For those interested in leveraged ETFs this post details the results of some backtesting for both long and short leveraged ETFs. Two pools were created one long and one short and then tested as separate portfolios holding two ETFs each month. All ETFs in both pools were two-times leveraged. The results follow are explained below.

2X Long ETFs

A pool of eight equity ETFs and five non-equity ETFs was used. Each month the ETFs were ranked by the momentum-volatility algorithm and selections of 2 ETFs were made. In cases where only one ETF satisfied the minimum momentum-volatility ranking, that ETF and SHY (a cash equivalent) were selected. Each selection represented 50% of the portfolio for that month. If no ETFs made the cut, then the portfolio was 100% SHY. SPY, the S&P 500, was used as a benchmark against which performance was judged.

The backtest return was run from January 1, 2007 through June 28, 2013. The total return for the 2 portfolio leveraged ETF strategy was 243%. That compared with a return for SPY (S&P 500 ETF) for the same period of 30%. The maximum drawdown was 46% versus a 55% for SPY. Volatility was 38% versus SPY of  24%. In 21 of the months, gains or losses were in double digits, several exceeding 20% and one nearly 40%.

The performance in terms of return was impressive, but the portfolio was too risky to trade even, I think, for the most risk-tolerant individuals.

2X Short ETFs

A pool of eight equity and four non-equity 2X Inverse ETFs was created. Each month the ETFs were ranked by the momentum-volatility algorithm and a selection of 2 ETFs were made. In cases where only one ETF satisfied the minimum momentum-volatility ranking, that ETF and SHY (a cash equivalent) were selected. Each represented 50% of the portfolio for that month. If no ETFs made the cut, then the portfolio was 100% SHY that month. SPY, the S&P 500 was used as a benchmark against which performance was judged.

These results were disastrous. The overall return for the period was a loss of 33% versus a gain for SPY of 30%. The maximum drawdown was 80% versus 55% for SPY. Volatility was 45% versus SPY of 24%. In 12 of the months gains/losses exceeded 10%. In back to back months there was a gain and a loss exceeding 30%.

In the first three years, 2007 2009, the strategy produced a 330% return. Short strategies work well when markets are collapsing. The rest of the time it gave up all of the gains plus 33% of the beginning principal.

Long and Short 2X Strategy

A strategy of combining the Short and Long 2X ETFs did not produce satisfactory results. It only equaled the SPY over the trading period with much greater volatility.

Conclusion

Trading 2X Leveraged ETFs does not seem useful. There was no surprise regarding this conclusion.

Long and short leveraged ETFs are profitable when the markets move in their direction. Yet the leveraged aspect clobbers a portfolio when markets move opposite the ETF. Volatility was extremely high, as was to be expected. Combining long and short pools into an overall pool did not provide advantages.

The next article will look at blending leveraged ETFs into pools of unlevered ETF to determine whether the momentum-volatility algorithm is able to distinguish when and whether to invest in them.  Backtesting will be performed to see whether they work in a broader trading strategy.