I have recently written a number of Seeking Alpha articles on Tactical Asset Allocation (TAA) strategies for growing retirement savings and mitigating risk. One such strategy is the popular Super Simple Savings Strategy, i.e. the SSS Strategy. It is composed of two sub-strategies: 1) the SSSEquity Strategy (also known as the CSD-Bond Strategy) and 2) the SSSBond Strategy. After publishing the article on the SSS Strategy, I was asked if it would be feasible to include multiple equities in the SSSEquity ETF universe. SSSEquity currently has the Guggenheim Spin-Off ETF (CSD) as the lone representative equity in its ETF universe. CSD has excellent five-year results, and is a logical choice if only one equity is to be used.
However, there is merit in including multiple equities rather than having just one. So I set out to incorporate three of my favorite high-growth equity ETFs into an equity ETF universe: 1) CSD, 2) First Trust IPO Index ETF (GURU) and Global X Top Guru Holdings ETF (FPX). Global X Solactive Social Media Index ETF (SOCL) was also considered for inclusion in the ETF universe, but it was rejected because its volatility is much higher than the other three. The idea is to have multiple equity ETFs that correlate well to each other, but may have different long-term growth rates. In a favorable trending equity market like we have today (well, maybe), it is advantageous to own the fastest growing equity ETF, and the fastest growing equity ETF may vary over time.
So I developed a new equity strategy I call SSSEquity2 that I am presenting in this article. For SSSEquity2, the entire ETF universe is composed of four ETFs: the aforementioned CSD, GURU, and FPX, plus a cash filter. As a first option, I decided to use Barclays Low Duration 2-Year Treasuries (SHY) as the cash ETF when none of the equity ETFs passed the moving average filter. Thus, the SSSEquity2 Strategy (option 1) holds one of the three equity ETFs, or else it owns SHY.
I used relative strength to determine the fastest growing equity ETF. The ETFs are ranked in terms of six-month growth and two-month growth, and the ranks are weighted 60% and 40% respectively. The weighted ranks are added together to arrive at a final ranking. The highest ranked ETF is bought at the start of each month, unless it does not pass a four-month moving average filter. If the top ranked ETF fails the filter test, then SHY is bought. This is a very simple strategy. Please note that monthly updating is employed for SSSEquity2.
The SSSEquity2 Strategy was backtested from 2007-2014 using the ETFreplay software. The backtesting started in 2007 because the inception dates of FPX and CSD were in 2006. I always desire longer backtesting, but there is no way to do it with the ETFs selected, and there are really no proxies for them that go back farther in time. The backtested results shown below have a CAGR of 21.6%, a Sharpe Ratio of 1.20, an overall volatility of 14.8%, and a maximum drawdown of 21.3%. Although these performance numbers are not too bad, they can be improved by using a different filter, as I will show below. Another deficiency of these results is that the yearly return for 2008 is negative. One of my goals in TAA strategies is to never have a negative return year.
After getting these results with SHY as the cash filter, I decided to try iShares Barclays Long-Term Treasury Bond (TLT) as the cash filter. TLT has a strong negative correlation with CSD, FPX and GURU, indeed TLT has a strong negative correlation for equities in general. But the question remains: will the negative correlation between equities and long-term bonds hold up in the current market environment? The negative correlation to equities held up nicely in 1987, 2001-2002, and 2008, all down cycles for the equity market, as well as the correction in the equity market in the second half of 2011. Whether it will hold up in today's conditions with the potential of rising long-term interest rates is yet to be seen.
The overall performance using TLT as the cash option is superior to the first option (cash=SHY). The backtested results shown below have a CAGR of 29.7%, a Sharpe Ratio of 0.91, an overall volatility of 18.3%, and a maximum drawdown of 19.1%. Every year produced positive returns; the minimum full-year return was 10.5% (2008), and the maximum was 54.3% (2009). The yearly returns substantially beat the S&P 500 ETF benchmark (SPY) every year, especially in 2008 and 2011 when the SSSEquity2 Strategy has over 47% higher gains than SPY. $100,000 invested in this strategy at the beginning of 2007 would now be worth $630,500. In contrast, $100,000 invested in SPY would now be worth only $146,000. The only real negative for this option is the Sharpe ratio of 0.91 that is lower than the first case because TLT was used as a filter instead of SHY.
And as a final cash option (option 3), I decided to use the bond that is tactically selected from the best of the bond strategies I have developed, the Bond-Only sub-strategy of the AssetBlend Strategy. In this option, when SSSEquity2 selected cash, the cash ETF is the top-ranked bond in the Bond-Only Strategy. There was a real challenge in backtesting this last option: ETFreplay was not able to handle a cash ETF that varied in time. So it was not possible to perform a backtest for this option without getting into an excel calculation. I decided to calculate the difference in performance between this third option and the second option (TLT = cash) by looking at each period individually. Here are the results when I substituted the top-ranked bond ETF in place of TLT.
Between 2008 and 2014, there were five timeframes when TLT was selected in SSSEquity2:
1) between February 1, 2008 and May 1, 2008. Option 2 (cash = TLT) had a total return of -1.4%, while option 3 (cash = top-ranked bond) had a total return of +3.8%
2) between July 1, 2008 and April 1, 2009. Option 2 had a total return of +20.0%, while option 3 had a total return of +27.2%
3) between June 1, 2010 and October 1, 2010. Option 2 had a total return of +9.5%, while option 3 had a total return of +4.5%.
4) between July 1, 2011 and November 1, 2011. Option 2 had a total return of +21.1%, and option 3 also had a total return of +21.1%.
5) between June 1, 2012 and September 4, 2014. Option 2 had a total return of -1.7%, while option 3 had a total return of +3.4%.
It can be seen from the numbers that option 3 produces better overall performance than option 2. Thus, I have decided to use option 3 as the cash option for SSSEquity2.
As a final exercise, I combined the SSSEquity2 Strategy with the Bonds-Only Strategy from my AssetBlend Strategy to produce a conservative tactical strategy I call the SSS2 Strategy. The mix is 60% SSSEquity2 and 40% Bonds-Only Strategy. The backtested results from 2007-2014 are shown below using cash option 2 (cash = TLT) for SSSEquity2. Based on my preliminary calculations shown previously, the results will only get better if cash option 3 is used for SSSEquity2 in place of cash option 2.
The CAGR is 23.9%, the Sharpe Ration is 0.84, the overall volatility is 13.3%, and the maximum drawdown is 13.6%. The annual returns are positive every year, with a minimum annual return of 7.8% in 2007 and a maximum annual return of 41.0% in 2011. The SSS2 Strategy has better returns than the benchmark Vanguard 60-40 Balanced Mutual Fund (VBINX) in every calendar year, and it avoids the large drawdown (36.0%) of VBINX. $100,000 invested in the SSS2 Strategy at the start of 2007 would be worth $456,100 at the present time. $100,000 invested in VBINX in 2007 would only be worth $151,400 at the present time.
I plan on investing in the SSS2 Strategy (using option 3 for cash) starting February 3, 2014. The recommended ETFs to be bought on February 3, 2014 are Vanguard Long-Term Government+Credit Bond ETF (BLV)(40%) and FPX(60%). I will post the recommendations for the SSS2 Strategy at the end of each period along with all of my other ETF recommendations. The posting will be on my Instablog.