Why Short Selling Matters

Arnold Kling (hat tip: Matthew Yglesias) argues that short-selling can’t “destroy a good company” because if the company is good, someone will always be willing to pay a good price for it and will be willing to buy it from the short-seller at that good price. But his analysis fails to account for uncertainty and imperfect information. We don’t really know which companies are “good.” Everyone has their own guess about how much a company is worth, but how much is it really worth? We won’t know until we see its results. Actually, we won’t even know then, because it will have more results in the future that we still won’t know about.

In general, since the market value of a company depends on an uncertain guess as to its true value, a good company can be destroyed simply by bad opinions, if people wrongly judge that it is a bad company. And, as I will explain, short-selling can exacerbate this problem in the case of a company that is in fact good but that, in terms of market opinions, is on the margin between being considered good and being considered bad. The “margin” idea applies when there is only one “marginal” short seller, but if there are many, they will effectively widen the margin, so that a clearly “good” company can turn into a clearly “bad” one.

For any given company, the market does not have one opinion, shared by everyone, about the value of its stock. Rather, everyone has a different opinion. People use different methods (including subjective ones, which differ from any one individual to another) to evaluate a stock. Those different methods lead to different conclusions about the stock’s value. One person may think a stock is worth $10; another may think it’s worth $9.99; another may think it’s worth $9.98; and so on down, and maybe, at the other extreme, there is someone who thinks it’s only worth $2.

So suppose the stock is trading at $10, and a large short-seller appears on the scene. The short-seller will satisfy the demand of the person who thinks the stock is worth $10, and then there will be no more demand at $10, and the stock will trade down to $9.99. Then the short-seller (who is, by my construction, “large”) will satisfy the demand of the person who thinks the stock is worth $9.99. Then the stock will trade down to $9.98, and the short-seller will satisfy the demand of the person who thinks it’s worth $9.98, and it will trade down to $9.97. And so on. If there are a lot of large short-sellers, they can bring the price down by quite a lot by, for example, satisfying the demand of everyone who thinks the stock is worth more than $7.

So let’s take that $10/$7 example, and let’s suppose that the definition of “good” requires the stock to be worth $8 or more (in a sense that will become more precise in later paragraphs). By sending the price down from $10 to $7, the short-sellers have effectively shifted the company’s “market reputation” from the “good” category into the “bad” category. No individual has actually changed their opinion, but because of the aggressiveness of the short-sellers (who obviously believe the company is quite a “bad” one), the “market” has changed its “collective opinion.”

Now let’s suppose that the company is (in actual fact rather than opinion) a “good” company but that it needs to raise more equity capital in order to stay good. (When you’re talking about a bank that is solvent but undercapitalized, this might be a reasonable description.) When it offers new shares, the company faces the same demand curve as the short-seller did, and since the price is now down to $7, the company will have to offer the shares at a lower price, say $6, for the offering to be fully subscribed, because it will satisfy the demand of everyone who believes the company is worth $7, $6.99, $6.98, and so on, down to $6.

OK, but suppose that the amount of capital raised at a $6 share price will not be enough to keep the company good. And let’s also suppose that it would have been enough if the offering price had been $9, which it would have been (approximately) if the short-sellers had not become involved. So the company still doesn’t have enough capital, and it needs to raise more. So it offers more shares. But again, the company faces a downward-sloping demand curve for its shares. Suppose it offers additional shares, which satisfy the demand of everyone who thinks they are worth more than $5, so the company offers these new shares at $5. But suppose that still is not enough to keep the company good. It has to offer more shares. But now that it has done two separate offerings, and it attempts a third, it probably won’t have the confidence of the market. Market participants will say, “If we pay $4 now, who is to say that the company won’t come back and offer shares at $3? We’ll wait for that.” And if they wait for $3, why not wait for $2? And if $2, why not $1? Of course at some point everyone is going to realize that the company is going to be unable to raise sufficient capital at any price.

In practice, potential buyers will have realized that much earlier. If they have a reasonable guess as to how much capital the company needs, and a reasonable guess as to what the demand curve for its stock looks like, then they will be able to come up with a reasonable guess as to whether the company can raise the necessary capital. Of course, there will be a variety of guesses, and opinions will differ. But at some point (as the declining stock value makes the problem clearer), only a few people (or none at all) will be of the opinion that the company can raise enough capital. At that point, the company is effectively ruined, the price goes down to near zero, and the short-sellers profit handsomely. If there are enough large short-sellers, not only can they destroy a good company, they can make a lot of money doing it.


UPDATE: Another way to argue that short-selling doesn't matter would be to argue that the person with the $10 opinion has an infinite amount of capital available, and therefore they will demand an unlimited number of shares at $10. Of course, it's rather silly to think that anyone has infinite capital. And for someone with limited capital, the more they pay for the shares, the more risk they are taking. So they will be willing to buy a certain number of shares at $10, but after that, the risk become too high, and the price has to go down to get them to buy more. Essentially, my argument above still applies, except that you can construct the demand curve from just one person's opinion. If there are many potential buyers, each of whom has limited capital and limited risk tolerance, but who have different opinions, the demand curve will still slope downward, and again my argument applies.