In partnership with

Day to day, what drives asset prices is some combination of changes in fundamentals and changes in investor sentiment. Sometimes a stock is up because the company won a big contract, and sometimes it's up because investors have the general sense that it's going to win one some time.

And sometimes, the logic of the investment flips around, and it's almost entirely driven by investor sentiment. We've written about this before, in the context of George Soros' theory of reflexivity: he noticed that when mortgage REITs' shares went up, their borrowing costs declined, so they could earn a bigger spread on their mortgage portfolios, which made the stocks go up even more. This kept going for a while until they ran into the problem that the new loans they were making weren't as good as the old ones, and more abundant financing meant that lenders were competing to get borrowers rather than the other way around, so the trade unwound.

MicroStrategy has what is in some ways a purer but in other ways a much weirder feedback loop going. (I'm short a small amount of it. Fun ride so far.) It goes like this:

• MicroStrategy used to have an operating business and a pile of cash. The business wasn't growing, but it made a bit of money. But they got worried about their cash pile, and moved $250m out of their $421m in cash reserves into Bitcoin, at a price of about $11.6k. Bitcoin prices ripped, and MicroStrategy kept buying. As a reasonably liquid stock listed in a major exchange, they were, in late 2020, one of the better ways for an investor to get exposure to crypto without doing anything weirder than they'd do if they bought shares of Exxon to get exposure to oil.

• MicroStrategy noticed that investors found this appealing, so they started issuing stock and convertible bonds and using this to buy more Bitcoin.

• They trade at a premium to their Bitcoin holdings, which are now by far their most important asset. So this secondary issuance is accretive to their value, i.e. every time they issue more stock and use it to buy Bitcoin, their Bitcoin per share goes up.

• They actually report a metric based on this, called "Bitcoin yield." It's not a yield, just their change in Bitcoin per share, in other words it's just a measure of how overpriced their stock is and how aggressively they exploit this.

So, what's a prospective short-seller to do? You have the classic conundrum that the stock trades at ~2.3x the value of its underlying assets, but if it trades at that premium, what's the argument against it trading at 3x (as it did last week) or 4x, or 10x?

This is a hard question. But the feedback loop works in the opposite direction, because it has a few dependencies: First, one of the reasons MicroStrategy issues convertible bonds specifically is that they monetize not just the price of the stock but also its volatility. A convertible bond is a bond paired with an option to buy the stock, so its value is partly a function of how much that stock bounces around. These bonds are often bought by convertible arbitrageurs, who are buying the bond and hedging their exposure to the stock price, the same way an options dealer would. And given high retail interest in the company, one way they can straightforwardly hedge is to write call options. That's an especially good deal if the convertible bond is being priced by bankers who know what the embedded option is worth, while the options are being priced by retail investors who see them as a source of leverage. (As a general rule, retail investors look at options as a way to get more exposure for a given dollar investment. Institutional investors look at options as a way to do some combination of betting on volatility separately from price direction and betting on discrete, time-sensitive catalysts.)

Meanwhile, MicroStrategy's trades have a price impact; they're buying Bitcoin, and traders are also buying it in anticipation of these trades (the net impact of both is that the price impact of their trades gets smoothed out, but it doesn't completely go away). So they're participating in a feedback loop where investor enthusiasm about MicroStrategy leads to inflows into crypto, which validates that enthusiasm (though these investors could just as easily express this view by buying Bitcoin directly, which is now much easier than it used to be—you can think of a share of a Bitcoin ETF as a business that owns Microstrategy stock and trades at 44% of fair value).

So, like a lot of other modern retail trading phenomena, the way it unwinds is not necessarily through a crash, but through gradually diminished interest. If the premium is lower and traders aren't racing to buy call options, those convertible and equity offerings are less exciting, and the company isn't putting so much upward pressure on the price of its main asset.

In the short term, owning MicroStrategy has been a fantastic trade; a recent drawdown notwithstanding, shares are up 416% this year. But it's a feedback loop that works in both directions. Structuring a bet against it is hard because there's a good reason to question every element of the bet: you can short outright, but you'll spend time gradually covering, and it takes conviction to double down repeatedly once the trade is working. You could write call options, but you'd be run over by the price appreciation, and hedging those call options by buying the underlying is also expensive given how volatile the stock is. (Options dealers obviously convert this execution cost into their options prices; there's an option price at which it's worthwhile to either pay the cost of constantly adjusting the size of the hedge or just pay the volatility cost of not keeping that hedge fully aligned.) You might say that if there's a feedback loop, you should own deeply out-of-the-money puts, but you're also paying plenty for those because the volatility is so high.

So you're stuck. MicroStrategy eventually converges on Bitcoin, whether in the base case where the premium drops because of a lack of investor enthusiasm or even in the bull case where it owns all the Bitcoin (in that scenario, its last buy is from someone hedging a short position in MSTR itself). But along the way, there's no particular reason to expect the gap to close at any specific time. That makes it an unusually pure version of a valuation short, which is a notoriously tough category. At the same time, there are plenty of notorious cases where a company did well fundamentally but investors broke even over long periods because their entry point was so high: if you bought Oracle at just the right moment in the year 2000, and waited a decade and a half, you owned a company that quadrupled its revenue and net income—and had net appreciation of precisely 0%. Which is a good way for short sellers to stay patient: over long periods, betting against a company that's going to go down when the market overall goes down, but that's going to struggle to go up because it's so expensive, is a lumpy, uneven proposition, but sized appropriately it's hard to go broke selling dollar bills for 230 cents.

Read More in The Diff

In The Diff, we’ve talked a few times about trends that keep going for a surprisingly long time:

• The Memo and the Multiples ($) considers what happens to tech revenue when tech companies spend less on each other.

• And tech had its own special macroeconomy for a while.

Sometimes, inefficiencies persist because it’s too hard to trade on them ($).

And note that, for a limited time, new Diff paid subscribers who choose an annual plan can get a free copy of Boom.

Ease into investing

Ease being the key word. With automated tool like portfolio rebalancing and dividend reinvestment, Betterment makes investing easy for you, and a total grind for your money.

Get started

Join the discussion!