Taking Advantage of Leveraged ETFs with a Risky Arbitrage Strategy

This post is yet another demonstration of why holding leveraged ETFs for a longer period is not worthwhile. It is also a brief examination of how the less risk averse among us might take advantage of this fact.

I wrote a while back about the dangers of holding leveraged ETFs over the longer term. As everyone should know, these instruments are designed to be used for short term trades. You can see this by looking at any of the leveraged bear ETFs' highs and lows. Even while the market is making new 52 week lows, many bear ETFs are off their highs, some by as much as 50% or more. This is because the longer one holds them, the more likely it is that volatility will kill performance. Morningstar recently summed this up in a two part video series.

So if holding the leveraged ETFs for a long time may not be very profitable, what about selling them? Here is a pretty risky arbitrage strategy that should work as long as the market stays volatile. This one is best left for the gamblers among us.

Sell short the leveraged ETFs in pairs. That is, sell short both the bull and bear ETFs for the same underlying index for the same dollar amount. For example, short $10,000 worth of BGZ and short $10,000 worth of BGU. This position is partially hedged. Since the ETFs essentially mirror one another's movements, as long as the market does not take one direction for too long, the position should be relatively stable. For example, suppose BGU goes up 10%. BGZ should fall 10%. You've neither gained nor lost anything. Any underperformance of the ETFs relative to the underlying index is the short seller's profit.

Take a look at the following most liquid 3x ETF pairs, and their performance since November 19, 2008 (the first date on which it was possible to buy/short both simultaneously).

Large cap stocks: BGZ and BGU.

click to enlarge

Energy: ERY and ERX.

Small cap stocks: TNA and TZA.

Since November 19, BGZ is up around 10%. Its counterpart BGU is down close to 50%. Shorting both of these in equal dollar amounts would have produced a gain of around 40%. Since November 19, TNA is down around 50%. Its counterpart TZA is also down (although not nearly as much). Although the ETFs trade in opposite directions, one could have made money on both. The same is true for ERX and ERY from November 19 until March 5th. ERY is down around 10%, while ERX is down over 50%.

I've used various different periods, and for the most part, one ETF in each pair is down more than its counterpart is up. Occasionally, over the very short term (i.e., intra day), both ETFs in a pair were up.

If this seems too good to be true, you are right to be skeptical, for there are plenty of risks.

I say above that the strategy of shorting the bull and bear pairs is mostly hedged. It is not completely hedged. For one thing, although the ETFs trade inversely (e.g., if one goes up 1% the other falls 1%), the movements are inexact. It often happens that one ETF rises more than its counterpart falls, and vice versa. A more important reason that the strategy is not completely hedged is that while the ETFs can go up more than 100%, they cannot go lower than 0. In shorting these ETFs, one's potential losses are unlimited. Consider UYG and SKF, where one would have lost money by shorting both:

One way to guard against this would be to purchase out of the money calls (say 100% out of the money) on each position, but the cost of this insurance would likely make the enterprise not worthwhile.

Other risks include liquidity issues as well as difficulty in borrowing the shares. One's broker may also close out a position at the most inconvenient time.

Although the strategy seems to work well over the medium/longer term, it might be safer and wiser to close out the position as soon as it is possible to realize some predetermined gain, say 5% or more. Close out the position then start one anew. One must watch it like a hawk.

I am not going to try this out myself. I am too scared of short selling (perhaps irrationally so) and like to replicate short strategies with puts. Unfortunately, the puts on these ETFs do not go out past October, the strike prices are too low, and the premiums are too high.

Comments

[squark62:]

in general the authors approach is a good place to start. there are other ways to hold leveraged ETFs LONGTERM to safely hedge downward market direction without all the hand holding. i hold several leveraged (long & short) ETFs w/o options using a quantitative arbitrage algorithm to the market. have been successful doing so since OCT-08. 1) the algo takes slippage and tracking errors into consideration. weights in the portfolio are based on relative changes and the model will tell when the hedge is about to broken and take everything off the table. 2) proshares SDS, SSO, DXD, DDM are very liquid and make the best pairs to trade based on the model used. bid/ask spreads are very very small so no major concern for arb attacks elsewhere. 3) options can't be traded in retirement accounts so some other method is needed to provide a hedge. bonds currently aren't very much a permanent safe haven here and who wants to hold treasuries or cash when it's possible smoothly hedge your way around volatility. 4) pair trading holding the underlying, not the options, in this manner does NOT require hand holding or watching the market like a hawk. quit the opposite. a quantitatively weighted pair such as SDS & SSO or DXD & DDM can be traded based on EOD results. i will grant that it is prudent to rebalance once a day unless the portfolio is weighted in favor of daily market direction. mathematics is the key and a steady hand. ofcourse most quants don't divulge key components of their plan. but either way, there's no free lunch.

Mar 08 09:30 AM

[Augustus:]

I have had this trade on for a while. Short both IYF and SKF and also Short both IYR and SRS. The problem with my trade is that the IYF has only gone one way, DN, do that SKF has only gone one way, UP. For the trade to work, there needs to be lots of wiggles both up and down. In those cases, both sides eventually lose as suggested. I'd sure be interested in reading about how the positions are adjusted to account for a continuing move in one direction.

Mar 08 02:07 PM

[squark62:]

On Mar 08 02:07 PM Augustus wrote:

> I'd sure be interested in reading about how the
> positions are adjusted
> to account for a continuing move in one direction.

i've recognized that around the web there aren't many places providing such a service to retail investors. i'm in the process of rolling out a service for a fee to do just this. the plan is not to recommend the sale of certain securities - only information to be used at the traders discretion.

Mar 08 04:09 PM

[NickH:]

Like anything else, you just have to try it and get used to it by watching how it behaves. This is a move for traders who have the time and patience to set at a computer and take advantage of the arbitrage. Another trade is to be long the 3x's etf and write the call option against it as a kind of cowardly way of locking in whatever gains there are. If it makes you more comfortable, while you are experimenting, try it. But the big gains are made by figuring out the trend, and then buying the directional that accentuates it and riding it.

Mar 08 04:32 PM

[BobC:]

This strategy is not a new idea. The difficulty in pulling it off (in part) is being able to short the shares. Some of these etf's are not very liquid and your broker may have trouble finding shares for you to short. A better approach that I have used is to obtain the short exposure synthetically by way of a collar (short OTM call and buy OTM put). Initially, this trade behaves like actually shorts on the stock. Over time, you build up a cushion via the price you entered and the OTM strike price.

Mar 08 05:04 PM

[kiltumper:]

Enough with the decay thing already. If you want to see for yourself how these things move over the long term, there are plenty of places where you can get historical daily prices for whatever index you want. Put the numbers in a spreadsheet and multiply daily % changes by 2 or 3. Things are not as simple as they seem.

Mar 08 09:17 PM

[Kevin Walmsley:]

What kind of margin would be required to short-sell offsetting ETF positions?

Mar 09 12:39 AM

[trader39:]

This approach is brilliant! This is like being the "house" in Vegas.
btw, I would suspect you would need to meet full margin requirements for both positions even if ofsetting ETFs

Mar 10 03:43 AM

[trader39:]

Tried to open up some of these positions this morning, broker indicated "shares not available to short" in many of these ETFs. Anybody have any advice how to proceed?

Mar 10 12:31 PM

[squark62:]

this sort of strategy can be accomplished w/o short selling per se nor using options. check my profile and send me an email.


On Mar 10 03:43 AM trader39 wrote:

> This approach is brilliant! This is like being the "house" in Vegas.
>
> btw, I would suspect you would need to meet full margin requirements
> for both positions even if ofsetting ETFs

Mar 10 01:45 PM

[trader39:]

OK I've thought about this a lot in the last day or so, some things to consider regarding this arbitrage strategy.

1. Many of these ETFs have distributions, a lot were fairly large and paid out in December, this is not reflected in the graphs shown above

2. This statement is misleading: For example, suppose BGU goes up 10%. BGZ should fall 10%. You've neither gained nor lost anything.

here is the explanation, suppose you open up a short etf pair position in BGU/BGZ and they are both trading at $100. after a 10% movement you have $110 in BGU and $90 in BGZ, no problem you say. But then on day 2, suppose BGU goes up ANOTHER 10%. now BGU is up to $121 but BGZ only goes down to $81. Your short position is now at a $2 loss. any time BGU goes up more, you will lose more money, you need a reversal to take advantage of the "slippage"

It seems that this strategy is only valid if it is certain that significant reversals will occur in these funds. (Like Augustus mentioned "lots of wiggles" are good) In general it appears that in the long term with any reasonable momentum either up or down will kill this strategy.

Conclusion: the term "risky arbitrage" is an oxymoron

Mar 11 03:23 AM

[squark62:]

all trader39's example illustrates is if you buy x # of shares of BGU and BGZ and ETFs hold them and do nothing else, you will loose money in the long run. this is not what the article advocates. the options trader readjusts buy buying further out of the money calls at a discount and sells in the money puts for a premium.

an alternative to trading options for an arbitrage risk/reward is to re-balance the pairs trade to compensate for the non-linear tracking errors (i.e. slippage) by quantitatively readjusting the weights of the ETFs. this requires a little more math than h.s. algebra. with patience and a steady hand, holding the leveraged ETFs longterm is no longer a liability but like a coiled up spring ready to release potential energy. follow the web link to learn more.

On Mar 11 03:23 AM trader39 wrote:

> OK I've thought about this a lot in the last day or so, some things
> to consider regarding this arbitrage strategy.
>
> 1. Many of these ETFs have distributions, a lot were fairly large
> and paid out in December, this is not reflected in the graphs shown
> above
>
> 2. This statement is misleading: For example, suppose BGU goes up
> 10%. BGZ should fall 10%. You've neither gained nor lost anything.
>
>
> here is the explanation, suppose you open up a short etf pair position
> in BGU/BGZ and they are both trading at $100. after a 10% movement
> you have $110 in BGU and $90 in BGZ, no problem you say. But then
> on day 2, suppose BGU goes up ANOTHER 10%. now BGU is up to $121
> but BGZ only goes down to $81. Your short position is now at a $2
> loss. any time BGU goes up more, you will lose more money, you need
> a reversal to take advantage of the "slippage"
>
> It seems that this strategy is only valid if it is certain that significant
> reversals will occur in these funds. (Like Augustus mentioned "lots
> of wiggles" are good) In general it appears that in the long term
> with any reasonable momentum either up or down will kill this strategy.
>
>
> Conclusion: the term "risky arbitrage" is an oxymoron

Mar 11 07:06 PM

[Augustus:]

BobC, thanks for filling in some blanks for me.
When I the SKF was recalled from my short position and I had to sell it, I sold the deep ITM calls. That would make it behave very much as the shart. The problem then came that the IYF did finally rally, the SKF did finally tank, but the calls developed a large time premium and quit tracking correctly. The puts would have really helped to balance that out. I'm not sure about the calculations on the balancing act and predicting how the deltas will change (effect of gamma) but the implied volatility premiums in these can get pretty big. I hate to be a buyer of that.


On Mar 08 05:04 PM BobC wrote:

> This strategy is not a new idea. The difficulty in pulling it off
> (in part) is being able to short the shares. Some of these etf's
> are not very liquid and your broker may have trouble finding shares
> for you to short. A better approach that I have used is to obtain
> the short exposure synthetically by way of a collar (short OTM call
> and buy OTM put). Initially, this trade behaves like actually shorts
> on the stock. Over time, you build up a cushion via the price you
> entered and the OTM strike price.

Mar 26 09:10 AM