The “risk parity” approach to asset allocation, which has grown tremendously as an institutional strategy since first being introduced 17 years ago, is now becoming widely available as an offering for financial advisory clients.

The idea behind a risk parity portfolio is that it equalizes risk allocation throughout a portfolio by overweighting lower volatility asset classes. Because such a portfolio’s construction differs dramatically from conventional asset allocation, advisors interested in using these funds should spend some time educating themselves and their clients on the underlying philosophy.

This article can only provide a brief introduction to risk parity funds, but it will attempt to give a brief overview of the theory and some insights into how two of the largest risk parity fund managers construct their portfolios.

Each asset class in this approach has an equal amount of risk. In their recent paper, “Leverage Aversion and Risk Parity,” in the Financial Analysts Journal, AQR founder Clifford Asness, Andrea Frazzini and Lasse Pedersen describe the strategy this way: It means to be diversified by risk, not dollars.

“To diversify by risk, we generally need to invest more money in low-risk assets than in high-risk assets,” they write. “As a result, even if return-per-unit-of-risk is higher, the total aggressiveness and expected return is lower than that of a traditional 60/40 portfolio. Risk parity investors address this problem by applying leverage to the risk-balanced portfolio to increase its expected return and risk to desired levels.”

Because bonds almost always have lower volatilities than equities (except when there is hyperinflation or danger of a government default), the risk-balancing effect of the risk parity approach will be to weight bonds more highly than stocks. Even incorporating commodities and sometimes credit asset classes, as most risk parity portfolio managers do, does not change the relative weightings too much. The result is a portfolio with an unlevered expected return much more like bonds than stocks, although significantly lower risk than the 60/40 traditional asset allocation.

So, why the interest in risk parity portfolios? The answer lies in the use of leverage. In their paper, Asness, Frazzini and Pedersen show that, since 1926, a risk parity portfolio of bonds and equities has an excess return over the risk free rate of 2.20% while the “market” portfolio (the market-value-weighted portfolio of all investable assets in those classes) has had an excess return of 3.84%. However, if the risk parity portfolio is levered using the risk-free rate to match the 15% annual volatility of the market portfolio, its excess return jumps to 7.99%. They show similar results over a shorter time period from 1973, with a risk parity portfolio that includes commodities and credit asset classes in addition to stocks and bonds. Additionally, they explore the impact of the leverage margin rates on the performance and still find that the risk parity portfolio outperforms, though not quite as dramatically.

So can risk parity portfolios really produce higher mean annual returns for the same amount of risk? Until recently, this anomaly had not been explained within the context of modern portfolio theory or the capital asset pricing model.
According to Asness, Frazzini and Pedersen, published results indicate that it is the aversion to leverage that creates the market inefficiency that allows risk parity portfolios to outperform. Essentially, an investment manager without leverage has only one way to increase expected returns, which is to increase his allocation to riskier assets (such as equities), thus lowering his Sharpe ratio and moving the return-to-risk profile away from the efficient frontier.

By balancing the risk across the asset classes, the risk parity manager starts with a portfolio much closer to the historically estimated efficient frontier. Leverage can then be used to increase the expected return of the portfolio while maintaining essentially the same Sharpe ratio.

While the academic arguments for risk parity investing are still being published, various practitioners have put the principles to use since 1996. The two largest institutional managers are Bridgewater Associates, with more than $50 billion in the strategy, and AQR, which had $21.8 billion in it at the end of 2012. Interestingly, both firms used the product only internally at first to manage wealth for their principals. Several other institutional managers have also used the strategy, and retail funds for the individual investor and smaller institutions became available in 2009 with the arrival of the Invesco Balanced Risk Allocation Fund (ABRZX). AQR followed in 2010 with the AQR Risk Parity Fund (AQRIX).

The original Bridgewater offering, pre-dating the term “risk parity,” is named the “All Weather Fund,” for its ability to protect assets in all economic cycles. Scott Wolle, a manager on the Invesco fund, describes the philosophy behind risk parity investing as “identifying different economic outcomes and defending against those outcomes.” Both Wolle and Michael Mendelson, an AQR portfolio manager, view risk parity funds as a core portfolio holding. They do not recommend, however, using the funds as a replacement for traditional 60/40 balanced funds. Instead, they view them as a complement to 60/40 funds, since risk parity portfolios have higher Sharpe ratios and diversification benefits.

After 2007 and 2008, of course, many investors are understandably wary of leverage.

“The risk of leverage is forcible deleveraging,” says Mendelson. “Leverage always has the risk of being forced out of positions.” If the investor is concerned that the manager “cannot manage leverage, don’t do risk parity.”

Just as index investing becomes complicated when you choose indexes, so does risk parity when you choose strategies. The approaches can be significantly different, depending on the provider. Which indices should be used for each of the asset classes? Since the funds are global, which equity, bond, credit and commodity markets should be included? Should credit even be an asset class, given its high correlation with equities? How is the leverage managed? For those asset classes you can’t get through exchange-traded futures, how are the positions collateralized? Also, as Mendelson points out, manager opinions vary about whether the leverage of the portfolio should be adjusted as market conditions change to minimize the variation of the risk over time. Mendelson estimates that 25% of the diversification benefit of AQR’s risk parity funds comes from performing this last adjustment, while the other 75% of the improvement over 60/40 portfolios comes from the risk balancing between asset classes.

The two largest retail fund managers of balanced risk funds, Invesco and AQR, do have some significant differences in their portfolio management. For example, AQR only does a strategic asset allocation, weighting each asset class in inverse proportion to the estimated future volatility of the asset class and adjusting the leverage to achieve the targeted overall portfolio risk level. Invesco also balances the risk and leverage to achieve an 8% targeted annual volatility, but overlays a tactical asset allocation to incorporate the portfolio managers’ views on future macroeconomic trends in inflation, economic growth and creditworthiness. According to Wolle, the tactical asset allocation has improved the performance of the fund by about 200 basis points since its inception, while taking no more than 2% annual tracking risk.

Besides the differences between managers in their use and adjustments to leverage, another important consideration for investors is the risk estimate being balanced between asset classes. Both Mendelson and Wolle agree it is difficult to forecast the future correlations between asset classes. Because of this difficulty, AQR relies predominantly on volatility forecasts in choosing its weightings. Wolle, meanwhile, says the “time frame of volatility and correlations is important” and that “Invesco uses longer-term estimates” of both in its strategy.

As risk parity funds have increased in popularity and size, critics have become concerned that investors are herding into a leveraged bond bet at a time the bull bond market could be facing imminent demise. Mendelson and Wolle understand these objections, but both say their firms have put significant research into understanding and quantifying the risks associated with risk parity investing in a bond bear market. Both also clarify that their funds are not explicitly levering their bond positions, but levering the fund as a whole to produce a risk level appropriate for their investors’ preferences.

Mendelson says AQR has done an extensive historical examination going back into the 1920s and has not found an increase in bond risks when yields were lower. Invesco, meanwhile, produced a white paper by Wolle examining whether a spike in sovereign debt yields could have a deleterious impact on bond prices—and asking how such a spike would impact other asset classes in a risk-balanced portfolio. (One interesting finding in Wolle’s research is that bond futures, with their variety of settlement options, actually outperform on-the-run Treasurys in distressed bond markets.)

As a relatively new offering in retail investing, risk parity or risk-balanced strategies are attracting both significant assets and meaningful questions about their complexity and structure. While it is true that risk parity managers typically use derivatives and leverage, and that they usually have some commodities holdings, to classify the funds as alternatives is a bit of a stretch, since by far the most exposure is in long traditional bonds and equities.

Both Wolle and Mendelson compare them to traditional 60/40 balanced funds. They emphasize that they invest mostly in exchange-traded futures, and have less than 1% of their assets exposed to counterparty risk. Both managers also emphasized that risk parity strategies are best used to complement traditional asset allocation because of their diversification benefits and explicit risk-balancing characteristics. The diversification and risk characteristics of the strategy have historically offered a smoother risk/reward payoff under a variety of economic conditions.

That historical performance may prompt some return-chasing among eager investors. But Mendelson emphasizes that the strategy should be used not alone but in tandem with 60/40 funds, for two important reasons:

Risk parity “investing is not way, way better than traditional allocation, and there is not an infinite capacity for risk parity.”

Michael J. Reed, formerly a senior managing director in Morgan Stanley’s process driven trading department, is founder of MJ Reed Investment Consulting. He has successfully managed multi-million dollar portfolios for individuals and small institutions for more than 10 years.