My mission: To construct a portfolio without U.S. Treasuries or U.S. Large Caps, that will have a higher yield, a higher total return over the past 3 years, and half the volatility of U.S. large caps.
I have seen a lot written about 60/40 portfolio asset allocation, usually referring to 60% equity and 40% bonds, although sometimes reversed for more conservative investors into 60% bonds / 40% equities. I wanted to describe a twist on this strategy that I am running in one sub-portfolio. It is a twist featuring 60% bonds, and 40% equities, and the allocations within each category are somewhat unconventional.
- 60% Vanguard long term U.S. investment grade corporate bond ETF (NASDAQ:VCLT)
- 20% international equity index ETF (NYSEARCA:VEU)
- 20% U.S. REIT ETF (NYSEARCA:VNQ).
Thus, there is no U.S. treasury bond component, and no general U.S. equity component.
Reaching for Yield
This approach reaches for yield in several ways:
- Longer duration bonds
- High-grade corporate bonds rather than Treasuries
- Significant REIT exposure
The blended yield of this portfolio is 3.9%, by the way. For reference purposes, the S&P 500 index ETF (NYSEARCA:SPY) currently yields 2.1%, and the 10-year U.S. Treasury ETF (NYSEARCA:IEF) yields 1.7%.
This is an all-Vanguard ETF portfolio. This means expense ratios among the lowest in the industry (as low as 0.12% for several of these ETFs), and also commission-free trading within a Vanguard brokerage account. This brings the costs and minimum investments down to a level that essentially anyone can implement.
I have previously written about the embedded "duration leverage" represented by long-term bonds. In this new analysis, I have also replaced U.S. treasuries with investment grade corporate bonds. I feel that high-grade companies exhibit better balance sheets than the U.S. government and have historically returned more than treasuries - especially in the current zero interest rate environment. The additional yield one gets (4.2% in long-term corporate bonds versus 3% in long-term Treasuries) should serve as an additional cushion in case interest rates finally do spike upward. Plus, in that kind of environment, the U.S. dollar will likely be getting devalued or inflated, and the international equity and U.S. REIT exposure may be advantageous and hopefully would offset any decline in the bond component.
If stocks decline, the relatively small exposure of 40% equities is a buffer. Also, if general U.S. equities decline, it is possible that REITs might fall less due to their higher yield, and also that international equities might outperform (since they are starting out at better valuations, based on things like price to book and price to earnings, at the time of this analysis and my initial investment).
Results of Backtesting
I backtested the results against the S&P 500 ETF , using strict buy and hold for three years. I chose SPY because as a U.S. large-cap index, it is the dominant component of any conventional portfolio that contains 60% or more equity exposure. Even if other indices are used, correlations are so high (>0.90) that the effect is the same. This was a pre-planned analysis of the asset mix described above, so I did not run the analysis multiple times, or test various mixes, or test the inclusion of SPY. The whole goal was to build upon prior analysis I have presented here on correlations and long-term bonds to see if I could achieve a higher-yielding portfolio with lower volatility that would outperform U.S. stocks.
The three-year equity curve, shown below, is slightly superior to SPY by 4% - not a huge difference, and with a large contribution from the long-term bond fund, as shown in the second figure below. (For reminder, I think it is worth repeating that the portfolio being tested only contains 40% equities!) But look at how much lower volatility that higher return was achieved with: 9% in this portfolio vs 18% with SPY. You can see this visually in comparing how much smaller the drawdowns are on the portfolio compared with SPY.
Additional statistics included below, thanks to ETFreplay.com:
Another comment I'll add is that the past three years were a very equity-friendly environment, so I find it noteworthy that a portfolio with only 40% equities was able to outperform the total return of SPY during the same time frame. I did NOT cherry-pick time frames for this analysis. I picked three years and ran with it. In fairness, the duration leverage of long-term bonds undoubtedly helped, and is unlikely to be repeated going forward. (However, if you know when interest rates are turning around and going up, please leave a comment below for posterity, so we can come back in a few years and check your results!)
Anytime you add multiple asset classes you will tend to get lower volatility - that is modern portfolio theory 101. But recall that this combination portfolio contains no general U.S. equities (such as SPY). The simplest explanation is that asset classes have become more correlated, so even without including a large-cap index like SPY, my portfolio in effect contains assets that behave like it (in my recent article on correlations, I found about 90% correlation with international ETF (NYSEARCA:EFA) and 84% correlation with US REIT ETF (NYSEARCA:IYR), vs total U.S. market ETF (NYSEARCA:VTI) - pretty high correlations post-2008).
Avoiding general (non-REIT) U.S. equities mainly carries the risk that U.S. stocks would significantly outperform all three asset classes in this portfolio. That certainly is possible, but as I presented recently, post-2008 most non-bond assets have become significantly more correlated. There is also the risk that spreads between corporate and Treasury bonds could widen, meaning that Treasuries could go up in value, and corporate bonds could go down. (Of course, that would mean new record-low Treasury rates, and possible a deflationary economy.)
In summary, I have shown what I consider to be a novel portfolio that captures additional income in several ways, and generates a higher total return and with half the volatility than the U.S. large-cap stock market, at least when backtested over the past three years. It also uses three extremely low-cost ETFs, which can be bought and sold commission-free at Vanguard. Finally, the current yield on this portfolio is 3.9% (compared with 2.1% on the S&P 500). While not for everyone, this analysis shows that one does not have to go far off the beaten path to achieve superior risk-adjusted returns.