The Relationship Between Volatility, Correlation & Portfolio Protection

As academic research has evolved from the single-factor approach of the capital asset pricing model (CAPM) in the 1960s to more sophisticated, multiple-factor models today, market participants who can separate noise from predictive signals have been able to generate excess returns versus the market. Over the last couple of years, I have written about asset correlations and predictive signals for spikes in volatility. In this article, I will take a step further, combining both into a single analysis that seeks to demonstrate a link between correlations and both dispersions and forward volatility. Most importantly, I’ll also identify specific strategies that could provide protection during periods of rising uncertainty.

Correlation Curves

In a recent post, I introduced the idea of correlation curves. To summarize, this approach plots a term structure of average correlations in the equity market in the same way traders think about the term structure of interest rates. That is, each point of the curve is the average of pairwise correlations between all stocks in an equity market across time on the x-axis.

In markets lacking systematic risk, the only difference between the calculation of near-term correlations (i.e., one or two months) and longer tenors (i.e., 24 or 36 months) is simply the addition of more historical data. Thus, average correlation numbers for different windows should not be statistically different. By examining this data, it’s now possible to track changes in the shape of the curve and assess how systematic risk in the market is evolving.

The table below illustrates the three most recent economic crises and how the correlation term-structure went into what the futures world refers to as “backwardation,” i.e., near-term correlation values are higher than far-term correlation values. Over longer periods, such as 36 months, correlations tend to converge toward the non-crisis average of 0.35, as represented by the last row of the table. In short, increasing correlation is a harbinger of increasing volatility.

Average Trailing Correlations for S&P 500 Index Stocks During Economic Crises

Wagging Tail of Correlations: Going Up or Down?

In the chart below left, I plot the correlation curve as of Q1, Q2 and Q3 2018. After the massive spike in volatility on Feb 5, the curve gradually started to flatten (i.e., near-term correlations were declining faster than medium and longer-term correlations). As the tail of the correlation curve went down over the following six months, the S&P 500 rallied by more than 11%.

In the chart to the right, I conduct a similar analysis, plotting the correlation curve every 10 days in October 2018. As correlations rose over the period, markets tanked by almost 7%.

For definitions of terms in the chart, please visit our glossary.

In my view, the wagging of the correlation tail is a signal for market behavior. If the tail goes down, the market goes up (and vice versa).

Accelerating Correlations: A Canary in a Coal Mine?

Taking this analysis a step further, it appears that average correlations could have a predictive power to forecast forward volatility.

In the chart below left, I show average six-month trailing pairwise stock correlations across all equities in the S&P 500 Index plotted against average VIX levels in the following month on the y-axis. In the chart on the right, I plot the ranges of the VIX for a given correlation bucket.

Our takeaways:

1. Generally, higher correlations across equities led to higher VIX levels the following month. Thus, rising correlations acted as a signal of higher VIX in the future.

2. Higher correlations in equities also tend to lead to a big dispersion in VIX ranges over the following month.

To summarize, accelerating correlations not only act as a precursor for spikes in the VIX, but also indicate potentially unstable swings in values of the VIX.

How Should Investors Allocate in this Environment?

To answer this key question, let’s look at few high-level trends:

1. After nearly a decade of supportive monetary policy that kept volatility constrained, we are entering a period in which interest rates are rising.

2. As markets adjust to a reduction in global central bank balance sheets, systematic risk has the potential to drive correlations higher, which in turn could increase volatility.

3. Given current trade tensions between the U.S. and China, market volatility appears to be increasing as investors grapple with uncertainty.

In our view, none of these challenges are insurmountable for equity markets if earnings continue to grow. However, an alternative method of stock selection and volatility mitigation may become necessary. As pioneers of Modern Alpha™ strategies, WisdomTree has a host of strategies that can help investors navigate the current environment:

1. WisdomTree U.S. Multifactor Fund (USMF): This strategy aims to provide multifactor exposure to U.S. equities through a combination of fundamental (value and quality) and technical (momentum and correlations) factors. USMF aims to deliver quality stock selection that can withstand periods of heightened systematic risk.

2. WisdomTree CBOE S&P 500 PutWrite Strategy Fund (PUTW)/Russell 2000 PutWrite Strategy Fund (RPUT): Both strategies aim to provide a positive correlation to equities, but with significantly lower volatility. These strategies sell one-month at-the-money put options, thereby generating income which could partially offset losses during market corrections. Should markets rise, the put options eventually expire, worthless, allowing investors to generate positive returns. PUTW and RPUT generate income by selling volatility and thus can help lower drawdowns during market corrections.

3. WisdomTree Dynamic Long/Short U.S. Equity Fund (DYLS) and WisdomTree Dynamic Bearish U.S. Equity Fund (DYB): These are long/short equity strategies, which seek to add value through security selection, as well as opportunistically hedging market risk, when fundamentals are mixed and volatility is increasing. DYB has the potential to go net short the market to profit from declines in equity prices.

Conclusion

Trends in correlation can be a powerful predictor of future volatility and risk in equity markets. As investors continue to grapple with short-term uncertainty, we believe our correlation signal can provide valuable insights into broader market trends. At present, we believe a more defensive positioning could be warranted as the current bull market enters its latter stages.