EquityMar 27 2017

S&P 500 and Vix's relationship reveals interesting insights

twitter-iconfacebook-iconlinkedin-iconmail-iconprint-icon
Search supported by
comment-speech

The CBOE Volatility Index (Vix) – the widely regarded market sentiment indicator and forward-looking estimator of S&P 500 volatility – remains low at 11.3 points, at the time of writing, and has hovered between 10 and 14 for most of 2017.

The only notable spikes in these depressed levels are the political risk events of Brexit and the US presidential election last year. It would not be absurd, therefore, to assume that markets will remain bullish on the much-anticipated reflation trade, and that investor ‘animal spirits’ have largely returned as equity markets continue their upward trend.

The combination of populism concerns in the Dutch general election with the rise of Geert Wilders’ PVV party and the much-anticipated rising of the US federal funds rate for the third time since the global financial crisis barely registered as a blip on the Vix. Thus when FE was asked last week whether volatility was expected to remain depressed, or if selling volatility was wise to enhance portfolio returns, we thought a closer consideration into the relationship that underpins the Vix and S&P 500 volatility might provide some insight. 

The Vix index is calculated using option prices on the S&P 500 across 30 days, thereby deriving implied volatility. Looking at the relationship between the implied volatility of the Vix against the realised standard deviations of the S&P 500, it seems abundantly clear that the S&P 500 volatility tends to lag the Vix. Any rise or fall is often followed by a similar move in S&P 500 volatility, lending weight to consensus that it is a forward-looking estimator of market fluctuations.

Nevertheless, a simple regression analysis of S&P 500 monthly volatility versus the Vix showed a statistically significant positive coefficient, with an r-squared of 42.7 per cent. This falls in line with expectations, even though the low level suggests a lack of accuracy in the Vix predicting market volatility.

Another observation is that implied volatility is typically higher than subsequent realised volatility. But this has not been the case recently as risk premium (the difference between Vix and S&P 500 volatility) is negative, with the Vix being much lower than S&P volatility.

The correlation between the realised S&P 500 volatility and Vix has ranged between 1 and -0.7 since 2000. Correlations moved to extreme negatives whenever the Vix over/underestimated the observed volatility. There is evidence of this in January 2003 (Vix rising, S&P volatility falling), June 2008 (Vix decreasing, S&P volatility increasing) and October 2013 (like 2003) where they travel in opposite directions.

FE’s results tend to agree with previous papers that the Vix explains the variation in S&P 500 volatility somewhat but is not a perfect relationship. A key takeaway from the negative rolling correlations and the examples highlighted is that the Vix regularly over and underestimates market volatility in different periods (stress/calm/benign). 

With no guarantees in the implementation of positive fiscal Trump policies, and fears of a rise in protectionism, we remain concerned of left tail risks that are currently being ignored by a complacent market.

Not much has changed in the fundamental drivers of growth, bar oil-led inflation. We remain focused on seeking managers who are consistent in their approach as we aim to benefit from the diversification in their differing portfolio tilts and strategies. 

Come what may, over a full cycle, this positioning should ride out any major corrections and stay on track within risk limits. 

Charles Younes is research manager at FE