ProShares UltraShort China ETF: Caveat Emptor!

Originally written on Feb. 9, 2008.

A week ago I concluded that the UltraShort FTSE/Xinhua China 25 ProShares fund (FXP) was not working as advertised. I decided to follow up with ProShares to get to the bottom of the mystery and ended up discovering that there is much more to these leveraged funds than meets the eye. In short (no pun intended) this fund is working as advertised but it's not what most people expect. And judging from the exchanges on the message boards I can assure you that there are a lot of people out there that don't understand what they are invested in. Apparently ProShares is used to the confusion because they have the subject aptly covered in all kinds of handout materials.

First, let my clarify what the target index is. It is not the Hang Seng and it is not the iShares FTSE/Xinhua China 25 Index as some people have suggested. The former is a Hong Kong index and the latter is an index fund marketed by Barclays which attempts to track the same index that FXP attempts to track. The FTSE/Xinhua China 25 Index consists of "25 of the largest and most liquid Chinese stocks (Red Chips and H shares) listed and trading on the Hong Kong exchange." Unfortunately, it's hard to find historic values of this index but I was able to pull a few key values off of stockcharts.com. Using those values I redid the analysis presented in my last blog post and pretty much got the same result:

So why am I not getting the result I'd expect? Given the 15.3% drop in the FTSE/Xinhua index I should have seen a 30.6% increase in FXP, right? Wrong! Here's why.

As explained in the ProShares Supplement of Additional Information [SAI] the fund's objective is to to match 200% of the performance of the index on a given DAY. The SAI goes on to explain that "for periods greater than one day, the use of leverage tends to cause the performance of a ProFund to be either greater than, or less than, the index performance times the stated multiple in the fund objective." It turns out that mathematically it is virtually impossible to match 200% of the index over longer periods as illustrated in this extreme example that I fabricated, where I assume that the index goes up 20% one day and down 10% the next day and the fund meets its objective each day.

As you can see, while the index ends up 8% the ultrashort fund would end down 28%, not 16%. A bit counterintuitive, isn't it?

As the ProShares folks explain it this phenomenon is caused by the leverage and the volatility of the index they are tracking. This is such an important concept that ProShares even went so far as to produce a table of estimates of how the fund would perform under different combinations of index return and volatility. What it shows is pretty staggering. With 40% volatility the fund can drop by 38% over the course of one year even if the underlying index is unchanged. With no volatility the fund can rise by almost 178% in one year while the index drops by only 40%. Unfortunately, this analysis is in the SAI and not in the prospectus. Even more unfortunate is the fact that it wouldn't have made any difference to me since I didn't read the prospectus anyway because the whole concept of this fund seemed pretty simple to me. Boy was I wrong!

So, the obvious question is why do they have a fund that tries to match the daily return of an index? Why can't you have it match a longer period? If I hold the fund for a year I expect it to match the performance of the index over a one year period. Well, the problem is that every shareholder has a different holding period and you can only get the math to work out for one holding period. So you might as well target a one day holding period.

The way I see it this holding period problem is created by the leverage. In a leveraged fund you end each day with different leverage than you started with. If you start the day with 2:1 inverse leverage and the index moves up 10% you end up with 2.25:1 leverage at the end of the day. If you were targeting the fund's long term performance for someone whose holding period began that morning you would let the leverage roll over each day since they would measure their performance relative to their starting point when the leverage was 2:1. But in that case someone who buys in on day 2 would not be buying a 2:1 leverage fund. They would be buying a 2.25:1 leverage fund. So I suspect that the fund manager rebalances the leverage at the start of each day.

This whole thing sort of reminds me of Einstein's theory of relativity. I think it would be a lot easier to just buy twice as much of an unleveraged short fund. That way you will know exactly what you are getting.

Comments

[Paulo:]

I purchased 400 shares of Proshares Ultrashort SDS on August 9, 2007 at $55 and they have, of course, bounced all over the place within a range (if you include intraday) of approximately between $50 and $65. Given the volatility and the risk, the payoff has not been worth it at all. When the going gets tough, you have to watch what is going on intraday like a hawk.
Anyway, I am down to 150 shares having sold the remainder between $62 and $64. But again, not really worth the effort and the risk.
However, I hung on to the 150 to hedge a purchase of 200 KBE close to the current trading bottom and on some days, to my horror, both KBE and SDS go up!!
You can also rationalize buying the S&P at a time which seems reasonably good to you using SDS to partially hedge.
But no, I would hesitate to do this again and a non-leveraged short position for hedging purposes seems a saner way to do things.

2008 Feb 18 05:46 PM

[root_thing:]

Wow, thanks! I own several of these funds, and I too noticed that they didn't move as expected. They should advertise these securities as short-term trading vehicles. Hold them for 2-4 days during a run-up, sell, then wait for the drop and buy them back.

2008 Feb 18 07:45 PM

[User 86999:]

Enlightening. Thank you.

2008 Feb 19 09:04 AM

[Edward Li:]

This phenomena has been well explained in Canada by Horizon Beta Pro Etf's. They even have a website with a rebalancing tool to help you achieve twice the performance of the underlying.
www.hbpetfs.com/rebala...

2008 Feb 19 12:20 PM

[Augustus:]

The problem you are recognizing is that it takes about 11% up to recover from 10% down.
If an index can move by 5% on a particular day - and the FXI has done that many times - then the leverage will cause seemingly small tracking differences to become very large. A 5% daily move will give that 10% daily leveraged move. Compounding is what causes the difference you are finding.

2008 Feb 19 01:12 PM

[S-P:]

Thanks for the very useful info. I held FXP for a while, noticed the tracking error, and being dissatisfied dumped it (of course, right before it flew up from 72 to 108...).

But given the inherent issues with longer-term tracking problems with FXP, why doesn't Proshares at least offer a non-leveraged China ETF inverse fund? All they have is Ultra... Likewise for the sector-based inverse ETFs... Anyone know?

2008 Feb 19 05:39 PM

[odenhoven:]

It is now obvious that there is a difference. Never took the time to find out why the ultra ETF did not perform as I anticipated... Thanks

2008 Feb 20 08:55 PM

[Questioning...:]

I found the article very helpful. Does anyone know what happened the price change last Friday. FXP lost more than 8% of its value on day when the market was down slightly and the Chinese market was down approximately 1.5% earlier in the day?

2008 Feb 20 09:39 PM

[locke:]

cynical: Proshares sometimes pays out very large, unannounced dividends. I held DDM for a while, and was floored when it dropped by a lot one day during an off market.

It turns out they just payed out a dividend that day.

2008 Feb 20 11:29 PM

[zcharles:]

"It is not the Hang Seng and it is not the iShares FTSE/Xinhua China 25 Index as some people have suggested. The former is a Hong Kong index and the latter is an index fund marketed by Barclays which attempts to track the same index that FXP attempts to track. The FTSE/Xinhua China 25 Index consists of "25 of the largest and most liquid Chinese stocks (Red Chips and H shares)

still unclear. the fxp is 2X of what index?

2008 Mar 12 05:27 AM

[zcharles:]

how does the fund manager rebalance the leverage each day as mentioned in the article. how does that affect performance on a given day?

2008 Mar 12 07:49 AM

[Sean M:]

I have held for quite awhile too and very very dissapointed. If SDS and QID can do it, I don't see why they can't do this either. It's just fuzzy math, and highly deceptive. Explain to me why today FXP is down 5% when last night the China index went down 5%. Shouldn't FXP be up at least 3 or 4%? And why all the intraday moves as well? Shouldn't it stay fairly stable during the US market since the Asian markets are closed??

2008 Mar 27 09:35 AM

[Anonymous16:]

THANK YOU VERY VERY MUCH !! BE WELL.

2008 Jul 29 09:34 AM

[Tehrooni:]

Your assumptions about FXI tracking the DAILY performance of Hang Seng wrong dude. If you find a correlation between the two, let me know cause I cannot find one. Take today (Dec 2/2008):

Hang Seng down by 5.1%
FXP down by 14.2%

How do you figure? Accident?

Let's try yesterday (Dec 1, 2008):

Hang Seng up 2.3%
FXP up 11%

Was this shorting Hang Seng??? Hum... doesn't look like it! The real correlation is with S$P 500. When S&P goes down, FXP goes up. When S&P goes down, FXP goes up.

Don't believe it? Just check the numbers for Dec 1 and Dec 2 for yourself!

2008 Dec 02 06:19 PM

[Gary Lucido:]

You need to reread the article. I clearly point out in the second paragraph that this fund is not meant to track the Hang Seng.

2008 Dec 09 11:23 PM