New methods for asset allocation

When you have a 60/40 allocation, your risk is actually 90/10.

So says Adrian Hussey, managing director of portfolio and risk analysis at the University of Toronto Asset Management Corporation at a recent Alternative Investment Management Association panel discussion in Toronto.

Read: Time for an asset allocation revolution

To reconfigure that portfolio to have equal equity and bond risk, you’d have to hold 41% equities and 95% bonds via leverage, he says. The efficient frontier, however, assumes no leverage; so Hussey suggests portfolio managers relax that constraint.

More asset classes

Another way to lower risk is to add asset classes like commodities and currencies, says Neil Simons of Northwater Capital. He also suggests momentum or active strategies could act like asset classes within portfolios, creating further diversification.

To create risk parity, you’d have equal exposure to your chosen asset classes.

Read: Faceoff: Asset allocation

A portfolio containing equal parts equities risk (S&P 500 and high-yield bonds); commodities risk (Goldman Sachs Commodities Index and gold) and fixed-income risk (30-year U.S. Treasuries and real-return bonds) would return 8.7% since 1977 with a 5.8% volatility and 0.58 Sharpe ratio.

A 60/40 portfolio would return 9.4% over the same period, but with 10.7% volatility and a 0.38 Sharpe ratio.

Read: From asset allocation to risk allocation

Simons also uses an economic model of allocation, which creates portfolios intended to respond to boom, bust, inflationary, stagflationary and deflationary environments.

First, he determines how various economic environments would affect each asset class. Government bonds tend to do well in low-growth and deflationary environments; equities, corporate credit and commodities usually thrive in boom times; commodities also do well in inflationary periods.

Then, he allocates assets so each economic environment is equally represented.

Read: Tactical asset allocation

“Theoretically, this smoothes out returns,” says Simons. But he adds not all asset classes respond as expected during economic turmoil, and that this approach doesn’t take asset class correlations into account.

Correlation corollary

To that point, Jason Russell of Acorn Investments says correlations don’t always behave as expected, either.

He says correlations between the S&P 500 and emerging markets are high and rising. And it doesn’t help to diversify by market cap or equity style: correlations are high between small and large caps, and between U.S. growth and value stocks.

Read: Downturn highlights value of non-correlation

Fortunately, there are still uncorrelated asset classes – and, says Russell, they can be either natural or synthetic.

A natural pair is 30-year bonds and crude oil. No law says they have to be uncorrelated, but crude oil tends to rise in inflationary times, while long-bond prices tend to fall as interest rates rise.

A synthetic pair is the S&P 500 and the HFRI Short Bias indices. The S&P 500 includes long-only large caps, while the HFRI only includes short stocks. It’s impossible for holdings to overlap, so correlations have been negative since 1991.

Read: ETF strategies for a downturn

Russell admits the Short Bias’s returns have lately been lackluster, but says such an index can be an important ingredient in an uncorrelated portfolio.

To find uncorrelated pairs, he suggests building them through a quant approach (synthetic) or understanding the drivers behind your desired asset class and finding its opposite (natural).

How much do you know?

To find the best allocation strategy, check the accuracy of your investment beliefs.

Mark Blair, director of fixed income and alternative investments at Ontario Teachers’ Pension Plan, says those skeptical of all overlays should use a simple equal-weighting strategy.

Read: Is smart beta smarter?

Or, if you understand volatility, use an equal volatility weight strategy. Believe you can figure out correlations? Use a risk parity approach. And if you think you can forecast returns, use a Mean Variance Optimization model.

Melissa Shin

Melissa is the editorial director of Advisor.ca and leads Newcom Media Inc.’s group of financial publications. She has been with the team since 2011 and been recognized by PMAC and CFA Society Toronto for her reporting. Reach her at mshin@newcom.ca. You may also call or text 416-847-8038 to provide a confidential tip.