Does Lightning Strike Twice?

overnight-gap

OneĀ of the things that makes trading individual stocks so risky is “event risk”. Which, to put it bluntly, is when all hell breaks loose because of some news item in the off-hours, causing an overnight gap downward from the close of the previous day to the beginning of the next day.

A stop loss can’t do anything to prevent these overnight gaps. The market doesn’t care if you wanted out at $9.35. If the next day opens at $8, well then that’s what you get.

I was reading a fine blog post by Cesar Alvarez, where he discussed the use of maximum-loss stops with mean reversion systems, and how they usually hurt results. I’ve found the same thing to generally be true, and wondered in my comment whether gaps could be avoided by getting them over with, so to speak. If you incorporate a downward gap as part of your mean reversion strategy, are you able to avoid additional gaps in the future?

In other words, does lightning strike twice?

Let’s go find out!

I picked a year: 2012. Why? Because it’s convenient. Recent but not too recent. Then I selected all of the current Russell 1000 stocks that were in existence at the beginning of 2012. 912 stocks to be exact. And since they’re currently in the Russell 2000 list, they made it to the end of 2012 intact.

During 2012, there were 527 gaps downward that were greater than 5% (as defined by the previous close divided by the current day’s open).

Assuming there were 252 trading days in 2012 (if there weren’t, then my numbers are slightly off but not enough to affect the results), that gives us 229,824 possible moments for a gap.

So on any given day, there was a 0.23% chance of having a gap down of greater than 5%.

Of the stocks that had a >5% gap downward in 2012, there was a 0.63% chance of that happening on any given day (because only 322 stocks accounted for those 527 gaps). Some stocks just behaved themselves in 2012 and avoided downward gaps.

I then looked at all the gaps, and determined if there were any additional gaps subsequent to the first gap. I was surprised by the results.

There was a 1.71% chance of a >5% gap happening in the next 2 days.
There was a 2.47% chance of a >5% gap happening in the next 5 days.
There was a 3.42% chance of a >5% gap happening in the next 10 days.

Gap me once, shame on you. Gap me twice, shame on me.

Mind you, those are still pretty small percentages. We’re only talking 18 instances of a second gap happening in the following 10 days. In 2012, only one ticker had two subsequent gaps following the first one.

But the results, as preliminary as they are, do seem to indicate that downward gaps have a slight tendency to cluster.

Lightning does indeed strike twice.

And you have a 1:280,000 chance of being hit by actual lightning this year.

Leave a Reply

Your email address will not be published. Required fields are marked *