Ultra inverse ETFs fall short of expectations

Aide: We’ve analyzed their attack, Sir, and there is a danger. Should I have your ship standing by?
Grand Moff Tarkin: Evacuate? In our moment of triumph? I think you overestimate their chances.

Intelledgement’s strategy is macro-based, which means we analyze broad trends and make investments designed to align with them. Thus if we expect energy and China and Brazil to do better than average in the long run, we will seek to go long (buy) securities that are likely to reflect those successes. Conversely, if expect an economic reversal that will hurt real estate values and depress consumption, we will seek to go short (sell) securities that are likely to decline in price.

Generally, we do not recommend owning (or shorting) individual stocks for most of our clients, because it is easier to be right with a more broadly targeted mutual fund or exchange-traded fund (ETF)—which invest in numerous stocks—than trying to divine the fate of a single company (which could run into accounting problems and tank even if the industry they are in does well overall, for example). And very few of our clients are comfortable with margin accounts and short selling, which is inherently more risky than buying stock (as we have explained elsewhere). Therefore, the tactic we employ to invest money on the short side when the macros so dictate is buying so-called “inverse” ETFs, such as our current position in the the ProShares UItraShort S&P 500 ETF, which is designed to “correspond to twice (200%) the inverse (opposite) of the daily performance of the S&P500^® Index,” using short sales, options, derivatives, and other relatively arcane maneuvers. This enables us to take a virtual short position without employing margin or undertaking the theoretically unlimited risk associated with an actual short position.

Now, while ETFs have been around for 20 years, it wasn’t until 1998 when State Street introduced “Sector SPDRs” that it became possible to employ a robust macro strategy using exchange-traded funds. And even then, it was not possible to bet against sectors or national indexes or commodities other than by selling the requisite ETF short. Until 2006, that is, when Proshares introduced their line of inverse ETFs, that go up when the targeted index goes down, and vice versa. So inverse ETFs, including “ultra” inverse funds—which aspire to double (or triple) inverse performance—don’t have a long performance track record.

Ultrashort ETFs—along with their ultralong cousins—have commonly been regarded as the Deathstars of exchange-traded fund investing. Afterall, it stands to reason, if you are absolutely convinced that the price of oil is going down, why be content with an mere inverse fund oil fund when you can buy an ultrashort fund that tracks 2x the inverse of any change in the price of oil?

One concern that commentators have pointed out is that the design of the funds—to reflect the inverse of fund performance on a daily basis—gives them an inherently bearish bias in a volatile environment, because mathematically, an x% move down always trumps an x% move up. That is, if you start at 100 and have a 10% up day followed by a 10% down day, you end up at 99; and by the same token if again starting at 100 you have a 10% down day followed by a 10% up day, again you end up at 99. A slow and steady move in either direction minimizes this effect, but it is likely to be augmented when volatility is high, as demonstrated by Eric Oberg in his thestreet.com article last month.

Of course, we don’t have a lot of performance data to analyze—as the inverse ETFs are so new—but let’s take a gander at the data we do have:

Security	Symbol	Inception	Cost	Value	ROI	Index	Index ROI	Rating
Cons Svs Ultrashort	SCC	02-Feb-07	46.47	84.78	82%	DJUSCY	-39%	+3
S&P500 Short	SH	21-Jun-06	55.81	72.02	29%	GSPC	-28%	+1
DOW30 Short	DOG	21-Jun-06	57.69	68.55	19%	DJI	-21%	-2
QQQ Short	PSQ	21-Jun-06	60.43	73.03	21%	IXIC	-26%	-5
Cons Goods Ultrashort	SZK	02-Feb-07	52.92	74.01	40%	DJUSNC	-24%	-8
Oil & Gas Short	DDG	22-Jul-08	57.72	68.16	18%	DJUSEN	-35%	-17
S&P500 Ultrashort	SDS	13-Jul-06	57.28	70.94	24%	GSPC	-27%	-31
DOW30 Ultrashort	DXD	13-Jul-06	51.33	53.56	4%	DJI	-19%	-34
QQQ Ultrashort	QID	13-Jul-06	60.79	57.35	-6%	IXIC	-23%	-52
Financials Short	SEF	22-Jul-08	68.32	76.03	11%	DJUSFN	-66%	-54
Financials Ultrashort	SKF	01-Feb-07	66.74	103.01	54%	DJUSFN	-55%	-57
Oil & Gas Ultrashort	DUG	22-Jul-08	26.45	25.04	-5%	DJUSEN	-35%	-75
R/E Ultrashort	SRS	01-Feb-07	62.07	50.71	-18%	DWRSF	-60%	-138

Security = the name of the exchange-traded fund (ETF)
Symbol = the symbol of the ETF
Inception = date the ETF started trading
Cost = closing price of the ETF on the first day it traded
Value = closing price of the ETF on 31 December 2008
ROI (Return on Investment) = on a percentage basis, the performance of this ETF from inception to 31 Dec 08
Index = the market index the ETF is tracking (inversely)
Index ROI = on a percentage basis, the performance of this index from inception of the associated ETF to 31 Dec 08
Rating = how well the ETF has performed relative to expectations (see notes below)

Notes: our rating is derived by comparing the ROI of the inverse ETF with the ROI of the index it is tracking and calculating how the ETF has performed relative to expectations. For example, if the underlying index declined 10% since the inception of the fund, we would expect an inverse fund to be +10%, and an ultra inverse fund—which, you will recall, strives to log 2x or 3x the inverse performance of the underlying index—to be either +20% or +30%. So, if the index has declined by 10% and an inverse ETF is up 10%, that yields a rating of 0 (zero), as it matches our expectations. For example, as of 31 December 2008, the Proshares S&P 500 Short fund (SH) was up 29% since inception while the S&P 500 index itself (GSPC) was down 28%…so that ETF has a rating of +1. In contrast, Proshares Real Estate Ultrashort fund (SRS) is down 18% since inception while the index it tracks is down 60%…we would expect SRS to be +120% and thus it has a rating of -138.

In general, the news is bad for the “ultrashort” 2x funds. Every single one of the eight ultrashort funds we analyzed are tracking an index that was down through 31 Dec 08…and therefore we would expect them all to be up twice as much as their respective index was down. This was true of only one: the Consumer Services Ultrashort ETF (SCC), which was up 82% while the Dow Jones US Consumer Services Index (DJUSCY) was down 39%. Four of the remaining seven ETFs were up, but only one of them was close to expectations: the Consumer Goods Ultrashort ETF (SZK), was up 40% while the underlying index was down 24%. The Financials Ultrashort ETF (SKF) was up 54%, but that was less than half what it should have done relative to its index, which was down 55%. And the S&P 500 Ultrashort ETF (SDS) was up 24%, less than half what it should have done relative to the GSPC, down 27%. The DOW 30 Ultrashort was up 4%…way lower than it should have been with the DJI down 19%.

From there, it gets really bad. The other three ultrashort funds were all down, even though with their respective index down, they should have been up sharply. We would have expected the QQQQ Ultrashort ETF (QID) to be up 46%…but it was down 6%. We would have expected the Oil & Gas Ultrashort ETF (DUG) to be up 70%…but it was down 5%. And the SRS was discussed above (should have been +120%, was -18%).

And it gets still worse! In every case where there are both an inverse ETF (targeting mirror image performance of the underlying index) and an ultra inverse ETF (targeting 2x inverse performance), the performance of the inverse ETF is relatively better than that of the ultra inverse ETF:

Index	Short Fund	Ultrashort Fund
S&P 500	+1	-31
DOW 30	-2	-19
QQQQ	-5	-52
Oil & Gas	-17	-75
Financials	-54	-57

In absolute terms, the SKF ultrashort financials ETF still outperformed SEF, the inverse ETF (as you would expect when they both underperformed their expectations by about the same degree). But in the other four instances, the four inverse funds not only did relatively better, but beat their ultrashort cousins in absolute ROI. The DUG and QID ultrashort ETFs actually lost money, even though their respective indices each was in the red.

While this does not prove Oberg is correct in his analysis that volatility is doing in the ultrashort ETFs, it does constitute prima-facie evidence that his conclusion—the ultrashort ETFs are underperforming—is on target.

And accordingly, it is time to consider evacuating the Deathstar.