The standard 50% stocks, 50% bonds portfolio outperforms 100% long-only stocks on a risk-adjusted basis, but can be improved.

A different allocation mix can improve the risk profile of the portfolio without hurting returns.

Those who achieve the highest returns over time will normally be invested in the highest-risk assets. Part of this is the fact that stocks have the highest duration and stocks are inherently leveraged investments. Namely, companies have debt, and in a theoretical liquidation scenario, stockholders will be paid last, increasing risk and therefore investors returns requirements.

Now some say having a diversified portfolio is best. But there is a lot of difference in opinion as to what having a diversified portfolio actually entails. A lot of investors are systematically biased to hold only stocks and to be long-only.

However, if diversified simply means you own a bunch of different stocks – and are long-only – this is still a very risky portfolio. The upside to this long-only, all-stocks approach is that it will perform the best over the long-run because of the risk youre taking on. But the returns will be very uneven and itll get creamed in down-cycles, losing about one-third its value from peak to trough and worse in some scenarios (e.g., 51% during the financial crisis).

When it comes to diversification into multiple asset classes, there tends to be too much of a focus on how to divide up the allocation purely in dollar terms, rather than focusing on risk and downside. A 50/50 stocks/bonds portfolio – a pretty common approach – sounds good on the surface because then youre diversified over more than one asset class and youre split evenly between the two.

The issue is that in a 50/50 portfolio stocks are about twice as volatile as corporate bonds, if not a little bit more. So if youre 50/50 then you still have about 70%-75% of your risk in stocks and only 25%-30% in bonds.

If youre aiming for higher quality securities on the bonds portion, then risk becomes even further skewed. A 50/50 portfolio that is split 50% US stocks and 50% intermediate-duration US Treasury bonds (NASDAQ:IEF)(NASDAQ:TLT) will have about 95% of the risk concentrated in equities. But if you pair SPY with a long-duration corporate bond benchmark fund such as CLY (CLY) youre going to be about 70/30 in terms of risk and have about 87% correlation to the stock market.

If you look at the charts below, between a 100% stocks portfolio and a 50/50 stocks/bonds portfolio, you do see a betterment in the risk profile of the returns in the case of the latter, back-testing from January 1994 to the present. Volatility of the 50/50 is 62% of the SPY portfolio and the maximum drawdown falls to 29.6% from 50.8%. This comes at the cost of just 58 basis points of annualized return.

But the overall trajectories of the two are fairly well similar given the stocks component chiefly dominates the risk profile.

If you want to bet equally on each, so that your portfolio can endure all environments equally, then you need to bet equally on each in terms of risk. However, if you buy more bonds then youre diluting your return and end up with a portfolio that tangibly lags in absolute return over time.

So how do you boost this 50/50 portfolio from its asymmetrical risk composition (i.e., risk slanted toward the equity portion) without diluting returns?

You either need to 1) drop your allocation of stocks to balance risk out 50/50 and then lever up the entire portfolio to match your volatility target or 2) you need to borrow more of the bonds up to a point where you get equality in the risk. Leverage doesnt make sense for lower-returning bonds as the cost of the financing can be as much as the return for some, but can make sense so long as youre getting sufficient yields for these assets and have access to cheap financing.

If you go with option 1, you would have roughly a 37/63 stocks/bonds portfolio. Volatility falls to just 56% of the S&P 500, market correlation is 73%, and maximum drawdown is less than half. The cost trade-off is a loss of 90 bps in annual yield, which is very reasonable.

If you leveraged this portfolio to match the volatility of the S&P 500 (1.8:1), returns would come to 15.0%. Assuming financing costs of 2.0% for each turn of leverage, this would come to 13.4%, or a gain of 4.2% over the benchmark.

If you wanted a slightly more complex portfolio than simply investing in stocks and corporate bonds to further dilute covariance (and therefore risk from inter-correlation), you could invest in a mixture of US stocks (NYSEARCA:SPY), long-term corporate bonds, high-yield credit, and REITs.

High-yield credit and REITs offer risky, cash-producing assets but thrive in a different type of environment than stocks and investment-grade corporate debt.

While stocks thrive in a high growth, high inflation environment, high-yield and REITs predominantly thrive in a high growth, low inflation environment. You could get more granular and say that it depends on the type of REIT and this is true. But on aggregate, low rates are more favorable for REITs though cash flows benefit from the higher growth. REITs are essentially a stock/bond hybrid given these entities are legally obligated to pay out 90% of their earnings to shareholders.

The long-term investment-grade corporate bonds allow for something that benefits in a low growth, low inflation period. However, as you go down the quality ladder from AAA-rated debt, the more these assets will rely on a high growth environment in order to perform well. That is, less creditworthy bonds are more likely to become distressed if economic growth underperforms. A pure low growth, low inflation instrument is basic Treasury bonds, ideally of longer duration for the benefit of higher yield.

But the issue with allocation toward Treasuries is that its impossible to achieve balanced risk with them in the portfolio given that you cant reasonably lever them up high enough. If a 30-year bond returns you 3% and each turn of leverage costs 2% (or more) on its own, then is basically undoable given the spread is so tight. Its better to find relatively safe long-term corporate bonds that yield about 4%-5%.

If we were to breakdown allocations of these four assets where risk is evenly divisible between the three different environments, it would look approximately like this:

This would give approximately one-third of your risk to high growth, high inflation (stocks); one-third to high growth, low inflation (split half and half between high-yield and REITs); and one-third to low growth, low inflation (long-term bonds).

Notice that nothing is allocated to the low growth, high inflation subset. This is potentially bad given that in a 1970s environment where you had stagflation, stocks performed poorly (no growth) and bonds performed poorly (inflation eroded yields), so this potentially would have performed at around a 2008-level drawdown, if not slightly worse.

What are good assets for a low growth, high inflation environment? Theres a few main ones – precious metals (e.g., gold), commodities, and floating-rate bonds (e.g., TIPS).

The issue with precious metals and commodities is that they dont produce cash. Its just supply-and-demand related return. They dont create economic value on their own. On top of this, theyre highly volatile. A 50/50 stocks/gold portfolio would have about 55% of its risk concentrated in gold.

Assets like gold and other commodities can go up or down 80% in a relatively short period of time. So youre looking at no long-term returns beyond its use an inflation hedge and plenty of volatility. This volatility is nonetheless counteracted to a degree through the portfolios broader diversification.

Im not against the idea of a small allocation of a portfolio (e.g., 5%) to precious metals. Its a diversifier, which holds value in itself. And it effectively has no correlation to the broader stock market over a large enough timeframe.

But one can argue that you can get stagflation asset exposure through other asset classes. For example, you can gain access to upstream oil and gas companies – which would benefit from an uptrend in oil – through the stock market. You can also add exposure to gold miners this way as well.

Stocks in these industries are effectively hybrid instruments in that they arent assets that would outperform purely in high growth, high inflation environment but could also benefit to a degree in a low growth, high inflation environment due to changes in the underlying commodity their business cycles are tied to.

But lets say we add a 5% allocation to gold. Decomposing the risk suggests that 3% of it is concentrated in gold, so risk is rebalanced toward the remaining three.

The first portfolio is termed portfolio 1 and the second is deemed portfolio 2.

If we compare returns from January 1994 forward, we get the following:

We can see that portfolio 1 and 2 are quite similar. Portfolio 2 does slightly better on a risk-adjusted basis, with a 19-bp fall in annualized volatility for only a 10-bp drop in annualized return. Its maximum drawdown is a little bit better as well.

Portfolio 2 holds just 53% of the volatility of the S&P 500. If we leverage this to match the S&Ps volatility (leverage ratio of 1.875:1), annualized returns would have come to 13.3% assuming a 2% cost of leverage.

Changing the benchmark to a macro-oriented fund

The Waddell & Reed Advisors Asset Strategy Fund (MUTF:UNASX) is one of the highest regarded global macro mutual funds. If we were to compare macro-themed portfolios we might as well compare it to a relevant benchmark to relate performances. At the very least it gives a frame of reference. This data available for this particular fund goes back to 1996.

We can see that risk-adjusted returns for these two portfolios notably outperformed this benchmark. The macro bent of the UNASX fund helped it beat the S&P 500 on a risk-adjusted basis, but its volatility is still relatively high for the amount of returns one gets in return. Portfolio 2 gives up just 72 bps in annualized yield for just 65% of the risk.

Past performance is no indication of future performance, but this illustrates that even moderately diversifying a portfolio can provide superior risk-adjusted returns. But the standard approach of a 50/50 portfolio can be improved as well. The more this can be done, in terms of collecting a variety of mostly uncorrelated assets that perform differently in different environments, the better your overall returns will be on a risk-adjusted basis.

It also avoids the conundrum of having to time the market. You also avoid the necessity of having to short to reduce market correlation, use derivatives, or deal with illiquid instruments. It can avoid be done in the context of avoiding stock/security picking altogether by using a collection of only 4-5 different ETFs. This could even be encouraged to benefit from the additional diversification benefits that these funds provide exposure to.

Disclosure:I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.