Some recent articles and blogs explain a strategy consisting in selling short a pair of opposed leveraged ETFs on the same underlying asset. Here is a non-exhaustive list of the pairs that people doing this usually use: FAS/FAZ, ERX/ERY, DRN/DRV, EDC/EDZ, TNA/TZA, AGQ/ZSL. The idea is to profit by the inherent decay of leveraged ETFs while keeping two market-neutral positions. This article demonstrates that it is a very bad idea.
First, if you want to profit by the decay of leveraged ETFs, you have to understand the main reason for this decay. This is called beta-slippage (contango and intrinsic volatility are secondary reasons for ETFs embedding futures and options).
To understand what is beta-slippage, imagine a very volatile asset that goes up 25% one day and down 20% the day after. A perfect double leveraged ETF goes up 50% the first day and down 40% the second day. On the close of the second day, the underlying asset is back to its initial price:
(1 + 0.25) x (1 - 0.2) = 1
And the perfect leveraged ETF ?
(1 + 0.5) x (1 - 0.4) = 0.9
Nothing has changed for the underlying asset, and 10% of your money has disappeared if you are long. If you are short, your gain is 10%. Beta-slippage is not magical: it is just the normal behavior of a leveraged and rebalanced portfolio. The previous example is simple, but beta-slippage is not a simple mathematical object. It cannot be calculated from statistical and probabilistic parameters. It depends on a specific sequence of gains and losses. It means that any strategy based on beta-slippage is questionable because its results are unpredictable, even with a statistical and probabilistic model of the game. This is a theoretical and sufficient reason to avoid doing that.
The second argument has to do with the market neutral balance. In fact, you cannot keep it market neutral. If you want to profit by the phenomenon of beta-slippage, the pair rebalancing must be significantly longer than the ETFs rebalancing, which is daily. So there will always be a discrepancy in the pair from the first day after every rebalancing. With so volatile instruments, it makes the strategy very sensitive to the rebalancing dates.
The following charts show simulations of holding equal-weight short positions on FAZ and FAS since January 2009, rebalancing them every four weeks.
The first chart is calculated with a starting date on 1/02/2009.
The second chart is calculated with a starting date on 1/16/2009.
Starting two weeks later not only transforms an apparently winning strategy in a loser, it also transforms a decent drawdown in a horror story.
The results above don't take into account the borrowing rate of such instruments, which are typically between 3% and 9% a year. This is an additional drag. Moreover, the rates are not constant, which makes the concept of "expected return" quite fuzzy.
Last but not least, the continued availability of the borrowed ETFs is not guaranteed. Your broker can ask you -at any time and without justification- to buy back a short position in the current trading day. If you cannot do it, the order is automatically forced. The only guarantee is to get a price between the low and the high of the day, and the "market neutral" balance falls apart (possibly at the worst time).
People making money with this technique can consider themselves as lucky up to now. It is much wiser to bet on proven and documented biases. For example, I have described paired switching and seasonal strategies, with references to academic articles. People wanting to verify, backtest themselves, and maybe improve these strategies can follow this link.
Disclaimer: Charts courtesy of Portfolio123.