In terms of equity styles, we’ve added some bias toward small-caps in the US, purchased downside protection in energy and view European small-caps as relatively attractive value opportunities. As we discussed in previous reports, we continue to be positioned for a modest economic recovery, with accelerating wage growth and stronger household and small-business formation in the US.
May’s performance effectively reversed some trends we saw in April (Performance). For example, the US dollar strengthened while oil and commodities declined. Since April represented a sharp reversal from the trend in the first quarter of 2015, we’re now simply back on the same path as earlier in the year. The most crowded trades outperformed in May after selling off in April; we don’t think these reversals are fundamentally important beyond the normal de-risking/re-risking swings.
A substantial part of our Multi-Asset Solutions group is dedicated to helping clients achieve long-term returns while reducing short-term drawdown risks. Inevitably, this effort involves a trade-off between the costs a client will bear to achieve a comfortable journey and the desire to meet long-term return objectives and financial outcomes.
The incremental cost of risk-management strategies can be frustrating for investors, in light of the nearly six-year equity bull market. The lengthy rally didn’t present many opportunities to demonstrate the effectiveness of risk-management approaches. But we see a compelling case for having more downside protection, not less, in the current setting. Equity markets have nearly tripled from their previous bottom, most markets are less liquid today, and the outcome of experimental monetary policy remains uncertain.
In this report, we look at the performance of a few risk-management strategies and discuss some of the approaches we favor in the current environment.
Wage Growth, Rising Correlations Could Test Effectiveness of Risk-Management Approaches
Since the last bout of volatility in mid-2012, global equity markets have been relatively calm, rising by more than 50%. Many investors are still concerned about the sources of economic growth and that decelerating emerging-market asset prices had been buoyed by easy-money policies. Central banks provided ample liquidity and essentially promised to keep interest rates low. There were few signs of rising inflationary pressures, and corporations continued to deliver strong earnings growth.
A recent acceleration in wage growth and an uptick in bond volatility may increase investors’ uncertainty and test the effectiveness of risk-management approaches (Display 1). It’s too early to tell whether stronger wage growth will erode corporate profit margins, or if it will translate into better earnings growth through increased consumer demand and a better pricing environment. During the first quarter, corporate profits grew faster than revenues, which suggested that operating leverage is alive and well. That’s encouraging, but because markets have been complacent for so long, it still makes sense for investors to reevaluate their positions. If this sets money in motion, it could boost volatility.
The significant pickup in correlation between stocks and bonds (Display 2) may further complicate portfolio positioning. Higher correlations make bonds a less effective diversifier than normal, while magnifying the impact of changes in interest rates on overall portfolio performance.
This is a very different environment from the one that prevailed after the global financial crisis, when the stock-bond correlation was persistently negative: Investors viewed bonds as a safe haven and a good hedge against deflation. Conversely, the market seemed to view higher rates as an indicator of better economic growth. As such, they were seen as a positive for equities. In 2015, the perception seems to have changed: higher rates have hurt both stocks and bonds.
A Three-Pronged Approach to Managing Risk
The tendency of historical correlations between asset classes to break down from time to time is one of the reasons we developed a holistic framework for managing downside risk (Display 3). The tool set covers three main areas:
1. Traditional diversification strategies from combining asset classes
2. Tactical asset allocation to adapt risky-asset exposure to the changing environment
3. Insurance against market downside through the use of derivatives
Of course, there are obvious trade-offs between the cost of protecting a portfolio and the certainty of achieving a specific outcome. But we think it makes sense to do more than simply make a strategic decision on the nature of risk-mitigation strategies. Investors need some flexibility to take advantage of more tactical opportunities the market provides.
We looked at three popular risk-reduction strategies to better understand these trade-offs: equity long/short, risk parity and the volatility-managed investments used in variable annuities. In our analysis, we used mainly mutual fund returns data. We expect to be indicative of overall market trends, although individual investors may have very different experiences. Precise numbers aren’t available, but we estimate that these three strategies capture almost $1 trillion in assets under management. This amount translates into almost $1.8 trillion of notional investment once we consider the effect of leverage (Display 4).
The Growing Challenge of Sameness
Many traditional risk-management strategies delivered acceptable returns, but they also relied more and more on similar signals to reduce portfolio risk, or “de-risk.” This sameness is a growing challenge.
Over the past decade, risk-managed variable-annuity-based products (we’ll refer to these as “risk-managed VA strategies”) delivered returns that were virtually identical to those of long/short equities, while risk parity outperformed (Display 5). Looking closer, risk parity has delivered superior risk-adjusted returns over the past decade—its Sharpe ratio of almost 1.0 is about the same as that of long/short hedge funds (Display 6). What may be more surprising is that over that same period, long/short mutual funds and long/short hedge funds delivered almost the same 5.3% annualized return—but hedge funds had roughly half the volatility. This big disparity in risk may relate to different incentive structures and the nature of the mandates—these tend to place a bigger burden on hedge funds to limit drawdowns.
Number and nature of VA strategies changed significantly in 2008. The strong after-fee performance of risk-managed VA strategies since 2009 is virtually matched 60/40 on risk-adjusted basis. It may reflect willingness (or a requirement) for these strategies to quickly shift into outperforming bonds as volatility picks up. It may also result from a low-volatility backdrop, which allowed risk-managed VA strategies to stay fully invested in equities without breaching their volatility targets.
Assessing Potential Performance in Rising Volatility
Equity volatility—and, to a lesser extent, bond volatility—has been subdued since the middle of 2012. In many instances, volatility touched lows set during the most complacent period in recent memory, 2007 (Display 7). But investors shouldn’t expect the current calm nature of markets to last indefinitely. Greece, rising wages and geopolitical conflicts are just a sampling of events that could disrupt markets.
What kind of performance might we see from these different strategies in the face of rising volatility? We looked at six periods surrounding peaks in the VIX, which measures the equity volatility implied by options on the S&P 500 Index. On average, the MSCI World Index fell roughly 18% in the month immediately after the VIX peaked. This was followed by a rebound that recaptured about 40% of those losses (Display 8).
Similarly, a typical 60/40 stock/bond portfolio lost almost 11% after the peak and recovered 40%–45% of the losses. Both risk-parity and long/short-equity strategies outperformed the 60/40 benchmark through the entire period surrounding the spike in the VIX (Display 9). Risk-managed VA strategies underperformed by roughly 1% in the month after the peak, but recovered a third of the losses by the end of the period.
However, much of the return difference among the strategies is driven by their varying beta, or sensitivity, to equity markets and by specific requirements of their mandates. For risk-managed VA strategies, for instance, the mandate translates to de-risking faster than either risk parity or long/short equities. We think it’s more informative to assess returns adjusted for the varying beta exposures (Display 10). On a beta-adjusted basis, risk-managed VA strategies fare a bit better, while risk parity lags.
This comparison highlights how critical it is to understand the uses of risk-reduction strategies in the context of a broader portfolio. It’s particularly important to distinguish risks due to equity market movements, security selection and dynamic beta timing. Armed with this knowledge, investors can better understand which products are appropriate for their specific objectives.
Comparing Beta Exposures over Time
How have the betas of the various risk-management strategies to the market and to other factors evolved over the years? Beta to the market for risk-parity and risk-managed VA strategies declined modestly after 2008, compared with 2000–2008 (Display 11); beta for long/short-equity funds increased. The decline in beta for risk-managed VA strategies may stem from the need for more frequent de-risking during 2010–2012, which reduced the average equity exposure. The increase in beta for long/short equities is likely a result of optimism over the market’s direction and/or a response to a more macro-driven market, which limited individual stock-selection opportunities.
We can also look at beta to other variables. The second row in Display 11 shows the beta to volatility. A positive number indicates that a strategy adds value in a volatile setting—in other words, it’s a “long-volatility” strategy. The ability to outperform in a volatile environment is very valuable in portfolio construction because most strategies tend to underperform in volatile settings. Based on their positive betas to volatility in the post-2008 period, risk-parity and risk-managed VA strategies could provide important diversification.
On the other hand, long/short equities have had a negative volatility beta—they tend to underperform in volatile environments. They behave this way because they don’t include bonds for diversification, as the other two strategies do. As a result, they often end up in crowded trades that underperform during volatility spikes.
It’s interesting that more than half the return volatility in all the strategies is explained by the equity market and its volatility level. In the case of risk-managed VA strategies, these two influences explained an amazing 74% of returns; for long/short equities, the number was 61%. The increased impact of the market and volatility on the risk of these products may be linked to managers’ greater reliance on similar, market-driven signals, as well as fewer opportunities to take cross-sectional or security-selection risks. This may lead to simultaneous de-risking in response to volatility increases and/or market declines.
The Role of Manager Diversification
Not all managers are the same, but how much diversification is really possible from manager selection? In the case of long/short equities, the case for diversifying active returns is strong (Display 12): The correlation of beta-adjusted active returns among managers has been remarkably low for the past decade. Meanwhile, correlation in absolute returns spiked after 2008, as the average beta of the strategies increased. This would suggest that investors should pay particular attention to the aggregate beta of a long/short multi-manager portfolio and consider using an overlay program to manage market risk.
In the case of risk-managed VA strategies, the correlations of both absolute and beta-adjusted returns have increased among managers (Display 13). This may reflect the deployment of similar risk-reduction strategies and a growing similarity in target-risk mandates among providers. The pairwise correlation of almost 85% among risk-managed VA strategies’ absolute returns suggests a worrisome similarity in behavior through the business cycle.
On the other hand, risk-parity strategies tend to deploy a much longer adjustment cycle, and they don’t respond as quickly to changes in volatility. As a result, their correlations tend to be lower (Display 14). Furthermore, managers tend to incorporate different assets into the mix, including commodities, credit, rates, REITs and equities. So we see even lower correlation among risk-adjusted returns.
The Impact of Broad-Based De-Risking
The prevalence of broadly similar risk-management techniques and an increased emphasis on managing short-term losses have made it more likely that a surge in volatility and/or a market decline would lead to simultaneous selling across strategies and asset classes. This would result in systematic drawdowns across seemingly unrelated strategies. What impact could an increase in volatility have on different investment strategies?
Here’s one example. As a group, long/short equity managers tend to de-risk by selling their biggest overweight positions; they may cover their largest short positions, too. These are generally the highest-conviction positions and the biggest contributors to overall portfolio risk. Managers tend to make these moves when volatility spikes and portfolio risk rises. As a result, crowded stocks (stocks that are consensus long positions and commonly recommended by many sell-side firms) and long positions that are most broadly held by hedge funds, as captured by the GS VIP Index (Display 15), have tended to underperform during volatility spikes. We believe that underperformance by crowded trades was responsible for the poor performance by active managers and hedge funds in 2014.
On the other hand, risk-managed VA strategies tend to quickly adjust their beta exposure in response to rising implied volatility. This is necessary to accommodate hedging strategies and to keep overall portfolio volatility below 10%. However, this can result in a dramatic reduction in equity exposure in the periods surrounding spikes in volatility (Display 16). For example, if volatility were to spike as it did in 1987, 1990–1991, 1998, 2001–2002, 2008 and 2011, risk-managed VA strategies would have had to slash equity exposure by more than 30% of total assets under management. As the amount of money deployed through these strategies continues to grow, it could contribute to a short-term drawdown akin to 1987, if implied volatility spikes sharply in response to a macro event.
What about the impact of rising volatility on risk-parity funds? In general, these strategies tend to take a longer view of volatility and are less responsive to short-term shocks. Selling pressure would have to come from a more prolonged volatility increase and environment of uncertainty than would cause VA strategies to sell. Display 17, attempts to quantify the value of assets that would have to be sold if volatility rises by 30% and stays at that level. Not surprisingly, the impact would be most extreme for bonds and credit, where the degree of leverage is highest.
It’s hard to say whether attempts to liquidate $100 billion to $150 billion in equities or bonds in a volatility spike would be enough to derail markets and create a systematic sell-off. But we think that type of event becomes more likely as central banks start to raise interest rates and limit liquidity. Regulatory capitalization requirements and fortified risk-management processes could keep traditional liquidity providers from providing offsetting flows. So we think it makes sense to maintain an arsenal of risk-management strategies that evolve with market conditions and risks.
Opportunity in “Least Crowded” Trades and Buying Equity Volatility
To manage risk effectively, we think it’s important to be thoughtful in deploying these strategies. In the current market landscape, one approach that features the “least crowded” stocks looks particularly attractive.
Over the past six years, we’ve researched crowding in equity markets, including maintaining a model portfolio of least crowded stocks. Generally, these stocks are broadly underweighted by active managers, disliked by the sell side and have had multiple years of underperformance versus their regional markets. We believe that a portfolio of these least crowded, contrarian stocks may offer valuable diversification of active risk during periods of heightened volatility or market drawdown.
As we discussed earlier, these periods tend to spur managers to reduce risk. This often involves selling the biggest portfolio-risk contributors, which drains liquidity from the most crowded stocks and leads to underperformance. We’ve found that a portfolio of least crowded stocks actually would have offered attractive alpha diversification during the past decade, especially in tumultuous markets (Display 18).
We also see an opportunity for risk-management strategies in the relatively low option-implied equity volatility (Display 19). Call options are particularly attractive, with valuations possibly being depressed by the proliferation of income-generating call-overwrite strategies.
We exploit this opportunity by obtaining some of the equity exposure in our risk-management strategies from purchasing S&P 500 call options instead of owning equities directly. This strategy produces a modest negative carry (loss) if markets remain flat, but a market movement of more than 2% either way in a month would produce a profitable trade relative to the benchmark—either by reducing losses on the downside or by taking part in the upside (Display 20).
We’ve used similar logic in building energy trades (Display 21). When the price of oil, as measured by West Texas Intermediate, drops to the mid-$40 range per barrel, we sell put options at about $40 to capture the significant premium. We believe that the marginal cash cost of production, $35–$40, provides a strong floor for the oil price, and we’d be comfortable adding to our positions below those price levels. On the flip side, a surge in the price of oil makes options cheaper. We’re using put spreads to protect our positions in case there’s a further decline.
Finally, we’re executing a hybrid beta-reduction and volatility-buying strategy in one of our portfolios to address the client’s desire for protection against a significant drawdown while maintaining upside participation. We set up a collar position, buying a put and selling a call when volatility and skew are low, while selling futures to manage equity exposure when volatility is high or rising. The end result of combining options and volatility management is the ability to limit the amount of equity that has to be sold to meet volatility targets (Display 22). This positioning creates the potential for better participation in a market rebound while controlling overall volatility.
We’ve modestly reduced exposure to risk assets, while implementing some of our equity overweight using call options. This positioning will automatically reduce our risk exposure if the market declines. Likewise, we’ve used put options to buy downside protection on oil prices. Outside those changes, we maintain a modestly procyclical stance, based on the assumption of a rebound in US small-business and household formation. Globally, we see relatively few deep-value opportunities, but the Greek crisis is starting to cause valuation spreads to widen in Europe—this is where we have our largest value exposure. The key variable we’re keeping an eye on is continued credit creation, which is necessary to sustain the global economic recovery.