A big topic of interest has been the recent spate of volatility, including the "volatility of volatility".  There are any number of ways to monitor that, but what we do on the site is compare the VIX gauge of implied volatility to its 10-day moving average.

Generally, when the VIX is more than 10% below its 10-day average, then the stock market is overbought; when it's more than 10% above the average, it's oversold.

Over the past week, the VIX has gone from more than 30% above its average to more than 15% below.  That gives us precious few historical comparisons, but when we relax the parameters to 15% both ways, we get a few more.

The table below shows how the S&P 500 performed in the weeks and months following each instance.

The results were weak across the board, with extreme volatility in the results.  During the next week, the S&P was +/- 3% on 7 out of the 8 occurrences.

During the next three months, the median drawdown in the S&P 500 was -15.5%, compared to a median maximum gain of only +5.5%, so certainly a negative skew there.

This is in direct conflict with a couple of the studies we've looked at over the past couple of days.  How do we reconcile that?  It just depends on how much weight you want to give each one, but personally I think the others outweigh the negative connotations from this single study.

If we look at how the VIX itself performed, something curious shows up:

In the very short-term, the VIX tended to rise, which makes sense, since it tends to rise when stocks fall.  But three months later, the VIX was negative every time, even despite the poor performance in the S&P.