Use access key #2 to skip to page content.

RVAspeculator (28.12)

The “Ultra ETF” CAPS experiment starts Tuesday



January 18, 2009 – Comments (12) | RELATED TICKERS: SKF , SRS , SDS

I have commented on this blog for sometime about how ultra ETF’s both short and long are horrible vehicles for periods even over 2-3 weeks.

Talking with a friend last week about what we would do with unlimited capital, I said I would short all of the ultra ETF’s both short and long and be guaranteed return.

The classic example is SKF (double short financials).   I sold it back in early July for $153 a share.  The index it was supposed to “short” is XLF (financial index). At the time XLF was trading at $20.32.
Today, just 6 months later financial sector (XLF) has lost more than half of its value and is trading at $9.68.  Where would you expect the “double short financials” ETF to be trading??

Well since the XLF lost 52% and the SKF is “double short” you should expect SKF to be well over $300 by now (+104%).  You would be wrong because it is actually trading at $154.  It is up $1 or less than 1%.   (For the more sophisticated readers out there SKF did NOT pay a distribution in December like some of the others.)

I understand that Proshares and the other guys who make these ETF’s warn people that they will underperform over the longer term, but I do not think people understand how drastic the underperformance really is.

To show people this I created a new account in CAPS called “UltraSuck”.  Here is the link:

What I did here is I found every “Double” ETF out there long and short.  To make the account completely “market neutral” I only added an ETF if there was a corresponding ETF on the long side. 

Because I had to find a matching ETF on the other side of the market (long or short) there are some ETF’s I had to leave out…  Trust me, I was tempted to just put red thumbs on all of them but I feel it would have put a bullish bias on this account and ruin the validity of the experiment.

ETF’s intentionally left out for that reason:
UltraShort 7-10 Year Treasury   PST 
UltraShort 20+ Year Treasury    TBT 
UltraShort MSCI EAFE     EFU 
UltraShort MSCI Emerging Markets    EEV 
UltraShort MSCI Japan     EWV 
UltraShort FTSE/Xinhua China 25    FXP 

I plan on never closing a pick in this account and only adding new Ultra ETF’s if more are created and they have a both a long and a short version..  For instance if a double long China ETF is created both that new fund and FXP (the double short) would be added to this account.

This should be an interesting experiment because for the first few weeks and months I expect this account will just be a middle of the road performer with 50% accuracy and a decent score.  For example, if the market were to go up the red thumbs on the ultrashorts will do well but the red thumbs on the ultralongs will do poorly.  In this example though, over time I suspect that even the ultralongs will underperform the market and over a great deal of time (multiple years) I believe ALL of these ETF’s will underpeform the S&P.   The only exception I can see is something like DXO (double long oil) which is down 90% and trading in the 2’s.. if oil were to make a big bounce it will take quite some time for that one to be to underperform the S&P.  Still at some point they are all going to lag the market and this account will be a measure of that underperformance.

Add “UltraSuck” as a friend in CAPS to follow and track the slow demise of these ill advised leveraged ETF’s.

12 Comments – Post Your Own

#1) On January 18, 2009 at 10:13 PM, stuckonempty (33.01) wrote:

I think the case against these funds has been oversimplified. A SeekingAlpha article played with some hypothetical money invested in a 2X leveraged fund and the underlying index starting in 1993. As you can see from the link below, between 1993 and 2006 the 2X fund returned 1108% the initial investment and the index (SPY) returned 407%

This is because there is a underlying long-term trend that the market goes up, so while the leveraging loss effect can really reduce your investment in bad years, the good years outnumber the bad (look at the value of the S&P overtime) and allow
the leveraged etf to make up its losses via its 2X/3X leverage.

For a more simple example, take an overused example but add on more days of good trading than bad and the 3X fund overtakes the 1X fund (each starting at a hypothetical value of 100) after just a couple days of solid gains.

1X Fund (100)         3X Fund (100)

-10% (90)                   -10% (70)
+10% (99)                  +10% (91)
+5%   (103.95)            +5%  (104.65)
+10%  (114.35)           +10% (136.05)

You can see that it's not a simplistic example. It is very  dependent on when you invest and how much more down the underlying index falls. If you catch it at close to the bottom and then see the underlying index return to when the dow was 14,000 (in who knows how many years), can you say that you won't be ahead versus investing in the index itself?

The fundamental principle of the mutual fund industry and therefore of most people's retirement account is that, on average, the value of the stock market increases over time. In fact, the S&P has given 10.4% yearly returns on average over the past 79 years through 2004. This gives you more days of green than red and being closer to the above example than what examples people tend to use when debating these funds.

A couple of articles have addressed what you speak of, but if you look at the data they portray, they do not seem to suggest these are always bad for long term investors.

Report this comment
#2) On January 18, 2009 at 10:26 PM, RVAspeculator (28.12) wrote:

Yes... I saw this article last week.  

I think the case FOR ultra ETF's has been oversimplified.  You can run all the models you want but if you look actually look at the way these guys are performing it is an unmitigated disaster.  

Look at FXP and then compare it to FXI…  look at SKF and compare it to XLF.  Look at SRS and compare it to IYR.   I realize in theory these should work out.  The problem is the “theory” is simply not working. 

The results since they were created speak for themselves.  I think over the long term this will be shown in the performance of this account.   The irony is I own some DXO but I shorted it in this account anyway just so I got them all.

Report this comment
#3) On January 18, 2009 at 10:52 PM, falang1 (< 20) wrote:

FXP and FXI are 2x short and 1x long respectively.  I'm not sure what you are trying to match up.  Overall I agree with your theory.  I tried an experiment of buying 2x gold and 2x short gold.  At some points they were both negative overall returns.  But I think the ETFs are a great tool if used properly.  And I still believe for example if you bought SSO and held it until the market went up (let's say 20% from here), it would go up more than 1x the market regardless of how long it took to get there.  It might not be 2x, but at least it would beat the index.  I just wouldn't hold on to the shorts forever.

Report this comment
#4) On January 18, 2009 at 11:59 PM, RVAspeculator (28.12) wrote:

Yes... they are 2x and 1x..  But pick any date just a few months in the past and find the value of both and then look where they are now like I did with the example of XLF and SKF which were also 2x and 1x.

I agree with you if the market went up 20% SSO would be up more than the market, but only if that 20% came in a quick period of time.  These suckers underperform so much that I believe if it took over a year to get that 20% SSO might just be underperforming, thats how bad they have been.

Report this comment
#5) On January 19, 2009 at 4:12 AM, kaskoosek (29.91) wrote:

Let us see how these work. Suppose I am bullish on oil.

This is a theoretical framework.

Oil is at 30$.


I also have 30$ and I want to buy the 3x leveraged ETF.

There are also 60$ levered money betting the same way my money is. Therefore the total amount of money used is 90$.

If p goes up to 50$, then I have made 20*3=60$.

I now have 50$ equity at oil and 100$ levered.


If p goes down back to 30$ then I go back to point a.


Suppose here I lost 5$ now price of oil is 25$.

Now I have 25$ equity and 50$ levered.

If there isn't too much transaction cost to change the weights of our levered money then these ETFs shouldn't underform by a big margin.


Report this comment
#6) On January 19, 2009 at 9:14 AM, RVAspeculator (28.12) wrote:


There IS too much of a transaction cost...  that is my point here.   What happens is every day these double longs and double shorts underperform what they are supposed to do just a bit.   So in your oil example there would be a day oil went up 10% and the double long oil only went up 18%.   That does not seem like a big deal, but this type of thing continues to go on day after day and pretty soon all the 1-2% underperformance adds up and you end up in a situation like SKF where the underlying index is off 52% yet the "double short" is up less than 1%

Report this comment
#7) On January 19, 2009 at 11:03 AM, anchak (99.89) wrote:

RVA...Great ....Wanted to do it myself.....Love the name -"Ultrasuck"

Incidentally, here's my latest blog on the subject of Leveraged ETFs...a lot of people are blaming the funds.....but you need to understand the underpinnings and then if you still want can deploy a similar strategy without most of the pitfalls:


Report this comment
#8) On January 19, 2009 at 12:32 PM, RVAspeculator (28.12) wrote:

Thanks anchak...   I read and rec'ed your post.   I actually own TBT and DXO.  I just wanted to see the underperformance of ALL Ultra ETF's.   The fact that I have both the long side and the short side of each of these will make it so UItraSuck's score = all Ultra ETF's underperfomance.    Should be interesting...

Report this comment
#9) On January 19, 2009 at 3:57 PM, jdawg1847 (37.70) wrote:

You won't notice as much underperformance in raw instruments like S&P futures.  But instruments with severe curves like in commodities will almost always underperform -- contango and backwardation assure it.  Take a look at what happened to USO.  When USO "IPOed"  it was almost at parity with the price of oil.  Today, it's quite different.

So the DXOs and what not will surely underperform over the very long term.

Report this comment
#10) On January 20, 2009 at 6:27 PM, ETFTimer (< 20) wrote:

Even though I'm using the ultra ETFs to beat the market here in CAPs, I have to admit that I like your idea, and I think your strategy will perform well in the long run. When the next bull market inevitably starts, you'll be earning 3x with the shorts and losing at most 1x with the longs. My quick first guess is that you'll achieve about 70% accuracy in the long run.

What's even more intriguing to me is that the derivatives market could implode one of these days, causing both the ultra-longs AND ultra-shorts to fall on the same day.  In that case you would clean up!

Yes, I'll be watching this strategy...

Report this comment
#11) On January 20, 2009 at 8:43 PM, RVAspeculator (28.12) wrote:

Thanks ETF,

I shorted both the 3x longs and the 3x shorts... so the account is supposed to be market neutral and not have a long bias.  I only included an ETF if I could get both sides of the trade shorted (IE: only picked DXO short because there were a double long oil to short)

So far its not working at all but I do not expect it to work for a long time.

Report this comment
#12) On March 04, 2009 at 9:44 PM, coralbro (94.19) wrote:

This strategy makes perfect sense, but would it work in real life?  If it would, why haven't I seen it mentioned more on CAPS?  If it wouldn't, why wouldn't it?  Thanks. 

Report this comment

Featured Broker Partners