“He who sells what isn’t his’n, must buy it back or go to pris’n.” That’s a quote from the legendary 19th century industrialist, some would say robber baron, Daniel Drew. These days, the penalties of short-selling don’t include prison time, but anytime a country or company have financial difficulties and their debt or stock begins to decline, they usually push for some type of short-selling ban. In fact, a short-selling ban was just in the news again last week when the UK’s FCA (Financial Conduct Authority) lifted its short-selling ban on Banco Espirito Santo shares. They weren’t the only one, Portugal’s financial markets regulatory authority banned short-selling immediately following the bank’s crisis.
Time and time again, those short-selling bans do more harm than good. The regulation is counter-productive and it results in less liquidity, greater volatility, and wider bid-offer spreads. But, what’s usually left out of the debate is the difference between naked-short-selling and legitimate short-selling. I support tight regulation for naked-short-selling, but legitimate short-selling is an important part of an efficient market.
The first known law banning short-selling appeared in Holland on February 24, 1610, when the government banned “trading the wind.” That meant selling shares that someone did not own. The law was enacted when the East India Company complained that short-sellers were driving down the price of their stock. Sound familiar? Years later, the British Parliament banned the sale of stocks that were not owned, blaming short-sales for causing banking panics. The British markets and courts ignored the ban until Parliament repealed it in 1860, only to be replaced 7 years later by a ban on short-selling bank shares only.
In the US, the newly created SEC enacted short selling regulation in 1938 to address one of the perceived causes of the stock market crash of 1929. Those rules sat virtually unchanged until 2004 when the SEC implemented the “uptick rule.” More recently, the SEC passed Regulation SHO which includes “locate” provisions and “close out” rules for fails.
In 1963, the SEC conducted a study which concluded that short-selling rules did not prevent any harmful effects of short-selling, the same effects the rules were targeting. Another SEC study (Release No. 34-42037) stated that short-selling increases pricing efficiency, concluding that “efficient markets require that prices fully reflect all buy and sell interest.” One private study by Jones and Larsen concluded that markets with short-selling bans or restrictions have higher than normal prices but are subject to sharper downturns. That makes a lot of sense to me.
More recently, in July 2008, just before the financial crisis, the SEC banned naked-short-selling for the shares of 17 American investment banks and banks, including Fannie Mae and Freddie Mac. We all know what happen to bank shares just a few months later. A 2008 Credit Suisse study found that the short-selling bans made stock prices less efficient and buying and selling a stock more expensive for investors. They found that bid-offer spreads doubled, on average, during bans and after bans were lifted those spreads declined by 65%. Then, an Oliver Wyman study in 2010 found that bid-offer spreads in the UK widened by 45% for stocks with additional short-selling disclosure rules, whereas during the same period, stock spreads without the rule only widened by 2%. For the overall market, wider bid-offer spreads increases the transaction costs for market participants.
The Real Problem And Solution
The studies are pretty clear: limiting short-selling does not promote an efficient market. Banning short-selling often miss-identifies the real problem – naked-short-selling.
Let’s take a normal short-selling transaction where a market-maker short-sells a security. It could be a stock or a bond and it could be to fill a customer order, to hedge another position, or even just to get short. In general, traders know which securities are “hard to borrow,” so let’s assume there’s no problem borrowing the securities. In this case, the trader is short and the securities are borrowed from an actual owner. The actual owner doesn’t wish to sell their securities at the time, but makes them available for someone else to sell, earning a little extra income in the process. This transaction is the normal part of an efficient market and it doesn’t distort the market in anyway. In economic terms, there was no change in the supply of the security.
Now, let’s take naked-short-selling with the same transaction above. This time, the trader sells shares they don’t own and cannot borrow. This selling can easily drive down the price of a stock artificially because it effectively adds supply to the market. That’s right, naked-short-selling creates additional supply. Since the trader never borrowed the securities from an actual owner they have no securities to deliver and the trade fails. Buyers believe they would receive the securities they bought but never receive them. This is not a normal part of an orderly market. The short-seller who cannot borrow shares is creating new supply in order to drive down the price of the stock. More supply with the same about of demand results in lower prices. The best way to limit or eliminate naked-short-selling is to make fails punitively expensive. We did it, to some extent, in the repo market back in 2009. It’s the best way to force market participants to cover their short or buy back the securities.
Overall, there’s still a global perception that all short-selling is bad, when in fact, only naked-short-selling is. Every time there’s some type of market panic, the first response is to ban short-selling, which actually does more harm than good.