ETF Trade Settlement Risk Raises Its Head Again

Failed ETF trades saw a spike this summer. Experts differ on how big a risk that poses.


James Angel isn’t shy when he sees something wrong in the markets. “I’m the kind of guy, generally, who complains about serious risk,” the Georgetown University McDonough School of Business finance professor tells Institutional Investor.

“I warned the SEC five times in writing about high frequency trading before the [May 2010] flash crash. When I see a big risk, I scream about it,” he says, referring to the many warning letters he sent about the risk of rapid, computer-driven trading to market stability. So it’s unusual to find Angel on the side of the everything’s okay crowd, but that’s precisely where he is in the debate over so-called failed trades in the exchange-traded fund market.

“Of all the things to worry about in equity market structure,” Angel says, “I would put [failed ETF trades] not at the bottom but well down on the list.”

The issue has raised its head again with the news recently that failed trades — that is, trades that have taken more than the industry standard three days to settle — have spiked dramatically in June. According to data from Boston, Massachusetts–based risk consulting firm Basis Point Group, failed trades for the 30 biggest ETFs hit a daily high of $3.96 billion on June 26.

The surge in failed ETF trades has reignited a debate that was last aired in summer 2011, when a similar spike raised questions about ETFs’ place in the market. Unresolved then was the question — asked in both the Economist and the Financial Times that summer — as to whether those failed trades are a simple by-product of the nature of ETFs or whether they they indicate something bigger and more worrisome.

Fred Sommers, co-founder of Basis Point Group, believes failed trades are a major systemic risk costing investors billions of dollars in lost liquidity. “High frequency trading occurs in nanoseconds,” he says, “So if it takes you three days to find a security in a nanosecond trading environment, what kind of liquidity is there in the market?”


Sommers was one of the authors of a report by the Kansas City, Missouri–based Ewing Marion Kauffman Foundation in 2011 that described failed trades as “canaries in the coalmine,” suggesting they could trigger a broader chain reaction in the markets. As they wrote at the time, “ETF fails account for approximately 60 percent of the nearly $2 billion of daily equity trading fails reported to the [Securities and Exchange Commission], and on some days they account for 90 percent of all exchange-traded fails.”

Sommers and his fellow authors — Basis Point colleague Robert Fawls and the Kauffman Foundation’s CIO Harold Bradley and vice president for research Robert Litan — went on to say that “every fail introduces a cumulative and potentially compounding liquidity risk into the orderly process of settling the $7.5 trillion of security transactions completed each day, which could be especially dangerous during times when financial institutions are short of liquidity.” (More recent reports from the Bank for International Settlements, the International Monetary Fund, the Bank of England and the Financial Stability Board all fingered ETFs as potentially posing systemic risks to the financial system, but did not mention failed trades as one of those risks.)

Angel, however, disagrees. He points out that the failed trades aren’t really failed at all, merely slower to resolve than the regulatory three days allows for. Market makers in fact have T+6 — that is, six days after the trade is executed — to complete ETF trades. And the vast majority of the failed trades, he points out, are completed in that time frame.

That’s not to say that the issue is problem free, as Angel wrote in a letter to the SEC back in 2008. “When a seller fails to deliver a security on the normal settlement date, an owner of the security is effectively forced into making an involuntary stock loan to the seller,” he penned at the time. “This violates one of the basic property rights of an owner, the ability to exclude others from the use of the property. In addition, it may deprive the buyer of the ability to earn stock lending revenue, as well as voting rights. The buyer may even have to pay higher tax rates on substitute dividend payments, which are taxed differently than normal dividends.”

Angel blames “the clunkiness” of ETFs’ creation-redemption system, with multiple participants involved in the creation and redemption of baskets of ETF shares (known as creation units), which is far more complex than investing in a mutual fund. “If it really were as simple as tap-tap-tap on a keyboard, we’d be happy ever after,” he explains. “But because of the costs in creating and redeeming shares, that makes it a little harder than plain-vanilla equities.”

In June, the problem got a little simpler, at least in Europe, when asset management giant BlackRock and clearinghouse Euroclear Bank announced plans to issue an ETF that can be traded across Europe but settled centrally by Euroclear, rather than in the current fragmented manner by one of the national clearing houses. But the failed trades are likely to continue, as will the debate over who is right: Angel’s description of the problem as a “minor blemish that offends my sense of order” or Sommers’ view of failed ETF trades as an unrealized systemic operational risk.