Short-sales have occupied the attention of regulators for a considerable length of time. Bans on short-selling were an immediate response to the crisis and to the precipitate declines in equity prices occurring in 2008 and 2009. In May, German financial regulators imposed a ban on ?naked? short-sales of Eurozone bonds and credit default swaps as well as on selected German financial stocks. In other countries, these bans were imposed on both the naked (selling securities that you have not been able to borrow or guarantee that you can borrow) and covered (borrowing and then selling) incarnations of short-selling.
The imposition of these bans is generally accompanied by a significant amount of rhetoric. For example, earlier this month, Nicolas Sarkozy and Angela Merkel wrote a letter to the European Commission President, Jos? Manuel Barroso, stating: ?Naked short-selling should be prohibited to refrain (sic) European markets from suffering a new wave of severe turbulence.? There are good reasons to believe that this rhetoric may be overblown.
Financial economists generally disagree about most things and short-selling is no exception. The theoretical models that have been written about short-selling offer different opinions about whether stock prices will be overpriced or underpriced if short-selling is banned. There is also little agreement about the effects of short-sales during market crashes?do they push prices to depths that cannot be justified by underlying economic realities? These disagreements make the effect of bans on short-sales an interesting empirical question?and markets have recently provided empiricists with an excellent experimental laboratory, now that sufficient time has elapsed since the bans were imposed.
In the US, researchers analysed the effects of the 2008 short-sales ban on over 1,000 stocks. They found that the liquidity of the stocks subject to the ban declined, but that the prices of these stocks increased sharply. These declines in liquidity make intuitive sense?if you eliminate one direction of trading, liquidity should decline. However, the price rise that they detected is difficult to interpret. This is because there are several competing interpretations of this result. Prices could have increased because banning short-selling was good for firms, i.e., short-sellers were hammering prices down below the level of fundamentals; or perhaps prices were artificially inflated because banning short-selling killed off negative but sound opinions about the price. A third possibility is that prices increased simply because other events with no relation to the ban on short-selling simultaneously boosted prices. An important confounding event in the US study was the simultaneous announcement of the Troubled Asset Relief Programme (TARP).
Another clever study uses the international imposition of the short-selling bans to kill off the third (TARP) possibility. The authors use a huge set of firms from over 30 countries, and show pretty conclusively that the imposition of both covered and naked short-selling bans reduce liquidity. In a US-free sample of countries, they find that bans actually decreased stock prices on average. The conclusion that we can draw from these analyses is that the imposition of short-sales bans, far from stabilising stock prices, actually have deleterious impacts on market liquidity.
There is another pernicious effect of short-sales bans that is not covered by these studies. I?m going to call it the ?round up the usual suspects? problem. The issue is that a partial imposition of a short-selling ban affects companies not covered by the ban. For example, the FSA in the UK put 34 financial sector firms on a list of companies in which short-selling was restricted. Consequently, companies in the financial sector that were left off the list suffered massive stock price decreases, as they became immediate short-sales targets. Many of these firms then went to the FSA, pleading to be put on the banned list. I would wager that counting this negative externality imposed on firms excluded from short-sales-ban-lists would significantly increase the measured costs of short-sales bans.
Is the right solution then to ban short-sales in everything if you are going to ban short-sales at all? If you ban short-sales in all stocks in a small open economy, you are likely to export the problem to your neighbours (or countries with stocks that are in similar industries). For example, a ban on short-selling in Cambodia could result in stock price declines in Vietnam. More generally, I think the ?round up the usual suspects? problem neatly illustrates the perils of interfering in the market too much. Any committee that decides the market is unfairly penalising certain stocks and decides to put them on a restricted-short-sales list is likely to have far less information than the market in its wisdom. Merton?s law of unintended consequences has serious bite here. (I realise that the market?s wisdom has taken a beating recently, but we?re still to come up with a better mechanism for information aggregation. Committees are clearly not it.) In the absence of any convincing evidence on the merits of short-sales bans, let?s agree to ban any bans on short-sales.
The author is a financial economist at Sa?d Business School, University of Oxford