The Lessons of Knight Capital

NEW YORK ( TheStreet) -- The Knight Capital ( KCG) incident shows how exchange-traded funds may be victims of market disruptions.

Like the Flash Crash of May 6, 2010, the Knight Capital glitch had an effect on the ETF market.

The Flash Crash started in the E-mini S&P 500 futures market and resulted in a "systemic problem with market-wide mispricings" says Paul Weisbruch, vice president of ETF/Options Sales and Trading at Street One Financial.

A trade was executed to sell 75,000 E-minis in just 20 minutes, which led to selloffs in the ETF market as well. Mispricings in the ETF market became so extreme -- trading almost 60% from previous prices -- that ETFs accounted for 70% of cancelled trades during the Flash Crash, according to the CFTC- SEC report on the flash crash.

Giovanni Cespa and Thierry Foucault offer one explanation to the ETF price deviations in 2010 in their paper "Illiquidity Contagion and Liquidity Crashes."

Cespa says the paper "argues that in today's markets, as liquidity suppliers in one asset increasingly rely on the price of other asset classes to set their quotes, markets have become more intertwined."

High illiquidity in the underlying holdings can be transmitted to the ETF as "dealers in that market find the prices of the underlying to be poorly informative," says Cespa. The price discovery process can be disrupted by these "illiquidity spillovers," making it more costly for arbitrageurs to engage in the market, which "further impairs price discovery".

In the paper, Cespa found that "a liquidity crash seems to be much stronger for assets that are close substitutes," which was consistent with how ETFs were "relatively more affected by the Flash Crash than other securities".

Unlike the Flash Crash, the Knight Capital glitch was caused by one firm, with no major data outages or mispricings across the market. Instead, Weisbruch says, "The Knight Capital incident opened up an arbitrage opportunity for other firms whose pricing mechanisms were still properly in place."

Over 100 stocks experienced volume and price spikes, but nothing as severe as the price swings during the 2010 Flash Crash. However, some utility ETFs, whose underlying holdings were concentrated in affected stocks, saw massive jumps in volume and price. More-illiquid ETFs also saw spreads widen significantly after the Knight Capital glitch due to Knight pulling out as lead market maker.

Some utility ETFs experienced record volumes and price jumps. The Vanguard Utilities ETF ( VPU), with an average trading volume of less than 100,000 shares, traded over seven million shares on the day of the Knight Capital glitch. The VPU also saw its price jump up and down 5%, which is highly unusual for utilities, according to data from Yahoo! Finance.

VPU, along with other affected utility ETFs, held heavy concentrations of stocks affected by the glitch. Three of the affected stocks, Exelon ( EXC), Dominion Resources ( D) and Southern Company, ( SO) make up over 15% of the VPU and over 20% of the Utilities SPDR ETF ( XLU), another ETF that experienced large volume and price movements.

"For many ETFs there is a basket effect, which damps the effect on the ETF from the movement of one stock," says Weisbruch. While large ETFs might have little potential to move abnormally, smaller and more concentrated ETFs carry "more risk of volatility and can be affected more by individual stocks," Weisbruch adds.

However, the utility ETF prices and volume levels came down to normal levels and the Knight Capital glitch did not seem to cause any lasting effects on those ETFs.

A slightly more concerning event was the widening of spreads of ETFs trading less than 50,000 shares a day.

ETFs trading over 50,000 shares a day did not see the same change in spread due the Knight incident. Other market makers quickly took over, profiting from an "opportunity to book arbitrage," says Dave Nadig, director of research at These more liquid ETFs are not as dependent on market makers and created more competition for other market makers.

The spreads of some more-illiquid ETFs with lead market makers rose from 0.5 to 0.9% as a result of the Knight Capital problem. The ETFs that held Knight Capital as a lead market maker rose from 0.5% to 1.4%, according to IndexUniverse data.

Lead market makers work to keep ETF spreads tight and centering around fair value. Premiums/discounts did not increase, but spreads of more illiquid ETFs widened. Knight pulling out as an ETF market maker was "disruptive over the very short-term, causing spreads to reach untradeable, wide levels," says Nadig.

These spreads could hurt retail investors who can't get out of the quoted spreads, especially those investors using market orders that could further widen the advertised spreads. Larger registered investment advisers can call liquidity desks and get a narrower spread, but most retail investors have no choice but to follow the spreads, according to Weisbruch.

"Market makers may be hesitant at first with certain low-volume ETFs, primarily those with very little assets under management as well as tenure in terms of years since inception," he adds. "This is a byproduct of 'seed capital' and 'holdings' issues as well as the nature of the underlying basket of the ETF in focus. Eventually market makers will step in and the spreads will tighten again."

The spreads did, in fact, narrow again quickly, dropping back down to 0.79% two days after the glitch, according to IndexUniverse.

Overall, the Knight Capital debacle has been a "story of brutal efficiency for ETFs," says Nadig. Only ETFs trading fewer than 50,000 shares a day and a few utility ETFs were affected, but only for a very short period of time. Other lead market makers stepped in relatively quickly and brought spreads and prices back to normal levels.

The ETF market did prove its efficiency, but ETFs are still derivative instruments, dependent on their underlying holdings. Thus, they may be more susceptible to these market disruptions.