A short in action.

Photographer: Jacob Kepler/Bloomberg.

It's Hard to Bet Against Houses

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

Here is a thing no one thinks:

The shrimp market is far less rational than even the often irrational stock market, for a couple of important reasons. First, most investors find it difficult to understand how shrimp supply responds to changes in demand. Only a small minority of people think carefully about such things. Second, it is very hard for the minority of smart-money investors who do understand such matters to bet against bubble-level prices in shrimp markets. In shrimp, the smart money has relatively little voice. Efficient markets require the possibility of selling short. Short-selling helps prevent bubbles from forming, but such negative bets cannot easily occur in the shrimp market. You can't routinely borrow a shrimp and sell it, promising to buy back the same shrimp later to repay the loan.

No one thinks that in part because it is so weird, and in part because you actually can short shrimp, I think.  But here is a Robert Shiller piece in the New York Times along the same lines,  except instead of shrimp he's talking about the housing market, and arguing that housing is particularly susceptible to bubbles because, unlike in the stock market, it is hard to short houses. But it is hard to short shrimp, too, and you hear rather less about shrimp bubbles. 

Part of the difference is that shrimp are a smaller portion of the economy than are houses. A more fundamental difference is that housing is part consumption (you get a place to live) and part financial investment/speculation (you might be able to sell your house at a profit), while buying shrimp is pure consumption (you get to eat shrimp) and not at all investment (please don't resell the shrimp). Bubbles can only really happen in investment assets; a bubble, by definition, requires the expectation of reselling at a profit. So a shrimp bubble seems unlikely.

Still, it might be fun to think for a minute about shrimp, or cars, or burgers, or iPhones, or any other thing that has a price but that is hard to short.  Not in any sort of rigorous way -- this is all super loose! -- and not as a rebuttal to Shiller, who is probably right that the difficulty of short selling increases the risk of housing bubbles. Just because short selling is a sort of misunderstood thing, and an interesting one.

So: What happens if the price of a good -- just a regular good -- gets too high? Well, you know, supply, demand, etc. If the price of a good is too high for you, you can do one of two things:

  1. Not buy it, or
  2. Make more of it and sell it.

For most people, not buying is the standard move. And this actually works great in housing markets. Housing is a big percentage of my net worth, as it is for most people who own housing. Choosing not to own would be a huge shift in my portfolio, a much bigger bet against housing than I could afford to make against any stock by shorting it. 

More generally, the way to bet against an asset's price is to have less than X of that asset in your portfolio, where X is, like, the "regular," neutral amount of that asset to own. Arguably -- traditionally -- the neutral amount of a stock to have in your portfolio is zero shares, so the way to bet against a stock is to short it (sell more shares than you own, so you have negative shares). But zero is pretty arbitrary. You have to own something; if you have totally neutral views on the prices of all assets, you will still own some assets. Plausibly, the neutral amount of stock to have is its index weighting.  On that theory, owning less of a stock than the index does is a bet against it, even if you actually own some shares. So some sell-side research uses the term "underweight" instead of "sell": If you don't like a stock, you should own less than X of it, but they don't assume that X is zero.

Similarly, it is not obvious to me that the neutral amount of housing to own is zero housing. If the neutral amount is more than zero, choosing to rent is a bet against house prices: You own less than the regular amount of housing, so you are effectively short.  Of course, it is not a bet that can be scaled up to hedge-fund size, which is important: If most people overestimate house price appreciation (risking a bubble), the minority betting the other way needs scale to influence prices. It is hard to scale "not buying a house." I mean, you can not buy two houses, but that doesn't give you twice the exposure to not owning houses.

But for most products, the standard economic story is less about not buying and more about selling. It's a story of competition: When prices are high, someone comes in and creates more of the product. If iPhones are expensive, other companies will make competing phones, driving down the price. If shrimp are expensive, people will farm or catch more shrimp.  Prices equilibrate supply and demand; if prices are high, supply will increase. 

This should work in the stock market, too. If stock prices are high, then someone should make more stock and sell it. But there are behavioral inefficiencies that prevent that. The standard story of the U.S. stock market right now is :

  1. Stocks seem expensive, and
  2. Companies keep buying their own stocks.

That's weird! Companies can make stock. If they were rational, when prices were high, they'd do that, and then sell the stock. But of course they don't. They tend to be run by confident managers who think that the stock is still cheap: If the price will get even higher, selling stock now would be a mistake. Also, they don't need the money: Stock prices tend to be high when companies are making a lot of money, so the companies see no need to issue more stock to raise more money. (In fact, they have so much money, and so little to do with it, that they end up buying back their stock for lack of better ideas.) Also, if they were to sell stock, with no use of proceeds other than "we think this stock is pretty expensive right now and want to cash out," who would buy it?

But you know who else can make stock? Short sellers. Short selling creates new shares of stock, allowing supply to increase in response to high prices. It lets stock markets be competitive: If stock prices are too high, and companies won't make more stock, then you can make stock and sell it. You may think that a company's stock is too expensive, but the company disagrees, so you go manufacture more of it and sell it while the selling is good.

But short selling is not the norm for normal markets; it's a weird way to introduce competition into weird markets. Here are Dan Davies and Tess Read explaining short selling through the example of borrowing overpriced shoes and selling them on eBay, and just reading it should make clear how weird short selling is. No one but a sociopath would actually borrow a friend's shoes and sell them on eBay. If you think shoes are overpriced, you open a shoe factory. Short selling is just a way to open a stock factory.

So perhaps: A market will be rational, in some extremely loose sense,  if it is easy to create supply to respond to demand. In markets for widgets, you create supply by building widgets; in markets for stocks, you create supply by short selling.

Is housing like stocks, or like widgets? Probably some of both. If house prices are high, you can build houses. Here's Shiller:

There is a way for smart money to profit from an understanding of high prices. It is to build new houses and sell them before prices fall. This is a time-consuming process but it is what we are starting to see now, as housing starts and permits data show.

That clearly happened during the housing bubble too. There are ghost towns of abandoned houses built during the boom. Those houses were built to meet demand perceived in the prices, though they didn't prevent a bubble. The reasons for that are ... let's say various and beyond my expertise. But obviously house-building does not frictionlessly respond to perceived demand. It takes time to build a house, there are zoning and land-use restrictions on new construction, it's hard to move houses around the country to respond to local demand, etc. When the bubble burst, we ended up with lots of people without houses, and lots of houses without people, suggesting that, in housing, supply and demand are not fully equilibrated by price.

One other problem in housing is leverage: Lots of people borrow money to pay for housing, meaning that house prices are only partly about housing, and partly about borrowing. And you can sort of short that. Shiller again:

During the financial crisis, some professional investors did manage to profit by correctly forecasting home price declines. They used mortgage derivatives such as collateralized debt obligations to place their bets. John Paulson of Paulson & Company is well known for very successfully profiting from his prediction of trouble in the housing market. But mortgages are not homes, and he and others like him did not beat down the emerging housing bubble before it grew out of proportion.

You can't quite blame them for that; homebuilders didn't beat down the bubble either, and they were selling homes. But it is worth remembering that short selling produces supply, "phantom" supply that competes with the regular supply. So when Paulson shorted mortgages, he created new phantom mortgages. Investors who were buying mortgages (or rather, synthetic CDOs of mortgage-backed securities) bought his phantom mortgages instead of real mortgages. Which was probably a good thing: If Paulson hadn't created the supply, the bank mortgage machine probably would have, meaning that even more dodgy mortgages would have been made, and the resulting crisis would have been worse. Think of that next time you get mad at banks for betting against the bad mortgages they were selling. The alternative was selling more and worse mortgages!

Short sellers are often unfairly vilified; betting on failure just icks a lot of people out. That is silly, but not because short sellers are really bold heroes standing firm against the evil of bubbles. The main point is that short selling is pretty normal, or at least, pretty close to normal. If you don't like something, you should own less of it -- maybe less than zero of it. And if something is overpriced, you might get into the business of selling it -- which sometimes can only be done by shorting. 

  1. A Google search for "shrimp futures" is illuminating. For instance:

    Frozen Shrimp Futures are traded on the Osaka Dojima Commodity Exchange. Each contract is for 108 kilograms of Frozen Shrimp. Prices are quoted in Yen per 1.8kg.

    Though, from the quotes I'm not so sure about shrimp market liquidity. Also, here is "They Trade Shrimp in Minneapolis? An Examination of the MGE White Shrimp Futures Contract." My research into this topic ends here, but I invite you to continue it.

  2. And in the same words, ceteris paribus, though I have omitted some portions in between to focus on the short selling.

  3. You can "short iPhones" by shorting Apple stock, or "short burgers" by shorting Shake Shack stock (or cattle futures), or whatever. Like, you can make bets against those products. But those are not straightforward bets that the prices of those products will decline; also, crucially, those bets do not in themselves reduce the prices of the products the way that a short sale of stock does.

  4. You know, times the size of your stock portfolio, for some relevant index, etc.

  5. The regular amount is, you know, half a house or whatever, the average amount of house that people own. If you own a house, you own more than the average amount, so you're long; if you don't, then you own less than the average amount and are short. This is super loose, but you get the idea.

  6. I should perhaps point out that the shrimp industry "has a horrific dark side," and I sort of regret using it as my lighthearted example, oh well.

  7. See, e.g., the Goldman analysts who argued that "acquisitions -- particularly in the form of stock deals -- represent a more compelling strategic use of cash than buybacks given the current stretched valuation of U.S. equities.

  8. The stylized fact is that public companies buy back stock when prices are high and sell more stock when prices are low. On the other hand, private companies are tempted to sell stock -- that is, go public -- when markets are strong. So that has some effect on increasing supply. But private companies tend to be small, so their public offerings can only do so much to increase supply. And private companies have their own, not-purely-financial, reasons not to want to go public.

  9. No one believes this but it is a straightforward fact. Like:

    • Company X issues 50 shares.
    • A buys 10 shares, B buys 10 shares, C buys 10 shares, D buys 10 shares and E buys 10 shares.
    • There are no other shares. There are 50 shares, owned by A, B, C, D and E.
    • F borrows 10 shares from A and sells them short to E.
    • Now A owns 10 shares, B owns 10 shares, C owns 10 shares, D owns 10 shares and E owns 20 shares.
    • There are 60 shares.
    • It's magic!
    • (F owns negative 10 shares, so there are 50 "net" shares in some sense, but so what? A has loaned out the 10 shares that she still owns, but again so what?)

    Herbalife's disclosed holders on Bloomberg own 126.93 percent of its stock. People get very exercised about the idea that naked short selling creates phantom shares, but clothed short selling creates exactly the same phantom shares. Don't think of them as phantom shares! Think of them as real shares created by short sellers to meet demand that companies are unwilling to meet.

  10. It also has lots of frictions, risks, etc. As Dan Davies points out, it is not the norm for overvalued assets to attract short sellers: You need a clear thesis that there will be a large and near-term correction for short selling to be attractive, not just, like, "Eh I think that price is a little high."

  11. Like, "It will produce prices that aren't crazy." Obviously, Shiller is talking about a more financial-market sort of rationality, in which prices predict future cash flows, and I am being super loose by conflating financial and product markets. There is no obvious analogue between stock and shrimp prices being "rational." Sorry! 

  12. I mean. It is not obvious what the equilibrium was there. Disclosure: I worked at Goldman Sachs, which became rather famous for this.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

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

To contact the editor on this story:
Zara Kessler at zkessler@bloomberg.net