Sector Volatility and What We Can Learn From It

The recent post looked at sector correlations in the stock market, which were helpful in identifying the recent peak in stocks.  Above is a chart of rolling 20-day volatility across the major stock market sectors.  We can see a similar pattern of low cross-sector volatility in advance of market peaks and high volatility at market bottoms and lift-offs from those bottoms.  

Note how volatility has trended lower across the sectors, recently touching a new low at recent market highs.  We have since bounced higher in volatility, but have not broken the pattern of generally lower highs.  As long as that pattern remains intact, corrections should be relatively modest both in duration and extent.  Indeed, a study of market corrections as a function of volatility regimes is quite useful:  expectable declines in lower regimes are quite different from those in higher ones.

Of course, this pattern will change at some juncture and we'll see sustained upward movement in volatility.  We actually bottomed in VIX back in July, though realized volatility across sectors continues its downward pace.  Because volume is highly correlated with volatility, an early tell of changing volatility regimes will be a breakout in share volume.  As long as volume remains within recent parameters, I don't expect market action wildly different from the recent past.

Short-term returns in SPY have been quite sensitive to volatility.  Going back to 2012, when sector volatility has been in its uppermost quartile (highest volatility), the next five days in SPY have averaged a gain of +.95%.  When volatility has been in its lowest quartile, the next five days in SPY have averaged a gain of only +.04%.  Adjusting volatility readings for the downward drift is useful in detecting relative volatility highs and lows and then gauging the impact of volatility on forward price action.  On that basis, we are nearing, but not yet at, top quartile relative values of volatility.

Further Reading:  Intraday Relative Volume and Volatility
.