U.S. Markets closed

Knight Stumbles, ETF Spreads Carry On

Dave Nadig


With the markets back from an extended and waterlogged vacation, all eyes—including ours—were on ETF trading activity today. And for the most part, it was a yawner.

Volatility and volume were nothing extraordinary. There weren’t a bunch of horribly broken trades on the open, or anything like that.

And then, Knight Capital had a hiccup, and oh how the market likes to poke at Knight.

When Knight reported that it had to shut down trading due to backup power failure, I got a dozen frantic phone calls asking if this was going to blow up trading in ETFs for which Knight is the lead market maker (LMM). The answer, it appears, is pretty much the same as it was last time:It depends.

So first, the ground rules:All we looked at were ETFs that traded over 1,000 shares today. So the truly abandoned ETFs and ETNs didn’t influence this list. If it didn’t trade, it didn’t count.

Let’s start with the most naive way of the day’s trading:average spreads. Here’s the chart, and then I’ll tell you why it’s bad:

So what does this chart show us? It shows us a typical early morning, where spreads start wide (as underlying securities start trading) and then tighten into a comfortable range, generally under 50 basis points wide.

By the end of the day, spreads pop out again as folks get ready to go home and see if the power came back on.

Wolverine looks terrible here, obviously, but remember, Wolverine is LMM on exactly six ETFs, only one of which, the JP Morgan Alerian MLP ETN (AMJ), has any real volume or traction.

It traded half a million shares today and at 2 basis points spreads. But since we’re averaging in this chart, that awesomeness is completely polluted by funds like the PowerShares Convertible Securities Portfolio (CVRT), which traded around 3 percent all day long.

It’s not Wolverine’s or PowerShares’ fault that individual convertible securities have wide spreads that pass through to the ETF, but it makes them look bad in this chart. And the 8,000 shares that CVRT traded today count just as much as the 500,000 shares that AMJ did.



But let’s focus on the other big line here, Knight.

Knight’s is the line that pops from right around everyone else at the 50 basis point level, to over 1 percent. That looks like an “aha” here, and technically, you’d be right.

Knight has just under 300 securities in our analysis that passed the threshold, and indeed, if you average out the quoted spreads on every ETF Knight is LMM for, those average spreads did widen out—just as you might suspect—when the power went out on the Knight trading floor and they could no longer make markets.

But here’s what really happened. Let’s look at the median spreads, not the average. For those of you who skipped stats in business school, the median is the imaginary middle instance of a distribution.

So looking at the median Knight ETF is like saying, “Let’s look at the whole list, and figure out an imaginary position in which 150 funds would be tighter in spread, and 150 would be wider in spread.”

That’s different than the average, which tallies a crazy 50 percent spread and a more normal 2 basis point spread and comes up with an average of 25 percent.

Wolverine still doesn’t look great here—after all, with only a handful of funds, there’s nowhere to hide—but have fun trying to pick Knight out of the sub-25 basis point spreads at the bottom of the chart.

So what creates this enormous difference between average and median? Outliers.

Knight is, quite literally, the only market participant on The Street paying attention to some of the smallest, most illiquid ETFs on the planet. When Knight’s in the market for these folks—their clients—they keep spreads rational and tight.

When someone has to break for the proverbial sandwich, there’s nobody in the market to step in and take their place. Here’s an example of what happened to the five most illiquid ETF—as measured by spreads—in the Knight portfolio today:

Those five funds from the AdvisorShares, FlexShares-Morningstar and db-X lineups completely blow up the averages.

Note that before the power hiccup, Knight was keeping these funds extremely tight for such illiquid securities. If you wanted 500 shares either way of the AdvisorShares Accuvest Global Opportunities fund (ACCU) at 10 a.m., Knight was there with a market just 30 basis points wide. But if you wanted to play in that sandbox at noon, forget about it. You needed to go home.

So the next time someone asks you about the role of market makers in the ETF business, just go back to the words themselves:They make markets.

They craft them out of client service and kept promises and present them for the trading community to use.

So, yes, when the only person willing to make your market has to go buy a new generator, it matters.



At the time this article was written, the author had no positions in the securities mentioned. Contact Dave Nadig at dnadig@indexuniverse.com.

Permalink | ' Copyright 2012 IndexUniverse LLC. All rights reserved

More From IndexUniverse.com