As I wrote about here and here, there is a simple 3-day pattern that I like a lot. I liked a lot again the last several days.
In nutshell:
Day 1) Closing price is below the low of the previous day
Day 2) The next day involves no action
Day 3) If the high today takes out the high for Day 2, then buy long and sell at the first profitable close (a stop-loss order is recommended but where to place it is an entirely different topic and one that might require a little experimentation on your own).
If the high for Day 3 DOES NOT take out the high for Day 2 then the pattern is invalidated.
Figure 1 displays all of the Day 3’s on which there would have been an entry signal in the eMini S&P 500 futures contract.
Figure 1 – Jay’s Simple Three-Day Pattern in EMini S&P 500
Last two signals:
#1
Day 1) On 7/24 ES closed below the low on 7/23
Day 2) 7/27 serves as Day 2
Day 3) ES opens on 7/28 above Day 2 high at 2071.25
ES closes on 7/28 at 2087.25
(2087.25 – 2071.25) * 50 = +$812.50 per contract
#2
Day 1) On 7/27 ES closed below the low on 7/24
Day 2) 7/28 serves as Day 2
Day 3) On 7/29 ES trades above Day 2 high of 2089.00. Buy at 2089.25 sell on close at 2101.50.
(2101.50 – 2089.25) * 50 = +$625 per contract
Summary
Like every other trading method I know of, it sure is great.
Especially when it works…..
Jay Kaeppel
Buy long, an sell at the first profitable close, theoretically sounds good by the way. Anyway, interesting details here about the strategy!
In the original article you presented hypothetical results going back to 2007. For that test did you have intraday data or end of day data? If it is end of day, how did you determine the entry and exit prices? I assume you made the entry on day 3 the same as the high on day 2 , or the day 3 open if it is already greater than the day 2 high. Was the exit then based on day 4 or greater? If so, that would give different results than entering and exiting on day 3 as in the examples in this article.
Thanks,
Kevin
Hello Jay
I backtested this with SPY all the way back to 2/9/93. I assumed you would not be able to catch day 1 of the trade and started looking for a profit at the close of every day starting day 2. No stop,I simply decided to end the trade after 5 days (which is all I could probably endure). There were 687 trades–446 winners and 241 loser—65% winning percentage but eyeballing the drawdowns on the losers they look pretty brutal.
Holding 7 days improves winning % to 72.