How might an investor reduce portfolio risk and protect profits when faced with the probability of a possible market decline? This question is particularly pertinent considering the market is bouncing around record highs. This third article on the Swensen 6 portfolio employs a strategy of reducing portfolio risk through the use of a cutoff or "circuit breaker" ETF. The same six asset classes are covered as recommended in Swensen's book, Unconventional Success. The one difference in this portfolio is the insertion of SHY as the cutoff ETF. This turns the Swensen 6 into the Swensen 7. To gain some background into the Swensen 6, here are two references to Seeking Alpha articles on the makeup of this portfolio.
Analysis: The seven ETFs used in the following analysis are:
- Vanguard Total Stock Market ETF (NYSEARCA:VTI)
- iShares MSCI EAFE ETF (NYSEARCA:EFA)
- Vanguard FTSE Emerging Markets ETF (NYSEARCA:VWO)
- Vanguard REIT Index ETF (NYSEARCA:VNQ)
- iShares TIPS Bond ETF (NYSEARCA:TIP)
- iShares 20+ Year Treasury Bond ETF (NYSEARCA:TLT)
- iShares 1-3 Year Treasury Bond ETF (NYSEARCA:SHY)
VTI covers the entire U.S. equities market and EFA does the same for international equities. Normally, I use Vanguard FTSE Developed Markets ETF (NYSEARCA:VEA) for developed international equities, but lack of historical data required substituting with EFA. VWO represents emerging markets and VNQ is our domestic real estate ETF of choice. TLT and TIP are the proxies for U.S. Treasury Bonds and U.S. Treasury Inflation-Protected Securities. Swensen does not specify specific securities in his book so the above ETF selections are mine.
The portfolio began on 6/30/2006 with $100,000 and ended on 8/15/2014. This starting date was selected as it covers both major bear and strong bull markets. ETFs have limited historical data making it difficult to go back further in time. Every 90 days following the launch, the ETFs were ranked using three metrics. 1) Performance over the last 91 calendar days. 2) Performance over the last 182 calendar days. 3) Volatility as measured by a 10-day standard deviation. A 50% allocation was assigned to metric #1, 30% to metric #2, and 20% to metric #3.
Where does the SHY (iShares 1-3 Year Treasury Bond) enter the picture? Any ETF under-performing SHY in the above ranking is sold out of the portfolio and invested in SHY. SHY was selected as the cutoff or "circuit breaker" ETF, as it has low volatility and behaves as a stable anchor to a portfolio. When any of the six critical ETFs are ranked above SHY, the percentage invested in each reaches a maximum or the percentage recommended by Swensen. Those percentages are as follows.
- VTI = 30%
- EFA = 15%
- VWO = 5% (In later writing, Swensen moved to equal percentages for developed international and emerging markets.)
- VNQ = 20%
- TLT = 15%
- TIP = 15%
In the following graph, the Swensen 6 plus SHY turned $100,000 into $224,643 while the VTSMX (Total U.S. Equities) increased to $186,379. If one had invested in the original percentages in each of the recommended six ETF and never made any changes, the ending sum as of 8/15/2014 would be $169,012. Of course, this portfolio would be out of balance by significant amounts and surely rebalancing would have occurred during the 14 years of operation. In the following graph, note how the use of SHY as a cutoff ETF protected the investor from large draw-downs during the Great Recession. More on this later.
Over the eight years of operation, the "Swensen 6" experienced only five down periods. In 2008, there was one decline of -2.7%, and in 2009 during one 90-day period there was a -2.4% drop. Those are minor considering what happened to the broad market during those painful 15-18 months. The largest decline actually happened in 2010 when there was a 4.7% draw-down. Two more drops happened in 2011 and 2013 where the dips were only -0.9% and -1.7%.
Volatility: The Swensen 6 had a standard deviation of 10.4%, while the VTSMX was three times greater at 33.1%. The "Swensen Index" came in with a 22% SD percentage.
The following information is an example of a slight modification from the "Swensen 6" described above. The "Aristotle" strategy uses a simple seven asset Strategic Asset Allocation (SAA) Plan and the portfolio is momentum filtered using SHY as a cut-off filter. This is exactly the same approach used above. Here we have equal percentages assigned to the two international equity holdings as recommended by Swensen in his later writings. Another big difference is the increased percentage assigned to VTI. TIP is dropped out of the portfolio and SPDR Dow Jones International Real Estate ETF (NYSEARCA:RWX), our international REITs of choice, is substituted.
The assets and SAA allocations are as follows:
- VTI - 40%
- TLT - 20%
- EFA - 10%
- VWO - 10%
- VNQ - 10%
- RWX - 5%
- AGG - 5%
The rules for portfolio construction are simple
- assets ranked higher than SHY in the momentum rankings are included in the portfolio, according to the SAA Plan allocations. The ranking metrics are described above.
- allocations assigned to assets ranking lower than SHY in the moment rankings are invested in SHY.
The results of the backtest are shown below:
In the first example, the portfolio was reviewed and rebalanced every 90 days. In this second example the "Aristotle" portfolio was rebalanced every 33 days. There are numerous reasons for selecting the 33-day period. 1) The review rotates the update throughout the month. 2) We avoid the wash-sale tax issue. 3) We eliminate the short-term trading fee as most of these ETFs are commission free.
Key performance characteristics are as follows:
CAGR - 9.41%
Volatility - 7.59%
Return/Risk (Sharpe) - 1.24
Max Draw-down - 6.13%
Volatility (Risk) is extremely low, especially since the test includes the 2008 financial crisis period. Holding down risk is every bit as important as going after high returns. The returns are there if one pays attention to reducing risk and uses SHY as a "circuit breaker."
This portfolio management model of risk reduction is undergoing further study. The next bear market will be the real test.
Disclosure: The author is long VTI, VWO, VEA, VNQ, RWX, TIP, TLT, SHY.
The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.