How the Market Performs Around Big Days

Here’s a research project I’ve been working on. I was curious to see if the market has historically evinced any sort of pattern before and after 2% down days.

I took all the S&P 500 daily closings since 1960. I realized I needed to use some kind of band so I sorted of all the daily moves between -1.9% and -2.1%. There were 91 such days which is about 0.6% of the time. I then pulled out 41 days: the 20 days before the big drop, the big drop and the 20 days following the drop.

I found that yes, there is a pattern. The market rallies before the big drop, about 1.5% in 10 days, which shouldn’t be too unexpected. However, the market tops out five days before the drop. The market then slowly slides before – wham! Down 2%.

Here’s what I found surprising. After the 2% drop, the market has continued to sink. It’s a momentum trade, and after about 10 days, the market has really started to fall. In the 20 days following a 2% drop, the market has historically lost another 1.73%, and that’s on top of the 2% loss.

Now I was curious to know what happens after a +2% daily jump. I assumed it would be the mirror image of the 2% fall. It’s not. I found 93 such days. In the 20 days before the big jump, the market hasn’t done much of anything. It has flatlined, but after the 2% jump, the market has lost ground, but not by much; -0.67% in 20 days.

Lots of questions. Is the market reverting? How come a 2% move prompts a countertrend while a 2% drop continues one? I don’t know. Perhaps down moves are more trend-sensitive.

I then tried the same experiment but with 1% up days and 1% down days. Again, I used a band of 0.9% to 1.1%. Now we’re dealing a lot more data, around 500 days for each.

Like the -2% day, the market has rallied into a 1% drop but not as much. Unlike the -2% day, the market has rallied a little bit (a very little bit) after a 1% drop. The move works out to less than 3% annualized. The 1% drop basically has canceled out the moves before and after the drop.

The time around a 1% up day is good for the bulls. The market has rallied 0.60% ahead of the bump and another 0.55% afterwards. That’s about the long-term average, but combined with the 1% jump it makes it good for investors.

The emerging picture seems to be that high volatility is simply bad news for stocks. Market students know that the market’s best days usually happen in the middle of lousy markets, but even modestly strong days aren’t that great for stocks.

One final test. What about flat days? I found more than 1,900 days between -0.1% and +0.1%. Flat days come amid good times. Twenty days prior to flatness, the market has gained 0.40%. Annualized that’s 5.2%. But here’s the surprise, the market has gained 1.07% in the 20 days following a flat day. That’s more than 14% annualized. I didn’t see that coming.

I also looked at 3% up and down days (2.9% to 3.1%), but there isn’t a lot of data so I’d want to be humble about any conclusions. However, the market has done horribly after 3% days of any direction. Down 3.25% after 3% losses, and down 5% after 3% gains. Yikes.

Conclusion: There seems to be a horseshoe effect. Big up and big down days are bad. I would guess there’s a sweet spot, probably a mildly positive day of 0.2% or so – that’s the ideal environment.

For any future research, I would recommend looking at more days. Maybe 100 before and 100 after. I’d also look to see if there’s a discernible trend towards an ideal spot.

Here’s a chart of the data which I hope is self-explanatory.

Posted by on February 28th, 2019 at 8:37 am


The information in this blog post represents my own opinions and does not contain a recommendation for any particular security or investment. I or my affiliates may hold positions or other interests in securities mentioned in the Blog, please see my Disclaimer page for my full disclaimer.