This Friday, March 20, is a triple witching day. This is when stock options, index options, and index futures all expire on the same day. In practice, these days tend to bring repositioning, options rolling, and general market “cleanup”. It’s also the day the S&P 500 is rebalanced to changes made to the index.
If history is any guide, volumes are higher, volatility is a little greater and returns tend to be a little lower. Let’s review.
Volumes:
The most dependable effect of this day has been mechanical: more contracts expiring means more investors making decisions at the same time. Looking at the data, triple witching days see more than 3x the usual volume.
Volatility is a little more nuanced.
It tends to run a bit hotter on triple witching days, but not dramatically so. These days may add friction, but broader market conditions still do the heavy lifting. The median VIX level was modestly higher on triple witching days: 20.76 versus 18.94.
Returns, however, are where the data gets the most interesting.
Triple witching days in this sample period posted a S&P 500 median return of -0.36%, compared with +0.10% on regular days. The win rate was just 25%, with only 6 positive sessions out of 24, versus 55% on non-triple witching days.
That doesn’t mean Friday has to be a down day, but it does suggest that expiration-related flows have historically carried a negative bias.
For the S&P 500 rebalance, when the index changes are reflected, updating for company additions or removals, index funds and portfolios that trade close to the index positioning need to make changes for this, forcing sales and buys. The new names can see buying volume while the stocks being removed can see selling pressure.
A day like Friday compresses a lot of activity into one trading day: hedges get adjusted, positions get closed, others opened, and exposures get rolled. It’s important to be aware of since it can make price action feel more exaggerated than the underlying news might justify. The negative return skew is also notable enough to consider.