In a separate post, I covered the importance of diversification, as holding multiple uncorrelated, (mostly) cash-producing assets will have a marked effect on improving a portfolio’s risk-to-reward ratio. This means holding stocks of various sectors and from different markets, high-yield bonds, corporate bonds, safe bonds, inflation-protected bonds, REITs, and some portion allocated to gold. This works to remove environmental bias from a portfolio, such that it can perform well in almost any environment. The only time when it won’t is during an economic depression. Financial assets will outperform cash provided enough time given that our entire capitalist system is predicated on this being true. Hence central banks have a clear incentive to get things moving again by pushing nominal interest rates as low as they need to in order to get financial asset returns above the return on cash. Most investors tend to concentrate their portfolios in risk assets given the returns are highest. But the high returns come at the expense of high volatility. To chase higher returns, the underlying risk profile of the portfolio needs to increase exponentially. Accordingly, a more efficient portfolio can be constructed with even a moderate amount of diversification, which is the topic of this article. Below I run through three separate portfolios – conveniently termed “aggressive,” “moderately aggressive,” and “moderate” – and discuss their historical returns and, more importantly, volatility. All have superior risk-adjusted returns relative to the S&P 500 and leverage can be employed to all three to boost returns. Each should be rebalanced just once per year. Rebalancing more frequently doesn't provide any additional benefit based on backtesting results. Aggressive The “aggressive” portfolio is geared toward achieving stock-like returns by keeping the bulk of the portfolio in risk assets. Its overall diversification still allows the portfolio to achieve such gains at just 70% of the volatility of the S&P 500. Risks between “high inflation” assets (i.e., stocks) and “low inflation” assets (i.e., bonds, REITs) are set equally with a 35/65 split. The allocations are as such: US Stocks - 35% High-Yield Credit - 40% REITs - 15% Long-Term Treasuries - 5% Treasury Inflation-Protected Securities ("TIPS") - 5% Technically this is a “50/50 portfolio” with 50% allocated to stocks and 50% to bonds, but REITs in many ways function like high-yield credit given these business are legally obligated to pay out 90% of their earnings to shareholders. With their business models, lower interest rates (and therefore lower inflation) tend to be favorable to their prospects, similar to bonds. This will naturally depend on the type of REIT, but this generally tends to be true categorically. Also, 90% of the portfolio is allocated to “risk assets” with just 10% in some form of US Treasuries. Standard and inflation-protected Treasuries of longer duration are added in as a hedge against volatility and downturns. The ideal hedge against a recession, in my opinion, are 30-year US Treasuries. Long-duration Treasuries can either be bought through an ETF such as TLT or EDV or purchased directly. Example ETFs for each: US Stocks – VT, VTI, VOO, SPY (former two provide international exposure) High-Yield Bonds – JNK, HYG REITs – VNQ, VNQI, SRET Long-Term Treasuries – TLT, EDV (or buy 30-year Treasuries outright) Treasury Inflation-Protected Securities – TIPS, LTPZ (or buy the bonds outright) If this is tested back to 2001 (the first full year of TIPS data), it would have produced a 37% drawdown, compared with the 50.8% of the S&P 500. (Click to enlarge) Using forward annual returns estimations, this portfolio – unleveraged – would perform worse by about 15 bps (per year) over the long-run using the median (58 bps at the mean). If we leverage by a factor of 1.3:1 to reflect its risk-adjusted return relative to the S&P 500, this would generate about 200 bps worth of “alpha” per year. (Alpha represents risk-adjusted returns over a benchmark.) We can also see a more even distribution in returns if we simulate out a S&P 500 portfolio over 50 years using my personal expectations of forward returns (starting at $10,000, with no additional contribution). Given the volatility of a long-only stocks-only portfolio, the returns distribution is heavily skewed right. Namely, given drawdowns on stocks can be so severe, the median portfolio doesn’t do as well as commonly expected. However, the volatility also gives a chance of a portfolio being wildly successful, so the mean is higher than the median. 100% SPY Portfolio The portfolio referenced above gives slightly more predictability in the eventual outcome, as we see the distribution become shorter and move slightly to the right. Aggressive Portfolio Moderately Aggressive This portfolio has many of the same characteristics as the first portfolio – i.e., more of an “offense first” rather “defense first” construction – but the following changes are made: 1) Stocks and high-yield are reduced by 2.5% each 2) REITs are cut back from 15% to 10% to keep its risk on par with that of high-yield 3) Long-duration corporate bonds are added 4) The allocation to regular and inflation-protected Treasuries is increased 2.5% 5) Gold was added at a 5% allocation Overall, risk assets are 70% of the portfolio rather than 90%. Allocations: US Stocks - 32.5% High-Yield Bonds - 27.5% REITs - 10% Long-Term Corporate Bonds - 10% 30-Year US Treasuries - 7.5% Treasury Inflation-Protected Securities - 7.5% Gold - 5% ETF examples for corporate bonds and gold: Long-Term Corporate Bonds – VCLT, CLY Gold – IAU This portfolio is still inflation-neutral, but maintains a bias toward a “high growth” economic environment (relative to embedded expectations). Long-duration corporate bonds are more growth-neutral. So long as they are slightly on the investment-grade side there will be no general bias for a high-growth or low-growth environment, unlike high-yield or US Treasuries, which are high-growth and low-growth assets, respectively. Gold is added as an additional hedge. It performs best in an environment similar to TIPS or inflation-protected bonds more generally – i.e., low growth, high inflation. It lacks correlation to other assets, making it a great diversifier. In terms of returns, over a multi-decade time horizon gold will likely return at only about the rate of inflation. So it’s not a great investment but what it lacks in returns it can make up with its ability to bring down the portfolio’s overall volatility given its low covariance with other assets. With a bettered risk profile, additional leverage can be used. Return-to-risk is improved. Note that annualized return figures aren’t totally relevant given backtesting is only done back to 2001 due to the TIPS data. "Portfolio 1" is the "aggressive" portfolio mentioned in the previous section, while "portfolio 2" is the "moderately aggressive" portfolio being discussed in this section. The maximum drawdown is improved by a bit more than seven percentage points. Relative to the S&P 500, expected forward returns on the unleveraged portfolio are approximately 17 bps lower in terms of annualized yield (using the median) or 110 bps annualized when using the mean. To match the S&P’s volatility, this portfolio would be leveraged approximately 1.67:1 and would be expected to generate around 400 bps worth of alpha on an annualized basis. The returns distribution also begins to center a bit more given the more equitable allotment to both “high-growth” and “low-growth” assets. 100% SPY Portfolio Moderately Aggressive Portfolio Moderate Exposure to risk assets further decreases, now down to 50% of the overall asset mix, from 70% in the “moderately aggressive” portfolio and 90% in the “aggressive” portfolio. 1) Stocks are cut back by 7.5% 2) High-yield is cut back by 7.5% 3) REITs are cut down to a 5% allocation to keep risk equivalent to that of high-yield 4) Long-duration corporate bonds are increased to 25% of the portfolio, up from 10% 5) The allocation to regular and inflation-protected Treasuries is increased another 2.5% 6) Gold remains at a 5% allocation Allocations: US Stocks - 25% Long-Term Corporate Bonds - 25% High-Yield Bonds - 20% REITs - 5% 30-Year US Treasuries - 10% Treasury Inflation-Protected Securities - 10% Gold - 5% Note that even with half the monetary value put toward risk assets (i.e., stocks, high-yield, REITs), we still have about two-thirds of the actual risk wrapped up in these given how they dominate the volatility profile. Even so, maximum drawdown is reduced all the way down to 21.3% against 29.8% for the “moderately aggressive” portfolio, 36.9% for the “aggressive” portfolio, and 50.8% for the S&P 500. Of the four asset mixes under consideration – the three portfolios mentioned here plus the S&P 500 – the moderate portfolio would perform best during bear markets, and worst during the best years for stocks (e.g., 2013). However, risk-adjusted returns are the best over the long-run, given diversification reduces the volatility profile while very minimally cutting into returns. Returns can be amplified through leverage. Given the moderate portfolio has just 50.5% of the volatility of the broader market, matching its volatility with that of the S&P 500 would entail leverage of 1.98:1. The “moderate portfolio” would expect to underperform the S&P by about 30 bps at the median. When leveraged 1.98:1, this would generate about 550 bps of annualized alpha over the S&P 500 or about 330 bps when going by the mean. Its performance distribution relative to the S&P centers even further given that returns become more linear as volatility is stripped out. It is also the most balanced of the four. Therefore, it shows the least degree of skewness in its distribution. The better a portfolio is diversified, the more even and predictable its expected performance becomes, reducing the discrepancy between the mean and median expected returns. In riskier, less diversified portfolios, the median is often far displaced from the mean given the inherent large range of eventual portfolio outcomes due to the volatility. 100% SPY Portfolio Moderate Portfolio Conclusion The three portfolios discussed above provide aggressive, moderately aggressive, and moderate strategies that improve on the return-to-risk profile of the S&P 500. To keep returns high, asset concentrations were slanted toward a "high growth" macroeconomic backdrop. But on the same token, this means less leverage can be used given assets that perform well in high growth environments (e.g., stocks, REITs, high-yield) are some of the most volatile. There is a cost to using it, though leverage is relatively cheap in today’s zero/near-zero interest rate environment, depending on where you live. The basic idea is to improve the return-to-risk ratio to its optimal point and then leverage back up to the desired level of volatility to maximize returns in the most efficient way possible. For those who are mostly hands off with their portfolios, using low-cost ETFs is generally recommended over picking individual securities. The purchase of just 5-7 ETFs can provide exposure to thousands of individual securities and set up a well-balanced portfolio with much superior risk-adjusted returns relative to the broader market.