When Correlations Rise…

Given the recent market shock, especially to those active in “volatility” markets, we thought it would be an opportune time to quickly re-visit a concept we wrote about two months ago. In that post we examined an often-overlooked cause of volatility (low or high), correlation. We showed that a large driver of low volatility in 2017 was historically low intra-market correlation. The math of volatility is quite simple; as the correlation of assets in a portfolio decrease, the ratio of average asset volatility to portfolio volatility increases. The opposite is also true, as correlations approach 1.0 the ratio of average underlying asset volatility and overall portfolio volatility will also approach 1.0, providing little to no diversification benefits.

There has been much discussion on why the VIX increased so much, so quickly, on what was a relatively small drop in the S&P 500 Index on February 5th. Volatility markets are complex, and given the implosion of several volatility focused funds this week, it’s clear that many involved in the markets may not have appreciated that complexity. However, much ink has already been, and will continue to be, spilled on those stories. Instead, we thought it would be interesting to view this week’s volatility through the lens of correlation.

Implied volatility represents 10 day put implied volatility of at the money options
(Bloomberg call: put_imp_vol_10d)

As can be seen above, S&P 500 Index implied volatility almost tripled on Monday while the average constituent volatility increased by only 33%. Index implied volatility and average constituent volatility were essentially the same, inferring that all stocks within the index were expected to be perfectly correlated, a practical impossibility, but still painful for anyone who happened to be short volatility. As would be expected, this extreme condition reverted rather rapidly, but not before much damage was inflicted to investor portfolios. While painful, this outcome should have been well within expectations. As we stated in a previous post: “Should correlations increase, market volatility could rise dramatically (not even considering an increase in underlying intra-market volatility). This could lead to challenging times for many portfolios.”

Read our last blog post, Trend Following in 2017: A Post-Mortem? >