing (or buying) and delivering the stock, but instead there'd be a whole 'nother flurry of option writing and they'd never deliver the stock, leaving Colak, as they say, "naked short": He'd sold the stock without arranging to borrow it first. 2

Here is a fun statistic:

Respondents sold approximately $800 million worth of call options and purchased at least $1.2 billion worth of common stock in over 20 issuers. Over the course of their scheme, Respondents reaped trading profits of approximately $420,000 on an initial investment of $100,000.

I don't know how the SEC is using the word "worth" there -- probably not "market value" -- but in any case, on $2 billion of trading activity, measured however the SEC measures it, these guys made $420,000. That's a profit of about 0.021 percent. Over almost two years. I think my checking account pays more than that.

Obviously, that's the wrong way to measure; that $2 billion of activity didn't really take much capital or involve much risk. (Except, you know, the risk that the SEC would catch them, which it duly did; they're paying some $670,000 to settle.) The trick is that, abstracting away from the flurry of offsetting options trades -- and, really, I recommend that you abstract away from them -- nothing was happening. Colak was short stock to Kostov. If the stock went up, Kostov made money, and Colak lost just as much money, and vice versa. But they were in it together, and at the end of the day, they'd agreed to split whatever profits either of them made -- 68 percent to Colak for funding the trade, and 32 percent to Kostov for executing it.

So ... what profits? Nothing was happening, so how were they generating money? Well, here:

When establishing their options positions, Respondents selected options of hard-to-borrow securities in which the price of the put options was higher relative to the price of the call options. Normally, the price of the put and the call will be in parity; however, when the stock associated with the options is hard-to-borrow, the higher cost of borrowing the stock is incorporated into the price of the put. By writing pairs of options in which the price of the put options was higher relative to the price of the call options, Respondents generated trading proceeds in excess of the proceeds that they would have been able to earn writing put options that were not associated with hard-to-borrow securities. The excess proceeds, which were derived from the underlying securities being hard-to-borrow, should have been offset by the cost of instituting and maintaining the associated short position that resulted from writing deep-in-the-money calls, for example, by effecting bona fide transactions to purchase, or borrow, the shares for delivery by settlement date.

Or if you don't like all the stuff about options, think of it this way:

  • A hard-to-borrow stock can be loaned out to short sellers for, I don't know, 20 percent a year. 3
  • If you own the stock you can make 20 percent a year just lending it out (but you have price risk).
  • If you're short the stock you have to pay 20 percent a year to borrow it (and you have price risk).
  • If you're long the stock in one account and short the stock in another, you have no price risk, can get paid 20 percent for lending, and pay 20 percent to borrow, leaving you with a nothing.
  • But if you're long the stock in one account and naked short in another, you have no price risk, get paid 20 percent for lending, pay nothing to borrow (because you're just not borrowing), and are left with nothing but a free 20 percent return. 4

Or whatever the numbers are. Their returns were somewhere between 0.021 and 420 percent over about two years so, you know, a wide range there.

Here's the point: "Naked short selling schemes," as the SEC calls them, are about making money on borrow costs. It costs money to borrow stocks to sell them short. If you don't pay that cost, you're cheating: You're getting something (stock borrow) for free, when everyone else has to pay for it.

Here's what "naked short selling" is mostly not about: manipulating stock prices. Destroying companies. Betting against companies. If you just want to bet against a company, you can go and borrow its stock and sell it short. You go naked short if you want to save on borrow costs.

Obviously you can do both at once, but there's no particular link. Most short sellers don't go naked short because -- well, one, because most people don't like doing illegal things (and naked shorting is illegal in the U.S. 5 ), and two, because short sellers in particular often like to bang on about how evil and criminal their targets are, and it doesn't look good if you're yelling about how criminal your enemies are and it turns out that you're breaking the law yourself.

And on the other hand, many (most?) naked short sellers aren't really short -- as Colak and Kostov weren't -- because, if you're just scamming some free money, why take the risk that a stock you're short will go up? Just be naked short in one account, long in another, and take your ill-gotten winnings without the risk of betting against a company.

There are those who think that naked short selling "doesn't have a very good case for illegality behind it." I'm not completely sure I agree, but it is true that the rules against naked short selling are what create the profit opportunity. If naked short selling was just allowed -- if you could always bet against a stock without borrowing it, just by promising that you'll deliver it in the future -- then the borrow costs for all stocks would be zero. You'd never need to pay to borrow a stock, since you'd never need to borrow a stock. 6 There would still be short selling, and much of it would be naked, but it would only be motivated by the fundamental desire to bet against a company. 7 Naked short selling schemes wouldn't exist, because their source of profits -- borrow costs -- would disappear.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.
  1. According to the SEC, "Kostov came up with the trading scheme at issue here, and Kostov presented the idea to Colak." What should you conclude about the fact that the engineering professor concocted the plan and the finance professor came along for the ride? Colak's research interests seem to be in IPO markets, not short selling; that would be too cute.

  2. Brokerages are supposed to be able to stop this sort of thing, but I guess a sufficient flurry of options -- at a sufficient number of brokerages -- will confuse them. Will confuse anyone; the SEC is pretty excited that it cracked this case, boasting that "SEC investigators pieced together the complex trading strategy - which involved literally thousands of trades - by tracing one of the trading sequences from start to finish."

  3. The numbers are all over the place. Herbalife seems to be only 1.25 percent, but really small really hard-to-borrow stocks can cost 20, 40, or more percent a year.

  4. And that's 20 percent of the price of the stock. But remember you've both bought and sold the stock. Presumably your short can fund your long so you're not even putting up (much) cash to buy the stock, so it's more than 20 percent of your capital. Maybe it's infinity percent of your capital, whatever.

  5. Mostly. Incidentally, since this is a bit of a throwaway footnote, here is a good visual joke about naked short selling.

  6. There might be other problems if this were the rule? It might be problematic if people were short, like, multiples of a company's stock outstanding. And obviously there are problems with voting -- if you're naked short, and I buy shares from you, how do I know if I have "real" shares that can vote or "naked" fake shares that don't actually exist. But some of these problems exist in some form even in the current system.

  7. Or convertible arbitrage and so forth.

To contact the author on this story:
Matthew S Levine at mlevine51@bloomberg.net

Before it's here, it's on the Bloomberg Terminal.