Roller Coaster Statistics

SPY 20high C to 20d low C

Within 11 days of hitting a 20-day high close, today the S&P 500 hit a 20-day low close. When this sort of roller coaster up and down happens, what can we expect from the future? Well as I often say: “the future can be revealed with statistics! But only 60% of the time.”

So I dusted off the ol’ trading abacus and looked back through SPY, an ETF which stands in nicely for the S&P 500. I looked all the way back to 1993 because, hey, I care about you folks. And really, it was no trouble at all.

I didn’t want to overfit my sample selection, so I set up these simple parameters:

• The close of SPY must be lower than the previous 20 closing prices.

• Sometime in the past 15 days, a close was higher than the previous 20 closing prices.

This yielded 150 instances since 1993, which should be enough data to chew on. Here’s what happened over the next five days, on average, if you’d bought at the close of that 20-day low close (not including commissions etc). The result is a decidedly positive win rate for all five days following, and an all-positive average gain over all five days too.

67.11% 1 day win rate
61.74% 2 day win rate
65.10% 3 day win rate
60.40% 4 day win rate
62.42% 5 day win rate
+0.34% 1 day avg G/L
+0.50% 2 day avg G/L
+0.68% 3 day avg G/L
+0.78% 4 day avg G/L
+0.97% 5 day avg G/L

Conclusion? You should have bought SPY at the closing bell today. But not to worry, the rest of the week could be awesome.

Unless, of course, it’s not.

Here’s wishing you an awesome week.

Leave a Reply

Your email address will not be published.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