August 24, 2015, was a day Chris Cook, founder and CEO of Beacon Capital Management in Centerville, Ohio, won’t soon forget. As the Dow plunged 1,100 points during the first five minutes of trading, he watched in disbelief as extreme market volatility halted trading among many stocks held by exchange-traded funds. Without enough liquidity to keep the ETF creation and redemption process flowing, market prices for some ETFs briefly dipped 20% or more below their net asset values before recovering their balance. 

“One health-care sector ETF showed a 30% to 40% price decline when the underlying stocks didn’t start trading properly at the beginning of the day,” Cook recalls. “I finally had to call the ETF sponsor to find out what was going on.”

Fortunately the carnage only lasted a few minutes and Cook’s firm didn’t use automatic sale triggers that would have kicked during the downward spiral. But the experience underscored for him, and others, that the unique structure of ETFs makes them a bit more complicated to buy and sell than mutual funds or stocks. 

Like stocks, ETFs have a bid price (the price an investor is offering to pay) and an ask price (the price an investor is offering to sell). The difference between the two is the “spread,” which can vary depending on market trading activity and other factors. ETFs also have the added element of market prices, the price at which they trade, and net asset values, the value of their underlying securities. 

The authorized participants that create and redeem ETF shares usually keep market prices and net asset values in line with each other. But there have been a couple of occasions when extreme market volatility upset the arbitrage process for brief periods (see the box below). Another source of potential pricing discrepancies is the growing number of ETFs that focus on smaller, more specialized corners of the market where relatively few shares of the underlying securities trade hands every day. And even with the largest, most liquid ETFs, if you trade at the wrong time of day or place the wrong kind of order, it can mean losing money.

While ETF trading practices might not seem that important, consistently getting the best prices possible can add to a portfolio’s total returns. For veterans as well as newcomers, reviewing ways to avoid disaster in a flash crash, or just to get the best possible pricing, is probably a good idea as year-end portfolio rebalancing draws near. Here are a few suggestions from experts.

Watch the clock. Prices on the underlying securities in an ETF may be unsettled when the market immediately opens, making it difficult for market makers to closely match net asset values to market values. “You need to wait until all the stocks in an ETF are open for trading and priced fairly,” says Cook, whose firm avoids placing trades during the first hour of trading. And just before the market closes, some ETFs may experience pricing glitches as market participants begin winding down trading activity. 

Use stop-limit orders rather than stop-loss orders. A stop-loss order specifies a price at which a stock should be sold, and once triggered it becomes a market order to obtain the best price available. The problem is that in a rapidly declining market that best price can plunge rapidly. A stop limit order offers more control because the trade is executed somewhere between the stop and the lower limit. 

Set price alerts. Cook prefers not to set automatic triggers of any kind. Instead, he uses alerts that indicate when an ETF is closing in on a desired sale price—and when it has hit that price. That way, a brief market anomaly such as August 24, 2015, won’t automatically trigger a sale he may not want. “We like to have a human be the last stop before a trade,” he says. 

Pay extra attention during Fed announcements. Sharp movements in an ETF’s underlying securities can lead to pricing disruptions. While especially volatile days are hard to forecast, they often happen when the Federal Reserve makes announcements about interest rates or the government releases economic data.

First « 1 2 3 » Next