In our last two editions of Factor Spotlight, we explored volatility’s unprecedented post-COVID rally as well as its recent hints at renormalization. As the dust continues to settle, there are strong indications that suggest that low volatility stocks are coming back into favor while high volatility stocks might take a back seat. During this transitionary period, we have seen quite a bit of turbulence in the Volatility factor.
Both low volatility and high volatility managers have been riding a rollercoaster, feeling strong pressures from both sides on portfolio performance. This week, in order to understand the practical impact of the Volatility factor on investment strategies, we will be applying our research to a real world portfolio management use case. The goal is not only to diagnose volatility’s influence on a portfolio but to take subsequent steps in mitigating unwanted risks and enhancing alpha potential.
For example’s sake, we have constructed a US-focused Long/Short portfolio that has been consistently overexposed to the Residual Volatility factor in the Barra US Total Market Equity Model for Long-Term Investors.
Although the overexposure to Residual Volatility would have had a significantly positive impact from March 2020 through February 2021, the factor has changed course since, creating complications with respect to recent returns. When isolating the time period from March 1st to July 22nd, we see that the portfolio’s performance has been largely factor-driven while alpha has been relatively muted. Factor performance has been incredibly volatile, driving portfolio return down steeply in March followed by a recovery in June before dropping off again in July.
Looking under the factor hood, the Residual Volatility factor tells a lot of the story. As of May 13th, the factor had driven the portfolio down to the tune of 599 bps and ultimately landed at 435 bps of underperformance through July 22nd.
Taming the Volatility Factor
Managers who might be feeling nauseous from volatility’s recent ups and downs are looking for ways to mitigate dramatic factor swings. A fundamental manager, in particular, who takes no directional stance on volatility and is not in the business of factor timing, wants to dampen the authority the factor has over his or her portfolio and, in turn, magnify the role of idiosyncratic alpha.
Below is a list of the ETFs with the highest exposures to the Residual Volatility factor as of July 22nd. ETFs are a popular method for quick, easy, and targeted hedges. What we start to see, however, is that, volatility comes with accompanying exposures and characteristics. Many of the top ETFs are focused industry and sector funds such as XBI, XOP, and IBB. Managers may not want to take such a heavily-focused bet against these sectors and industries on their own.
ARKK presents its own set of issues as well. The ARK Innovation ETF is heavily concentrated, holds a significant portion of shares outstanding in its underlying companies, and is expensive to short. As a result, many investors are looking for alternatives.
A Re-Introduction to the JP Morgan iDex Indices
In order to target the volatility risk of our demo portfolio, we’ll look to the iDex Indices developed by our partners at JP Morgan. The JP Morgan iDex US Volatility Indices are constructed from a US broad-market index. After applying various investibility screens, the top and bottom 150 securities by trailing 6M-1W volatility are selected and rebalanced monthly to form the constituents of the JP Morgan iDex US High 6m Volatility and US Low 6m Volatility indices, respectively.
Two characteristics that make these indices unique are the chosen weighting scheme and the sector normalization. The indices are liquidity-weighted and capped at 2% concentration which is an interesting alternative to the standard cap- or equal-weighting that we typically see with market indices. Sector exposures are capped at 25% to reduce the heavy sector bias that can be at times be evident within factor indices and the ETFs we observed above. These methods provide proper diversification while preserving factor isolation.
To apply our hedge, we will start by simulating a pair trade of both the JPM iDex 6m High Volatility and Low Volatility indices, allowing us to short high volatility names and go long low volatility names. The pair trade creates a combined hedge driven almost entirely by style factors. Over 90% of the predicted risk of the hedge is driven by style factors in the Barra US Total Market Risk Model for Long-Term Investors, well over half of which is made up by Residual Volatility and Beta.
By shorting the JPM iDex High Volatility index and applying a long position in the JPM iDex Low Volatility index, we were able to reduce Total Risk by 178 bps while also reducing factor contribution to Total Risk by almost 5%, much of which was driven by the Residual Volatility factor.
To get a better sense of the effectiveness of this hedge in practical terms, we applied the same trade, rebalancing monthly back to the beginning of March and compared the performance of our new portfolio to the original. Below, we can see relative performance of the Demo US Long/Short Portfolio with the iDex index hedge vs. the same portfolio without the hedge. From March 1st to July 22nd, the hedged portfolio showed 149 bps of outperformance. Residual Volatility and Beta accounted for 124 bps and 38 bps, respectively, which greatly reduced the previous 435 bp drawdown we saw in the original portfolio.
Volatility will continue to be a factor to watch. Given its recent turbulence, even small exposures to broad volatility and residual volatility factors can have massive implications for portfolio risk and performance. We will be digging for more insights at Omega Point to provide further transparency. In the meantime, if you are interested in a complimentary portfolio evaluation or would like to discuss our research in greater detail, don’t hesitate to reach out!
US & Global Market Summary
US Market: 07/19/21 - 07/23/21
- The major US indices started the week with a significant sell-off on Monday as concerns around the spread of the Delta variant shook the market.
- This gave way to a strong rally over the next four days, with the S&P 500 ending up +1.96% and Nasdaq up +2.84% on the week (as both reached new all time highs). The Dow crossed above 35,000 for the first time as well.
- The bond market made some waves on Monday as well, when yield on 10Y Treasuries fell to a five-month low of 1.13%. Yield bounced back to 1.29% on Friday, helping to assuage investor concerns over the economy.
- 2Q earnings season is in full swing, as we’ll see a flurry of big reports next week including Apple, Amazon, Alphabet, Microsoft, Facebook, and Tesla.
- Investors will also spend part of next week digesting the 2Q GDP print and picking apart statements coming out of the Fed’s meetings on Tues and Wed.
- IHS Markit reported that its flash U.S. Composite PMI Output Index, which tracks the manufacturing and services sectors, fell to a four-month low of 59.7 from 63.7 in June.
Normalized Factor Returns: Axioma US Equity Risk Model (AXUS4-MH)
Methodology for normalized factor returns
- Earnings Yield was the week’s biggest winner, moving up another half a standard deviation as it crossed into positive normalized space after hitting a 6/24 trough of -1.86 SD below the mean.
- Profitability continued to climb higher into positive normalized territory, now sitting at +0.59 SD above the mean.
- Value is slowly climbing back from a low of -2.22 SD below the mean on 7/8, remaining an Oversold factor.
- Growth increased by +0.13 standard deviations and is now an Extremely Overbought factor at +2.12 SD above the mean.
- Market Sensitivity continued to decline after exiting Overbought space last week, falling 0.34 standard deviations back towards the mean after a recent peak of +1.7 SD above the mean on 7/1.
- As discussed above, Volatility was again the worst performing factor as it fell another 0.37 standard deviations to +0.49 SD above the mean.
- US Total Risk (using the Russell 3000 as proxy) decreased by 7 basis points.
Normalized Factor Returns: Axioma Worldwide Equity Risk Model (AXWW4-MH)
Methodology for normalized factor returns
- Global Earnings Yield vaulted into Overbought space, now sitting at +1.14 SD above the mean.
- Profitability ditched its Oversold label as it rallied by +0.36 standard deviations towards the mean. It now sits at -0.94 SD below the mean.
- Global Growth climbed higher into Overbought space as it moved up to +1.55 SD above the mean.
- Value was flat on the week, and remains an Extremely Oversold factor at -2.19 SD below the mean.
- Exchange Rate Sensitivity had a significant down move, falling 0.31 standard deviations to -0.79 SD below the mean.
- Global Volatility continued its free fall, crossing into negative normalized space , falling out of Overbought space with a -0.54 standard deviation move, landing at -0.8 SD above the mean.
- Global Total Risk (using the ACWI as proxy) decreased by 21 basis points.