Do ETF Trades Put the Market at Risk?

Value Added

Do ETF Trades Put the Market at Risk?

The proliferation of exchange-traded products could result in more crashes, such as the May 6, 2010, episode, unless regulators take action.

Exchange-traded funds and similar products have grown like kudzu in recent years. If, however, two top financial experts are right, excessive trading in ETFs poses a huge danger to the stock market.

The U.S. now has more than 1,000 ETFs and other exchange-traded products. Their assets total some $1.2 trillion. Harold Bradley, former longtime head of trading at American Century funds and co-author of a study published by the Kauffman Foundation, says that about ten new ETFs launch in an average week.

ETFs that invest in large, liquid securities (those that can be easily traded without disturbing their price) pose little risk to markets. It’s the ETFs that invest in stocks of small companies and other illiquid securities that are the potential problem.

Sponsored Content

In their 88-page report, Bradley and Robert Litan, an economist at the Brookings Institution who has served in a variety of federal agencies and White House posts, say that trading of ETFs was the main cause of the May 6 “flash crash,” during which the Dow Jones industrial average plunged more than 600 points, or about 6.5%, in less than five minutes.


A report by the Securities & Exchange Commission blamed the flash crash largely on a sell order for stock-index futures entered by a single mutual fund trader. But Bradley says that big investors have made that kind of trade for decades, and it couldn’t have triggered the flash crash. Instead, he says, trading of ETFs that invest in small companies, such as iShares Russell 2000 index (IWM), triggered the collapse in share prices.

The problem isn’t the ETFs themselves. The problem is with the Wall Street brokerages, hedge funds and computerized high-frequency traders who use arbitrage strategies to profit from tiny discrepancies in prices among ETFs, stock futures indexes that hold the same securities, and the individual securities themselves.

In a typical arbitrage trade, a trader will buy a stock-futures index and simultaneously sell short an ETF that invests in the exact same securities but is priced a tiny bit higher than the futures index. In a short sale, a trader bets that the ETF will decline in price. Because the futures index and the ETF own the same securities, their prices will almost certainly come back into equilibrium -- and the trader exits the trade with a small profit.

These kinds of arbitrage traders have been around for decades, but ETFs give them a new tool. The trouble is that trading in ETFs that invest in illiquid securities has mushroomed out of control. On May 6, for instance, the volume in the iShares Russell 2000 ETF was 56% of the volume in the underlying securities. As a result, trading in the ETF had an enormous impact on the stock prices. As Bradley puts it, “The tail is wagging the dog.”


In a statement, BlackRock, which owns iShares, said that its Russell 2000 ETF accounts for just 1.1% of the market value (stock price times number of shares outstanding) of the underlying stocks in the index. That “is clearly not the impactful ratio the authors would like people to believe,” BlackRock said. But it’s not the size of the ETFs that worries Bradley and Litan. Rather, it’s the heavy trading in them.

What’s more, a lot of traders are shorting the ETFs -- that is, betting that they will fall in price. As of September 30, the most recent date for which data was available, $17 billion worth of the iShares Russell 2000 ETF had been sold short. In a short sale, the seller borrows the shares from someone who owns them, then sells them, hoping to buy them back at a lower price. “Naked shorting,” in which the short seller doesn’t first borrow the shares he or she sells, is illegal.

But the Bradley-Litan report says that buying the underlying stocks -- an essential step in creating ETF shares -- would have been virtually impossible. Assuming a buyer bought 10% of the trading volume of each stock daily, so as to not drive the price of each stock up too sharply, it would have taken 42 days to buy the average stock in the ETF. To buy the stock of the tiniest company would have taken six months, the report found. The obvious conclusion: Naked shorting of ETFs is rampant.

Another way of looking at it: The iShares Russell 2000 ETF in mid October was the largest institutional holder of 99 of the stocks in the ETF, and among the ten top institutional owners of another 1,638 of the stocks. “My big worry is that when someone starts to sell the ETF, who will be buying?” says Bradley.


In a calm market, that’s not a concern. But in a market that’s highly stressed, problems can develop quickly.

The market was under considerable pressure on May 6, mainly due to bad news about Greece and its financial problems. Bradley and Litan say that the selloff started in the Russell 2000 ETF. Arbitrageurs, they say, then bought the Russell 2000 ETF and shorted Standard & Poor’s 500-stock index ETFs, betting correctly that the two indexes would converge. But that drove down the shares of the larger companies in the S&P 500.

Bradley is blunt: “More flash crashes are inevitable.” It was, Bradley says, arbitrage trading of stocks and stock-index futures that caused the 1987 crash, driving down the Dow Jones industrial average 22% in a single day. “These are exactly the same issues as the 1987 crash, except done at hyper speed,” Bradley says. “The flash crash was the first electronic market crash,” the report adds.

Bradley also compares ETF trading with the trading of subprime mortgage securities that contributed to the financial crash in 2008. “Just because mortgage-backed securities were packaged in a way that made them distributable around the world, it didn’t make each mortgage more liquid,” Bradley says. “They traded like water, but when you try to break them up, they were more like sand.”


In the report, Bradley and Litan suggest that some hedge funds are even going to ETF firms and asking them to create ETFs that trade in specific securities. This is precisely the same thing that happened in the mortgage market.

The authors also argue that ETFs are responsible for the growing correlation in the performance of different kinds of stocks. Together, ETFs and futures, they say, now trade far more frequently than the stock market itself. Thus, the components move together.

The solution? Like much of the Bradley-Litan report, it’s complicated. But the authors’ primary recommendation is that the Securities and Exchange Commission impose strict limits on the short selling of ETFs. I can only hope that the SEC is listening.

Steven T. Goldberg (bio) is an investment adviser in the Washington, D.C. area..