Is There a Social-Network Bubble?

I want to consider what an investment bubble is, why it arises, whether it's irrational, and whether the current valuations of social network enterprises such as LinkedIn and Facebook are a bubble phenomenon.

Many finance theorists regard the price of an asset, such as a corporation, as the discounted value of its predicted profits. Assets thus are overvalued ex ante if the prediction is unrealistically optimistic and are overvalued ex post if, though it may been the most sensible prediction given what was or could be known when made, it turned out to be exaggerated. Because of uncertainty such disappointments are inevitable. But the premise is that investors are driven by predictions of profits.

Finance theorists who think that all trading is guided by profit predictions exclude the possibility of bubbles, in which asset prices rise steeply and then collapse, seemingly without regard to estimations of future profits. An extremely high price-earnings ratio is a symptom of a bubble, since a very rapid and steep increase in future earnings would be necessary to make the asset in question a "good" investment in a conventional sense, and such increases are rare.

Although a bubble thus violates the assumption that asset prices are driven by profit (or loss) expectations, that doesn't make buying in a bubble, even by a speculator who thinks it's a bubble, irrational. As Keynes pointed out, a stock market speculator is not interested in the profitability of the firms whose stocks are traded in the market, as such. He is interested in the behavior of the other traders. If he thinks they will bid up the price of a stock he will see an opportunity for gains by buying the stock even if he thinks it's a dog. In other words, he buys not because he thinks the stock is undervalue but because others are buying it. This is a rational strategy, although risky. It is rational because if prices continue rising the speculator will make money, at least for a time. It is risky because other investors may stop buying, and start selling, and prices will fall. The bubble speculator can try to protect himself by moving in and out of the market rapidly, enabling him to lock in gains before deciding whether to try for a further profit. Increased trading activity is in fact a symptom that a bubble is reaching its peak.

Bubbles flourish in periods of "new era" thinking. The late 1920s were heralded as a new economic era because of the tremendous boom in automobile production and the advent of new methods of consumer credit, such as installment buying. This set off a stock-buying frenzy that set the stage for the stock market collapse that began in October 1929. The rise of dot-com commerce in the 1990s was thought to herald a "new era" in commerce, and again there was a stock bubble and eventual bust. The advent of novel methods of financing home purchases, including adjustable-rate mortgages and mortgage securitization, gave rise to the housing bubble of the 2000s that ended abruptly in the housing crash of 2008.

These bubbles were based on calculation rather than irrational exuberance. People who bought early in the bubble and sold before it burst did well. There are plenty of suckers and fools, but that is a constant, and bubbles are only occasional.

Although rational, bubble trading creates external costs, and so should be discouraged, or at least limited. The externality is macroeconomic. A bubble causes asset-price inflation, which increases borrowing because asset owners have more to offer as collateral for loans. The recent housing bubble caused an enormous increase in consumer debt. When a bubble bursts and asset prices plummet, loan collateral falls in value, resulting in contraction of credit. Debt cannot easily be rolled over, defaults skyrocket, and a downward spiral in buying and selling ensues.

A bubble in a single stock would not have macroeconomic consequences. But a bubble is likely to involve all the stocks in a particular segment of an industry (or even the entire industry), because whatever is pushing up the price of one stock is likely to push up the stocks of companies that sell the same or similar goods or services. The belief in a "new era" is unlikely to affect only one company.

The simplest way to try to control bubbles is to limit borrowing for stock purchases (that is, buying stock on margin). But that is unlikely to have much effect on a bubble limited to one industry, such as online social networking, unless the industry is huge, like housing (the housing market is larger than the entire stock market). The dot-com bubble of the 1990s, when it burst, caused a recession that would have been of little consequence had it not been for the Federal Reserve's incompetent response (pushing interest rates way down and by doing so setting the stage for the housing bubble, because houses are bought largely with debt and therefore housing prices can soar when interest rates are very low). Similarly, if the social-network stock-buying frenzy is a bubble, its bursting is unlikely to have a substantial effect on the economy as a whole. The online social network industry is tiny—the entire industry employs only a few thousand people—and the collapse of its stock values, which is no sure thing—though the astronomical price-earnings ratio of LinkedIn is a bubble symptom—would not be an economic disaster, even in the current fragile status of the U.S. and world economy.