ETF Volume Spikes Don’t Indicate Direction

July 02, 2014

Why reading too much into one-day-ETF volume anomalies will lead you down a primrose path.

Here’s a fairly typical email I had in my box this morning:

“I know that daily volume drives the price of individual stocks and shows market conviction. Since ETFs use creation units, I suspect daily volume (in comparison to the 50-day-volume-moving average) does not drive ETF pricing as it does in stocks. If that is true, how should daily ETF volume spikes be interpreted, and what exactly are they telling me?”

First off, I love getting emails like this. It’s an articulate understanding of how ETFs and their underlying securities interact. The problem is, the relationship between ETFs, volume, creation/redemption activity and underlying securities isn’t clean. It’s highly complex and fuzzy. And fuzzy drives investors nuts.

Let’s take a very recent example from Monday. Here’s the chart for the Guggenheim S&P Equal Weight Industrials (RGI | B-81):


This certainly seems like a poster child for a volume-spike day. Its average daily trading is under 10,000 shares. It traded more than 200,000 shares on Monday. So how should we interpret this spike? Well the first thing I generally look at is what the volume looked like on the day. Here’s the tape for RGI, only looking at trades of more than 1,000 shares:


On this 200,000 share trading day, there aren’t a ton of big blocks changing hands. There’s the big one in the afternoon for 45,335 shares, or roughly a $4 million block. That’s not so exciting. Running the rest of the tape shows just a ton of small trades.

When we look back on flows (which take at least a day’s lag) we will likely see a creation unit on the tape—that $86.56 was slightly over the market at the time, which means it was likely a buyer, which means an AP could offset the “overpriced” sale with a fair market value creation and book the arbitrage.

So what does it “mean” to see a spike like this, getting back to the emailer’s question? The somewhat snarky answer is “someone wanted to own a lot of equal-weight industrial S&P 500 stocks.”

To see if we’re talking about just a handful of investors or some sort of cataclysmic flood, you can simply look at the “hot” stocks inside RGI and see how they did on this same day. Here’s how Iron Mountain (IRM) did during the run-up to our spiky day:



Find your next ETF