VIX data goes back to 1990, so we have five presidential elections to weigh the behavior of uncertainty and therefore VIX levels. What could the election mean to the performance of volatility shortly afterward?
In two of the five elections, the market was in a major downward correction. In 2008, the S&P was en route from 1400 to 666. VIX levels were extreme as the market gyrated by hundreds of points per day. In 2000, the internet bubble was coming to an end, and VIX levels were also quite high as the S&P was making a 400 point correction. Since we are not in a market similar to that, I am going to exclude those years as outliers.
The years 1992, 1996 and 2004 are the few, but enlightening data points. The average VIX move by Friday of election week was -13%. By midweek the following week, the average VIX move was -20%, the starting point being the Monday of election week, essentially 10 days.
Could there be extra anxiety this year due to the perceived closeness of the race? In 1992 and 1996, the VIX went from 17 to 13 and 18 to 13.5 respectively. These races were not close come election day as Clinton's electoral votes exceeded 370, far higher than the 270 needed to claim the White House.
In 2004, the race was much closer, the electoral tally being 286-251. The VIX went from 16.76 to 12.77. That was a drop of 15% in the election week and over 23% during the two week period.
As the VIX crossed 18.50 today, it may appear ready to deflate shortly after the election.
Additional disclosure: I am long VXX puts.