Why Short Sector ETFs Aren't So Smart

I am sure most people are familiar with the concept that if you go down 20% one day, then up 20% the next, you are still worse off than when you started (100 times 0.80 equals 80, and 80 times 1.20 equals 96). This is similar to what happens with these double-levered short side ETFs (the two-times long-side ETFs look like they do what they should), you get shorter on the way down, making bounces hurt more, because you lose more of your capital account.

So when you're frequently rebalancing, volatility nibbles away at your returns. When volatility goes to extreme levels, it eats away at your returns ... and with leverage, it devours your returns. This is essentially a short volatility position, and the short volatility position can outweigh the short index position, as evidenced by the returns in the chart. So these ETFs are not quite as effective as one would think as a mainstay in the portfolio, as a hedge or otherwise; in fact, they may be completely ineffective, or even counterproductive, at achieving objectives.

What's worse, though, is that by their very construct, these ETFs exacerbate the volatility. By bifurcating an index into long side and short side ETFs, they eliminate an "out" for the market maker, causing the market maker to actively hedge in the underliers. With a normal security, all buyers and all sellers come to a central meeting place, and buyers can be matched easily with sellers, and we reach price discovery. But when you set up a specifically one-sided instrument, rather than one common product that people can be either long or short, you contribute to dislocations.

Very few people would decide to go long an index by shorting the short-sided index ETF -- they'd just go buy the long-sided ETF. These products purposefully segment the longs and the shorts, and that, by definition, creates illiquidity. (Although I have to admit, this is an ingenious idea for the fund manager -- if they just had one product where longs and shorts could meet, some of those would cancel each other out, and they'd have less assets under management than they get by herding the bulls and bears into different products.)

So if someone buys that short-sided ETF from a market maker, the market maker does not really have "the other side" to mitigate his risk, thus he either waits for someone to unwind a pre-existing position or he goes out and shorts the underlier. This puts pressure on the underlier, which creates more interest in being short. This, magnified by the leverage, magnifies the volatility, which magnifies the negative convexity, which eats into returns. Thus the "savvy trader" who thinks he or she is doing a "smart trade" is contributing to his or her own underperformance while still having the right idea -- the wrong execution of the right concept.

Now here is a key point: This short-volatility position is kind of a compounding issue. If you compound at low yields, it is only slightly noticeable. If you compound at high yields, it becomes meaningful. Only in this case, instead of yield, think volatility. The more volatility, the more these levered short ETFs get clipped.