Ultrashort ETFs Can Work if You Trade Them Correctly

Bashing the ProShares UltraShort ETFs seems to be a favorite blog topic these days but it has also incited a few defenders of these ETFs to respond.

To review, those who dislike these ETFs have pointed to the fact that they do not seem to have come anywhere near delivering the expected result of double the inverse of the underlying index. Indeed, they might not have shown any gain at all even if the underlying index lost significant value.

Evaluating the arguments --

Let's take a look at the following scenario: buy an UltraShort ETF when the corresponding long index ETF falls below its 200-day simple moving average.

We looked at four pairs of ETFs as described in the following tables. The Start Date corresponds to when the index ETF fell below its 200-day MA. Standard Deviation is included as a measure of volatility.

This first table shows the results using ETFs that correspond to two major market averages.

	QQQQ	QID	SPY	SDS
Start Date	1/9/08		1/7/08
End Date	1/16/09		1/16/09
Initial Price	47.75	37.01	137.87	47.48
Recent Price	28.90	58.95	83.11	81.11
Gain/Loss	-18.85	21.94	-54.76	33.63
Percent change	-39.48%	59.28%	-39.72%	70.83%
Standard Deviation	7.30	11.57	19.28	13.83

This next table shows the results of using ETFs that track two individual sectors: real estate and financials.

	IYR	SRS	IYF	SKF
Start Date	6/18/07		7/23/07
End Date	1/16/09		1/16/09
Initial Price	76.30	77.93	109.27	75.94
Recent Price	32.22	59.32	33.86	158.78
Gain/Loss	-44.08	-18.61	-75.41	82.84
Percent change	-57.77%	-18.61	-75.41	82.84
Standard Deviation	12.63	21.69	20.68	27.35

The outcome --

What we found was that one of the sector UltraShorts, SRS, the UltraShort Real Estate ETF, did indeed perform horribly over the long haul. The other ETFs we examined didn't do too badly. They may have lagged the hoped for 2X inverse of the underlying index but they did way better than the underlying indexes did as stocks plunged this year.

At this point it is important to reiterate what ProShares says about these ETFs: they are meant to double the inverse of the underlying index on a daily basis, not over extended periods of time.

Others have written how compounding over time can affect these kinds of leveraged ETFs so I won't go into detail here; suffice to say that they are correct and that it tends to keep the gains less than the expected 2X.

Use the right tool for the right job --

What is important to point out is that indulging in these ETFs is essentially a commitment to, at the least, market timing or even active trading. This by itself implies that the intent is probably not long-term hedging.

So if you are willing to actively trade the UltraShort ETFs, then they will work quite well. This does not necessarily mean day trading. Handsome gains were had in SKF, for example, if an investor bought this ETF each time it fell back to around its 200-day MA, held it for a few weeks or months and sold each time it spiked. For most folks, of course, this is easier said than done but if you were good at determining stops, it was still quite possible.

What is more noticeable (and useful) from the data presented above is that a bearish investor, betting against a major market average like the S&P 500 or the NASDAQ 100, could indeed have held the corresponding UltraShort ETF and realized a quite reasonable gain over the course of the last 12 months.

Sour grapes?

So in my opinion, it seems that those who are railing against these ETFs are not using them as they were intended and are judging them by criteria they were not intended to be judged by. In addition, this attitude that the UltraShorts are awful investments is very much a result of what end date is used. A few months ago, when SRS on two occasions rose to over $175, I don't think anyone would have been complaining about this ETF. Furthermore, it is clear that the volatility of this ETF is very high and wild price swings are to be expected. Lesson: be careful, pay attention and take profits when you can. Otherwise, don't complain.

Major market averages do pay off --

Playing individual sectors, especially the highly volatile financial and real estate sectors, may yield unpredictable results.

In a cyclical bear market, however, using the UltraShorts that correspond to the major market averages has actually worked out quite well, even if holding these ETFs for longer periods than recommended. I have provided the details for the S&P 500 and the NASDAQ 100 but the results are the same for the Dow 30 Industrials (comparing DIA and DXD) and the Russell 2000 (comparing IWM and TWM).

The rules of the game --

It seems there are four rules for using the ProShares UltraShorts:

1. Pay attention. When an ETF jumps 20 or 30 points in a day, it is probably a good idea to take some profits.
2. Some of the most volatile sector ETFs should only be used as trading vehicles over shorter periods of time. This should be evaluated on a case by case basis. Also, see rule #1.
3. The ETFs that correspond to major market averages (referred to by ProShares as "short market cap" ETFs) can be considered "dual purpose": use them for longer bets on cyclic bear markets or as trading vehicles if you are nimble enough.
4. Have an exit strategy. Set a point at which "enough is enough" and have the discipline to take profits. Don't assume a 2X return over longer periods of time.

Are there any other rules you think should be added?

Disclosure:

small positions in SDS and QID

Comments

[Lawrence J....:]

Trading, per se, may not the be the key to using Ultrashorts. Balancing one's holdings may be sufficient. For example, if you buy an ultrashort and the index moves down 5%, your ETF should move up 10%. Now ask yourself what would a move in the index back to the day before's level mean to your ETF. The answer is that it would fall by more than it went up the day before, and you'd be out money. If, on the other hand, you sold a carefully computed chunk of the ETF, you could expect that a fall of 4.76% in the index (about what it would take to get back down to the original level), and the resulting 9.5% loss in the ETF, would lose a dollar amount equal to the previous day's gain. My arithmetic says you need to unload about 13.6% of your ETF holding to reach that result. I may be wrong, but it seems to me that balancing one's holdings on a daily basis would give you a pretty good chance of having the ETF do over the long haul what it is designed to do on a daily basis.

Jan 27 06:57 PM

[Fitz919:]

I've both traded, and paper traded these leveraged and inverse leveraged ETFs since March of 2008. You can easily capture gains in less than 1 minute, and can suffer losses just as quickly.

You need to know which direction the market is going, and why. If you don't know that much, you shouldn't be trading that day.

Once you buy one of these, do not walk away from your screen, set a market order to sell, so that you can sell instantly if it goes into reverse. If it's going up you either grab a predetermined gain, or set your sights on a larger move...say 5% or more.

A cautious trader can make money every single day.

Jan 27 07:01 PM

[cma:]

"from the data presented above is that a bearish investor, betting against a major market average like the S&P 500 or the NASDAQ 100, could indeed have held the corresponding UltraShort ETF and realized a quite reasonable gain over the course of the last 12 months."
This is absurd. His own stats disprove it. IYF was down 75.4%. By arbitraging with SKF, you would have needed + 150.8%; instead you got + 82.8%, i.e., SKF fell behind by 68 percentage points? Tracking error, anyone?

Jan 27 07:21 PM

[Lawrence J....:]

Sorry, I screwed up my example.

Trading, per se, may not the be the key to using Ultrashorts. Balancing one's holdings may be sufficient. For example, if you buy an ultrashort and the index moves down 5%, your ETF should move up 10%. Now ask yourself what would a move in the index back to the day before's level mean to your ETF. The answer is that it would fall by more than it went up the day before, and you'd be out money. If, on the other hand, you sold a carefully computed chunk of the ETF, you could expect that a gain of 5.26% in the index (about what it would take to get back down to the original level), and the resulting 10.52% loss in the ETF, would lose a dollar amount equal to the previous day's gain. My arithmetic says you need to unload about 13.6% of your ETF holding so that the 5.26% gain in the index will lose the same amount of dollars as you made the day before. I may be wrong about this, but It seems to me that balancing one's holdings on a daily basis would give you a pretty good chance of having the ETF do over the long haul what it is designed to do on a daily basis.

Jan 27 07:36 PM

[Fighting Yoda:]

The author has made extremely poor (dumb) arguments. One would expect the short ETFs to do a exact job of tracking, or at worst a decent job. As all the data has shown they do not track at all- they did the opposite – these have gone down when the underlying index has fallen significantly. That is how bad these are. Just because they went high at some point or other does not mean or prove anything. These at best are trading vehicles for a day trade nothing else. If you want a week or month trade use options.

These double shorts etc should be banned or investors should simply stop buying them.

Jan 27 10:24 PM

[SHB:]

If holding for long term does not deliver, why not short them for the long term? In fact, short both the long and short to maintain a market neutral position.

Jan 28 12:42 AM

[Lawrence J....:]

SB-Tiger: The math on inverse funds makes clear that daily inverse percentage moves will leave an inverse fund below the tracked index over periods of market fluctuation. Because this result is algebraically inevitable, one should not expect accurate tracking and so should not be surprised - or even angry except at oneself - when it doesn't track..

The math is simple. Suppose the index and the etf are both at 100 on Day 0. On Day 1, the market goes up 5% to 105 and your etf goes down 5% to 95. On Day 2, the market returns to 100. What does your inverse do? It goes down 5/95 (the inverse of the 5/95 increase in the index). But 90/95 of 105 is only 99.47, less than 100. Poof, you're tracking is off. Compound that result over several fluctuations, and of course the inverse does not "track."

2x inverse funds are even worse. Same 5% drop in the index produces a 10% gain in the fund. Return to 100 produces a drop of 10/95 from a level of 110, leaving 98.42. That means the index hasn't moved, but you're down 1.58%. Compound that.

The only way out is to trade for a daily burst, or adjust your holdings frequently so that the value of your etf holding as a percentage of your portfolio stays constant.

Jan 28 12:55 AM

[Lawrence J....:]

Damn. I quit. I apologize to readers. I don't seem able to proof my ups and downs. Seriously, I'm sorry. But if you do the math right, you'll get the results I was trying to show.

Jan 28 12:57 AM

[bobbobwhite:]

All this talk about accurate tracking! All I know is I bought a 2X reverse commodity ETF at 40 and it is now over 80 in less than 4 months. I don't care about exact tracking(picky picky), I care about gains. And, a double is always real good.

However, for safety I do have a following sell stop set at 10% under close.

Jan 28 01:07 PM

[todehls:]

SHB, there has been some discussion about shorting the "ultras", and I have done some analysis and done exactly what you suggested and shorted both the ultraLong and ultraShort. I've done this with UYG/SKF and DIG/DUG. For some smoothing I sold short the positions on two different dates, Jan 12 and 21. I shorted as close to $500 of all four as the price at the time would allow. Results (mid day as I type on Jan 29): UYG up $26.33 and SKF up $82.98. DIG down $20.54 and DUG up $38.56. The resulting net across all four shorted positions is $127.35, which is 3.16% of the $3944.70 in positions shorted. I'm doing it with my Tradeking account which requires a 75% margin on these ETFs, so only $2958.53 of my purchasing power was used to take these positions making the gains in half a month around 4.3% of purchasing power used--not too bad.

see also:
www.stockpickr.com/mem.../
which contains a link to another blog post I've authored about my analysis as well as a link to a couple of Excel files I used to do my analysis.

Jan 29 02:02 PM

[Rich from W...:]

This ignores (as it says) distributions/reinvest... such as, about 30% from SRS about a month ago. In the long run that may balance out, but right now the effect is distorting.

Jan 30 09:10 AM

[todehls:]

SRS declared a short term capital gain distribution of 4.56656 at a time when it was trading around 57 which is about 8%. SKF declared none. The calculations in the Excel file I used for analysis use dividend/distribution adjusted figures.

Jan 30 12:02 PM

[todehls:]

Full disclosure: my figures include a $5 transaction fee per, taking $40 off the profit to arrive at the $127.35 (on $3944.70). I guess if you trade for free (or trade larger amounts) it could be better.

Jan 30 04:32 PM

[sunbakd1:]

How are you not stopped out as these inverse funds can gap down more than 10% on afterhours or open much lower than the close? Just wondering, because it has happened to me before.


On Jan 28 01:07 PM bobbobwhite wrote:

> All this talk about accurate tracking! All I know is I bought a 2X
> reverse commodity ETF at 40 and it is now over 80 in less than 4
> months. I don't care about exact tracking(picky picky), I care about
> gains. And, a double is always real good.
>
> However, for safety I do have a following sell stop set at 10% under
> close.

Feb 03 05:29 PM