New Rules For ETF Trading An Upgrade

Ugo Egbunike
August 23, 2013

This week, ETF investors saw the effects of the phase II rollout of the limit-up/limit-down program (LULD). They needn’t worry—at least not as much as they think they should.

Despite the fact that trading halts in more than 50 ETFs and ETNs on Aug. 19 left investors worrying that something had gone awry in the markets once again, it was quite the opposite. Things worked exactly as they should have.

The idea of the new LULD system stems from the “flash crash” of May 6, 2010, after which it became clear there were significant issues needing to be addressed regarding market structure—especially during times of peak volatility.

After months of review and mulling through data, the Securities and Exchange Commission approved a national market system plan—limit up/limit down. The idea was to prevent problematic trades in individual exchange-traded securities by setting percentage price bands above and below the reference price of a security over the preceding five-minute period.

So how do the mechanics of the system actually work?

ETFs are covered under phase II rules of the program. Under the rule structure, price bands are constructed by applying a formula to the reference price of an ETF. The price band formula is fairly basic:

Price Band = (Reference Price) +/- [(Reference Price) x (Percentage Parameter)]

The percentage parameter is determined by the securities’ previous closing price, and under the LULD plan, most ETFs will have a percentage parameter of 10 percent.

Price Band
Price Band

*For phase II securities, price bands are doubled during the first 15 minutes and last 25 minutes of the regular trading day.

However, it’s how that “reference price” is calculated that might be problematic for the program.

The reference price for securities is calculated using either the arithmetic mean price of the ETF over the last five minutes, or the midpoint of the opening quote of the ETF from its primary market listing.

The reference price is updated after 30 seconds only if a new reference price is at least 1 percent away from the current reference price.

This can and does become an issue for ETFs that don’t trade, because the reference price is entirely based on bid/ask quotes, which are often wide and may not change as much as quotes in actively traded ETFs.

So when do halts occur?


There are really two stages to it:the “limit state” and the “straddle state.”

The Limit State

When trading for an ETF enters a limit state, the national best bid (NBB) for an ETF equals but does not cross the upper price band. It may also occur if the national best offer (NBO) for an ETF equals but does not cross the lower price band.

Under the system, if the market doesn’t exit a limit state within 15 seconds, the ETF will experience a trading halt.

View photos

The Straddle State

When trading for an ETF enters a “straddle state,” the national best bid is below the lower price band or the national best offer is above the upper price band. If the ETF is in that straddle state and trading in the ETF deviates from normal trading characteristics, the ETF will experience a trading halt.

View photos

The intention behind the program is honorable. There are serious issues that should be addressed when it comes to market structure. However, the biggest flaw comes back to the calculation of the reference price.

When the 50 or so ETFs experienced a halt on Aug. 19, the triggers were likely a result of a reference price being calculated from quotes and not actual trades. Again, that’s  because many of those securities don’t trade often enough to provide a “last price” within the five-minute time period.

Themis Trading noted in its blog this week that an algorithm might have even triggered the halts. If so, that’s completely unacceptable and regulators will have to follow up.

However, it’s important to note that, in the end, investors were protected.

Also, if the calculation for the reference price is correct, it’s unlikely that a deeply liquid security such as the SPDR S'P 500 ETF (SPY) could experience a manipulated halt due to its reference price being based on last price data.

Although there are some kinks to work out with LULD regarding quotes, I welcome it precisely because it still protects investors from executing trades against those quotes.

The program is still taking shape and likely will be amended as time goes along.

But frankly, I have no problem seeing numerous halts in especially illiquid ETFs. It forces ETF issuers to pay attention to their products, but more importantly, it keeps investors from shooting themselves in the foot when making trades.

At the time this article was written, the author held no positions in the securities mentioned. Contact Ugo Egbunike at

Permalink | ' Copyright 2013 IndexUniverse LLC. All rights reserved