Even ETF shorts care about exposure.
I love trolling through the short data and seeing how the other half lives. I’m pretty much a long-only, long-term investor (though I’ve put on a short once or twice back in the day). But there’s this whole world of hedge funds and speculators that live on pairs trades and short-selling that I find fascinating.
For a long time, the data has suggested that ETF shorting was largely a liquidity-driven game. The biggest, most liquid ETF in any space got all the short action. It makes sense—bigger ETFs are easier to get a borrow on, and the high level of liquidity means you don’t have to worry so much about the ketchup bottle problem (you can get in, but not out).
Underlying Coming Into Focus
But increasingly I’m starting to see short activity that suggests that the exposure of the underlying, not just size, is important.
Consider the case of the two biggest industrials ETFs, the nearly $9 billion Industrials Select Sector SPDR (XLI | A-88), and the second-place finisher, the $1 billion iShares U.S. Industrials ETF (IYJ | A-92). Over the long term, you could be forgiven for thinking any difference between these two ETFs was incidental: just over 3 percent in total over five years.
And indeed, for most of those five years, short-sellers who wanted to bet against General Electric and Union Pacific and Caterpillar used the giant XLI as their proxy.
But that may be changing. Here’s the chart of the dollars short in the two funds over the last year:
When I see something like that, it’s pretty clear to me that someone—a large someone, as we’re talking about roughly $300 million in short exposure—has abandoned one fund in favor of another.
Why Change Underlying
So why might the shorts be more interested in being the bad guy in IYJ than XLI? A few reasons.
The first is simply cost. When you’re a short-seller of an ETF, expense ratios are your best friend. XLI is cheap at 16 basis points. IYJ is comparatively expensive at 46 bps.
If you look at your actual realized tracking difference, it’s the same: On a median one-year holding period, XLI will be 28 bps behind the index it tracks, whereas IYJ will be 59 bps behind. Think of that as a source of revenue for the short-seller.
After all, if the actual indexes are completely flat, the value of his short will increase by those amounts every year, just as a long holder of each would see their investment decline.
So that’s 30 bps, guaranteed, in favor of IYJ, and in fact, explains 1.5 percent of the difference in returns over the five-year window all by itself. But there’s another thing going on here.
XLI and IYJ don’t track the same index. XLI tracks a concentrated portfolio of 64 companies pulled from the S&P 500, with the top 10 holdings responsible for 47 percent of the ETFs. IYJ by contrast has 224 holdings, with less concentration.
In general, we’d tell investors that for a sector play, you want to minimize single-stock risk and capture the full market, and that would lead you to IYJ.
But as a short-seller, IYJ’s different portfolio has another characteristic that makes it attractive: its lower yield. IYJ’s current portfolio holdings have a dividend yield of 1.75 percent versus XLI’s 1.99 percent. Since short-sellers must pay the dividends out as an expense, that’s another 24 basis points in favor of IYJ.
Of course, it’s also possible that the short-seller in question who moved $300 million from one position to the other may also think that the stocks inside IYJ are more likely to go down—smaller-cap stocks, rather than General Electric. Professional services firms (which IYJ favors) versus, say, aerospace (which XLI favors).
Either way, I see moves like this as an enormously good sign, because it suggests investors—even the negative nellies of the world—are really starting to look under the hood of their ETF exposure, and looking past the on-screen size and liquidity.
At the time this article was written, the author held no positions in the securities mentioned. Contact Dave Nadig at [email protected].