In recent articles (see here, here and here) I have discussed simple equity-treasury hedge strategies that short inverse leveraged ETFs like the Direxion Daily S&P 500 Bear 3X Shares ETF (NYSEARCA:SPXS) and the Direxion Daily 30-Year Treasury Bear 3x Shares ETF (NYSEARCA:TMV). For example, one simple hedge strategy is to short SPXS and TMV in a 50%/50% split with monthly rebalancing (covering at the end of the month and then shorting again). The reason this type of hedge strategy works is because of the negative correlation between equity and treasuries in general.
In my previous articles, these hedge strategies have been backtested using ETFreplay software and have shown the potential of excellent growth without significant drawdown under most market conditions. If we have a bull equity market and a bear treasury market (e.g. 2013), the strategy proved to be successful. And if we have a bear equity market and a bull treasury market (e.g. 2011 correction), the strategy proved to be successful. And if there is both a bull equity market and a bull treasury market (like 2012 and 2014 so far), the strategy proved to be successful. The only market condition that is of concern is when the equity market and the treasury market are both bearish at the same time. This has not happened in the recent past, but it could easily happen in the near future. And we need a way to tactically handle this double bearish market scenario in our hedge strategy.
This article develops a defensive add-on methodology to the basic equity-treasury hedge strategy presented in previous articles. The goal of this article is to show the effectiveness of this new defensive sub-strategy in getting out of the equity and/or treasury holdings before excessive drawdowns occur in the equity-treasury hedge strategy. In this way, we can see excellent annual growth without excessive drawdowns in all market conditions.
Although I have previously developed some simple tactical strategies for defensive purposes, I decided to take a new look at defensive methods. I really wanted a method that was more responsive (i.e. not waiting for semi-monthly updating). Thus, I decided to take a closer look at daily moving average methods to accomplish this goal. With daily moving averages, I am able to respond quickly on crossover days, when the value of the ETF crosses its moving average. This appeals to me more than waiting for two weeks for semi-monthly updates, especially if market conditions change rather quickly.
Another good reason for developing a moving average strategy is because I can backtest to the year 2000 in ETFreplay if I use appropriate proxies. This gives over 14 years of backtesting, a timeframe that includes many market conditions and increases my overall confidence in the methodology. And additionally, ETFreplay has implemented a double moving average capability, thus enabling the use of a short duration simple moving average and a long duration simple moving average, and their corresponding crossover days.
I decided to use the SPDR S&P 500 Trust ETF (NYSEARCA:SPY) for the equity representative of leveraged equity S&P 500 Index ETFs, and Vanguard Long Duration Treasury Mutual Fund (MUTF:VUSTX) for the long-term treasury proxy for leveraged long-term treasury ETFs. The idea is that when the methodology says we should be out of SPY, it is a good indication that we should also be out of the family of leveraged S&P 500 Index ETFs. The same can be said about VUSTX and the family of leveraged long-term treasuries.
Rather than just using the moving average of SPY, it turns out that using the moving average ratio of SPY and a short-duration treasury like Vanguard Short-Duration Treasury Mutual Fund (MUTF:VFISX) gives improved performance. (I used VFISX as a proxy for the iShares 1-3 Year Treasury Bond ETF (SHY) for short-term treasuries to enable backtesting to 2000.) The idea is that if the SPY/VFISX ratio is above its moving average ratio, then a risk-on position is held, i.e. SPY. If the SPY/VFISX ratio is below its moving average ratio, then a risk-off position is taken, i.e. VFISX.
I also used the moving average ratio feature on the treasury side because it produced better results. The moving average ratio that I used was VUSTX/VFISX. So either you are in VUSTX (in a bull treasury condition, i.e. falling rates) or VFISX (in a bear treasury condition, i.e. rising rates). That is the intent by using the moving average ratio methodology.
To start, I looked at the equity side of things and the SPY/VFISX moving average ratio. I tried various time durations, and the one that gave the best performance was a 200 day moving average ratio. Updates occurred at the end-of-day prices on crossover day. When the SPY/VFISX ratio went below the 200 day moving average ratio, SPY was sold and VFISX was bought. When the SPY/VFISX ratio went above the 200 day moving average ratio, VFISX was sold and SPY was bought.
ETFreplay calculations of this process are shown below for a timeframe from 2000-present. The overall total return is 170.4%. The maximum drawdown is -17.6%. This compares to an overall total return of 78.7% for SPY in a buy & hold approach, and a maximum drawdown of -55.2%.
I then included a second moving average in the calculation. Use of a second moving average means the calculation moves between SPY and VFISX as determined by the crossover days of the two moving averages. The use of a second moving average not only improved performance, but it also reduced the number of trades significantly. The best second moving average duration was 5 days.
ETFreplay calculations using the crossover determined by 200-day and 5-day moving average ratios of SPY/VFISX are presented below. The green curve labeled "Equity Curve" is really the strategy's performance. The total return is 273.6%. The maximum drawdown is -11.7%. The median win percentage divided by the median loss percentage is 2.05. The total number of trades over 14 1/2 years is 44 (22 risk-on trades and 22 risk-off trades). I think these are pretty good numbers.
During the two major bear equity periods (2001-02 and 2008), the moving average ratio strategy kept us mainly out of SPY and in VFISX. In the two major bull equity markets (2003-2007 and 2009-present), the strategy was mainly in SPY as desired. It is interesting to note that the correction seen in 2011 was also well captured by the strategy, taking us out of SPY at that time.
Please remember that the intent of using this moving average ratio methodology is to get us out of the SPY family (i.e. out of the short shares of SPXS) in bear equity markets. When a bull equity market exists, the hedge strategy will be shorting inverse leveraged equity ETFs like SPXS. And the rewards for shorting SPXS will be substantial in bull equity markets (in comparison to holding SPY long or the Direxion Daily S&P 500 Bull 3X Shares ETF (NYSEARCA:SPXL) long). So the intent here for the moving average ratio methodology is not the growth of the strategy, but how well it gets us out of SPY when the equity market is bearish (or has bearish corrections). And I think this moving average ratio methodology might be more than adequate in providing this function. And it is very responsive to market conditions, getting us out of SPXS the day after the crossover occurs.
Next, I turned my attention to the treasury side of things. Long-term treasures do not follow long duration trends like equities have tended to do (unless one says the entire time from 2000-2013 is one general upward trend); rather long-term treasures increase and decrease in short timeframes, months rather than years. Thus, shorter duration moving averages are required for the methodology to be successful in treasuries. This will mean that more trades will be necessary.
I found the double moving average ratio method with VUSTX/VFISX to be superior to a single moving average ratio approach with VUSTX/VFISX. The best moving average ratio durations were determined to be 25 days and 15 days. The ETFreplay results are presented below for the 2000-present timeframe. The green equity curve is mislabeled, and is really the strategy's curve. The total return is 203%, and the maximum drawdown in -10.6%. There were 74 risk-on trades (selling VFISX and buying VUSTX) and 74 risk-off trades (selling VUSTX and buying VFISX) in this timeframe, or about ten total trades per year.
Although the trades are not included with these published results, it was seen that during bullish times, we would have been in long-term treasuries (i.e. shorting TMV in our hedge strategy) for the most part, and in bearish times, we would have been in short-term treasures (i.e. VFISX) for the most part. This is exactly what we want to accomplish with this defensive strategy.
In conclusion, I propose using the defensive approach developed in this article to get out of our short equity position of SPXS and our short treasury position of TMV during bearish times. When equity and long-term treasury markets are bullish (as determined by the defensive sub-strategy), we should short SPXS and short TMV in a 50%/50% split with monthly rebalancing. When the markets go bearish, the defensive sub-strategy should place us in SHY. There will be times when the equity market is bullish and the treasury market is bearish, and vice versa. There will also be times when both markets are bullish and times when both markets are bearish. The equity-treasury hedge strategy combined with the defensive sub-strategy should be able to successfully handle all four scenarios.
I am not able to carry out the combined calculation (i.e. the hedge strategy plus the defensive sub-strategy) on ETFreplay, but the pieces taken separately have been shown to be very successful in backtesting, with the potential of 50% annual growth and maximum drawdowns of -15% or less. This hedge strategy with the defensive sub-strategy is best used by investors who can stand some volatility, and it should constitute only a small portion of their retirement portfolio. This strategy is definitely a high risk/high payoff type of strategy.
Disclosure: The author is short SPXS, TMV.
The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it. The author has no business relationship with any company whose stock is mentioned in this article.