UltraShort ETF Losses Go to 'Money Heaven'

So since the autumn low close on Nov. 20, the IYF is up 8.0% through Friday night. This implies a 2x short return of -16.1%. Yet the SKF is down 69.2% over that period, reflecting an underperformance of 53.1% vs. if one had simply shorted the IYF with two times leverage.

Given that the IYF closed below its NAV and the SKF closed above its NAV, this return discrepancy is not caused by NAV differentials (and for what it is worth, on Nov. 20 the SKF closed about 6 points below NAV). So where did the money go? The answer is in the title of this article.

In previous pieces I have noted how these levered short ETFs are implicitly short a path-dependent volatility position. Sometimes the path dependency can work in your favor -- note that since March 6, the levered ETF has "outperformed," being down only 68% when the implied return suggests a loss of over 100%. (I am sure the buyers of the SKF on March 6 are comforted by that.) But this path dependency is a second-order bet that, as one can see, overwhelms first-order returns. It isn't as simple as saying there is an "arbitrage" here and that these levered funds can recapture that loss -- that money has gone to "money heaven."

The volatility in the relative performance, however, should show people that this is an entirely inefficacious "hedge" or term position vs. other alternatives. And it isn't just the financial sector -- check out how the China and real estate sector levered ETFs have performed vs. their unlevered counterparts:

3-Apr From 3/26 From 3/6 From 2/23 From 12/31 From 11/20 FXI 30.94 8.0% 28.9% 24.8% 6.4% 47.2% 30.4nav FXP 22.17 -17.0% -45.9% -46.5% -37.3% -76.2% 22.22nav implied -16.1% -57.8% -49.5% -12.7% -94.4% over/(under) -0.9% 11.9% 3.0% -24.6% 18.2% 3-Apr From 3/26 From 3/6 From 2/23 From 12/31 From 11/20 IYR 29.33 13.8% 35.0% 18.7% -19.4% 21.5% 29.36nav SRS 38.1 -28.8% -61.6% -53.8% -24.9% -84.1% 37.96nav implied -27.5% -70.1% -37.4% 38.9% -43.1% over/(under) -1.2% 8.4% -16.4% -63.8% -41.0%

The levered short China ETF UltraShort FTSE/Xinhua China 25 ProShare(FXP Quote) again shows a mixed bag vs. its unlevered counterpart iShares FTSE/Xinhua China 25 Index (FXI Quote) in terms of relative performance, whereas when it comes to the real estate sector, the levered UltraShort Real Estate ProShares (SRS Quote) has underperformed in nearly all of the time periods shown. (And as with the financial pairing, the SRS closed above NAV, and the iShares Dow Jones US Real Estate (IYR Quote) closed below.)

To be fair, the marketing material for these levered short-side ETFs says explicitly that these ETFs are only meant to track daily index movements. But every time I see that, I have to wonder why a fund company would create a fund meant for a one-day hold -- in theory everyone should go home flat every night, and the manager would have no assets under management. So it seems as though there is an implicit "greater fool theory" in place here -- by definition, the company is reliant on someone holding these positions overnight and bearing the volatility risk.

I wrote earlier that there are only three reasons I could think of why someone would buy these: 1) they were uninformed, 2) they were trying to skirt the margin rules, or 3) they were trying to engage in market manipulation. I still stand by all three of these statements. Hopefully if one reads enough of these articles, we can eliminate reason No. 1.