Risk Parity: Taking the long view
In the wake of the latest market crash, the name alone garnered a lot of attention for risk parity from US investors. But as investors calm down, reassured by rising equity markets, Stephanie Schwartz asks if it has staying power
The financial crisis highlighted the extent to which equity risk dominates most ‘diversified' portfolios. A traditional 60% equity/40% bonds portfolio gets 90% of its risk from the equity portion. Investors tolerated this higher short-term risk because of the promise of long-term returns, but the market collapse proved to be more than many could accept. The result is a newly focused attention on risk and risk management on the part of investors worldwide.
In the US, where the cult of equities is very strong, there has been an alternative around since at least the 1990s. Risk parity is a new investment paradigm based on a very established principle - modern portfolio theory - that seeks to define diversification based on risk rather than asset class or investment type.
"The idea has its origins in the 1970s with modern portfolio theory. You can take the best portfolio and lever or delever it," says Michael Mendelson, a principal at AQR. "However, the instruments you needed to do this - such as futures - did not exist. The products we needed have only been around for the last 10 years or so." It took a long time to see some basic applications of established ideas.
Bridgewater Associates pioneered the risk parity style of risk management with its All Weather fund, launched in 1996. "It's interesting that the idea of risk parity goes back to the 1990s, when equity-centric portfolios served investors very well," notes Joe Flaherty, chief investment risk officer at MFS.
Risk parity "is about putting better balance in the portfolio, balancing risk exposures to generate more consistent returns," says Bob Prince, co-chief investment officer at Bridgewater. "Nearly every pension fund has concentrated exposure to the US stock market," he says, pointing out that 98% of pension fund returns are more than 75% correlated to the S&P 500.
"It is an interesting question - why do they do that? It is on purpose; they are choosing to have a concentrated portfolio," Prince observes. "The most common answer is that it makes economic sense for the fund because they see themselves as long-term investors and they can tolerate short-term risk for long-term returns. There is an important fallacy here. If you have a concentrated portfolio, you also have higher long-term risk, and you can have multiple decades of poor performance. We believe that investors fail to recognise that any one asset class presents a substantial long-term risk of underperformance, which jeopardises the ability to meet future pension obligations. Even though there are plenty of examples of this long-term risk, it seems to be underappreciated and not adequately communicated."
Although Bridgewater was the first to market a risk parity product, the firm did not coin the name. That credit goes to Edward Qian, chief investment officer and head of research, macro strategies, at PanAgora Asset Management, who used the term in a 2005 white paper. Panagora followed this up by launching its first risk parity product in 2006. "You need a good name to start a revolution," jokes Qian. The name was popularised by consultants who advocated the approach, says the firm's head of business development, Robert Job: "Some of us started talking to the marketplace, consultants picked up on the phrase, and from there it evolved into a category."
AQR started to offer risk parity investment to its clients around six years ago. "We noticed that the partners at AQR had most of our personal investment in market-neutral funds - but we were not telling any of our clients to do that," Job recalls. The firm did not launch risk parity investment as a response to investor demand - rather, it had to introduce the concept to its clients and win support.
According to Clifford Asness, AQR's founding principal, there is no doubt that risk parity is the right way to invest for the long term - the question is why people are starting to listen now. The answer seems to be the ongoing shock of the financial crisis.
"Performance during the last crisis is one factor - portfolios that used some leverage held up better in the crisis than those that did not," he notes. "But the idea I like least is that risk parity has done better in the last two or three years. The important point is that the long-term evidence is really strong. An extreme event just shows that the nuts and bolts hold together."
Some of the newfound popularity of the risk parity approach is due to a higher concern with risk management across the spectrum, says MFS's Flaherty. "There was a widespread failure of risk management in so many institutions, along with an over-reliance on traditional quantitative models," he explains. "Understanding risk management as an integral part of the investment process helps you to meet your return expectation."
While the principles underlying risk parity portfolios are shared, the various providers of the strategy apply them differently.
Bridgewater sees its risk parity fund as the ‘optimal beta portfolio'. The firm balances the portfolio by associating diversification with exposures to different drivers in the economic environment - a rise or fall in inflation or a pick-up or contraction in economic growth. "This provides a way that a fund will always be diversified no matter what the economy does," Prince explains. In other words, the All-Weather approach balances the drivers of returns, so that poor performance in one asset will be offset by good performance in another asset, based on its sensitivity to economic fundamentals. "The most significant difference in what we do is that we do not rely on any correlation assumptions," he says. "We look at the pricing of an asset related to its economic environment."
PanAgora views risk parity as balancing a portfolio based on equal risk allocation. The firm focuses on limiting tail risk and being very careful about the correlations. "We create our own risk parity underlying asset exposures," Job explains. "There will be deviation over time against traditional benchmarks but you will create a more consistent return pattern with less volatility and less fat-tail risk. That is a key appeal."
PanAgora establishes a balanced portfolio and then uses leverage to get returns from bonds. Qian calls this "a great application of modern portfolio theory". In addition, PanAgora does not sacrifice liquidity in order to gain the right balance of risk exposures, as it utilises exchange-traded derivatives that settle daily. In addition, Qian uses a proprietary method to adjust asset exposures in an extra layer of dynamic allocation, whereas many other risk parity practitioners use a static risk framework or arbitrary risk trigger to determine portfolio exposures.
To AQR, risk parity means "making everything in the portfolio matter but nothing matter too much," as Mendelson puts it. Asness elaborates: "There are two key factors. One is balancing assets by risk, not by dollars. We have essentially been doing this in some form since 1994 at Goldman Sachs. Then there is the academic notion that you take the best portfolio and then apply leverage to increase returns." For AQR, risk parity is a strategic, not a tactical, approach. "We believe one should start from risk parity, then if you do have the ability to time the market, you can start from this basis," he says.
AQR distinguishes itself on several counts. "We are among the most diversified," says Asness. "Some risk parity portfolios will skip credit because it is comparable with equities - we do not." AQR also uses a variety of vehicles, including ETFs and futures. In addition, AQR balances risk over time, not just across asset classes. "Once you have accepted that risks vary, they vary over time. We take that into account."
Implementation presents very real challenges. "It appears that theoretically there are benefits to the risk parity approach, but practically there are challenges," notes Karyn Williams, managing director at US pension fund consultant Wilshire Associates. "In the practical world, sitting in front of trustees, it may be difficult to maintain as an ongoing strategy that potentially can have long periods of underperformance or one that requires leverage, liquidity and margin."
The first step in implementing a risk parity strategy in all or part of the portfolio takes place on the governance side. "As with all things risk, I feel strongly that organisations must have governance principles in place," says Williams. "Before implementing a risk parity approach, plans need to define the objective of the programmes and understand the risk of the fund. They also need to consider risk tolerance, recognising the downside potential and operational issues." Williams notes that it could be extremely challenging to convince an investment committee to undertake a risk parity strategy if there is not a lot of understanding of risk or of the instruments used to engage the strategy. "Risk parity is a significantly different approach than traditional asset allocation," she says. Ultimately she advises incorporating risk tolerance statements into the investment policy statement.
Those CIOs who have succeeded in implementing risk parity strategies have worked with boards that have been willing to put in the time to learn about risk and then about risk parity in particular and to feel comfortable with adopting something ahead of the curve.
There is ‘maverick risk' to adopting a risk parity strategy, maintains John Meier, a consultant with Strategic Investment Solutions (SIS), which advised the State of Wisconsin Investment Board on its allocations with AQR and Bridgewater. "It takes a lot of education of the board. Even once the policy was approved, it is constantly being reviewed."
Job at Panagora sees a variety of options for institutional investors that want to implement a risk parity portfolio, either as a discrete allocation or as an overlay. "We will work with a plan sponsor to construct a risk parity overlay," he explains. "This will allow them to keep some of their existing positions while enhancing the likelihood of achieving their desired risk target and return expectation for the overall plan. A client may also invest in one of our risk parity funds, in one strike obtaining risk diversification and gaining immediate broad exposure."
Alternatively, a client can create a total return portfolio by allocating a portion of assets to risk parity as a beta portfolio and giving the balance to alpha strategies. Job also notes that clients can now choose to invest in commodities, equities, and other individual asset categories, where PanAgora has applied risk parity. "We see many applications to risk parity and offer a suite of alternatives," says Job. "We can customise solutions for plans, consultants, and platforms."
In terms of allocations, AQR takes a realistic look at what institutional investors can do. "There is a theoretical answer: we believe that risk parity is better than traditional allocations, and we would run the core portfolio on that basis," says Asness. "In the real world, everyone is benchmarked against the traditional 60/40 allocation. Any form of change is an attempt at alpha, to outperform, but risking short-term underperformance." He advises funds to be conservative, and to assume that ‘going forward is doubly bad as history for the next two to three years', while putting in ‘as much as they can tolerate with an outside manager', seeing it as an unlevered investment into a levered vehicle. "It is more important to put in the amount they can tolerate for the long term," he says.
Panagora's Job notes that it is challenging for funds to implement risk parity at the entire portfolio level because of the leverage involved. "This is a big change to make, and it can be expensive," he says.
Bridgewater avoids this problem by managing both leveraged and unleveraged portfolios. "You can build a better portfolio if you use leverage because leverage allows you to create a broad choice of assets at any level of desired return or risk, which allows maximum possible diversification," says Prince. "But you do not have to use leverage. You can apply the same concepts by buying long-duration bonds and by utilising assets that have higher volatility on a nonleveraged basis."
Indeed, it is the leverage aspect of risk parity that has created the most controversy. Risk parity advocates all assert that the traditional 60/40 portfolio relies on leverage, but this leverage is invisible because it is held by the underlying equities, embedded in companies' debt-to-equity ratios. "With PanAgora's risk parity, the leverage is transparent," says Qian - it is achieved through the use of exchanged-traded instruments, such as futures.
Asness of AQR sees the criticism of leverage in risk parity as a ‘"Luddite attack", assessing much of the criticism as being "quite naïve". Mendelson observes that investors face a choice between "moderate leverage risk or excessive concentration risk".
In practical terms, leverage presents issues. "Return expectations for asset classes change through time," says Wilshire's Williams. "Given the same level of risk, do you want to lever a portfolio? Sometimes the answer is yes, and sometimes leverage may not improve the efficiency of the portfolio. Sometimes the cost of leverage is higher than the expected return benefits of the strategy." She also points out that plans need to determine the type of leverage to use.
"Leverage does create additional risk," says Joe Flaherty of MFS, "including the ability to borrow in all environments, maintaining your line of credit." He also notes that the evolving shape of the yield curve means that leverage is not always as attractive as it is today. Flaherty also points out that risk parity's heavier allocation to fixed income exposes those portfolios to greater fixed income risk and a greater sensitivity to rising interest rates.
In some ways, the criticism of risk parity is an indication of its strength. As Qian notes: "A lot of asset managers have a lot invested in the 60/40 approach, and they have an entrenched interest in preserving this framework."
Its advocates maintain that risk parity as an investment approach has legs. It will be a long time before people fully forget 2008, as Asness puts it, and even when they do, that change in sentiment does not mean that risk parity portfolios will underperform. His colleague Mendelson concurs: "It is a solid concept. The fear level may wax and wane, but good ideas stay around."
Asness likens risk parity to the idea of international diversification - despite some rough years for internationally diversified portfolios, the principle has persisted. However, Flaherty wonders whether asset liability management (ALM) might not be the more apt comparison: "There was a lot of talk, but very few investors acted on it." He does not doubt that risk parity has some value. "Some managers have done it for a long period of time, and if you know what you are doing, it can make sense," he acknowledges. "However, given investors' tendency to lose patience, if they had put a risk parity portfolio in place at the start of the 1990s then missed out on the equity boom, would they have stuck with it?"
Even though risk parity managers have modelled performance of a strategy going back decades, these strategies remain broadly untested, and it is this need to maintain the strategy for the long term that presents one of the biggest challenges to risk parity. "Our research shows that risk reduction can be large if you can hold the portfolio for a long time, but there are periods when the strategy does not work," notes Williams. "It will take time to learn whether risk parity is supportable among investment committees for the long term. I do not see broad adoption, but rather the gradual implementation of different forms of risk parity within portions of the portfolio. Plans will test the strategies out that way. But even now, that is a very young conversation."