You cannot sell a share you do not own until somebody lends it to you. A short seller borrows the stock, sells it straight away, and waits for the price to fall. If it does, they buy the share back for less, return it to the lender and keep the gap.
The borrowing is where the trade gets its texture. There is no central exchange for lending shares, so finding stock to borrow can be hard and sometimes impossible, according to a review of the research in the NBER Reporter. The seller has to find an owner willing to lend.
Who lends the shares, and for what
The lender is paid. The short sale raises cash, and the lender earns interest on it, minus a rebate handed back to the borrower. That rebate works as the price of the loan. When a stock is in heavy demand it can turn negative, and the short seller then pays the lender every day for the right to keep the borrowed share.
The NBER piece puts a number on why that matters. Take a stock worth $100 and a $1 cost to short it. Nobody can profitably stop the price drifting to $101. Make that $1 a daily holding cost and the short becomes a bet that the stock falls at least $1 a day.
Big, widely held companies are usually cheap and easy to borrow. Small companies, those with few institutional owners and those everybody wants to short are not.
The lender can ask for the share back
A borrowed share stays on call. The lender can recall it at any time, and a short seller who cannot find a new lender must close the position, whatever the price.
That recall risk makes some lenders better than others. Economists at the Federal Reserve Board, Bastian von Beschwitz, Pekka Honkanen and Daniel Schmidt, studied stocks either side of the Russell 1000/2000 index cutoff. They found that more ownership by passive index funds raised both the supply of lendable shares and the amount sold short, while lending fees barely moved. Their explanation is that passive funds are less likely to recall a loan, which makes their shares better to borrow.
How much can go wrong
Dan Moczulski, UK managing director of the trading platform eToro, described institutional shorting to the BBC as carrying unlimited risk. On eToro, retail customers short through a contract for difference, start closing once losses pass 50% of the account, and can never owe more than they put in.
The professionals lose often too. Carson Block of the activist short seller Muddy Waters told the BBC that of his first 100 shorts he reckoned he won about 30% and lost 70%. Peter Roscoe, a trader with 18 years in the markets, tells followers who want to try it that without four to five hours a day of watching prices they have no chance.
Why markets let people do it
Defenders argue that short sellers help find what an asset is worth and sometimes expose fraud. Block calls activist short sellers providers of transparency.
The Adam Smith Institute made the same case during the GameStop episode, when short sellers lost an estimated $20 billion and calls for a ban followed. Its argument, which leans on the economist Robert Shiller, is that a price only reflects every view if people who think it will fall can act on that belief.
The NBER review adds evidence from the other direction. When shorting is costly or restricted, stocks can stay overpriced and go on to deliver low returns until the mispricing corrects. The Federal Reserve paper found that extra short selling linked to passive ownership made prices respond better to bad earnings news.
Not everyone is persuaded. Short sellers face waves of hostility from governments and the public, usually after a crash, when they are the ones blamed.
