Any news on this one? I just saw it now. To be clear - is the pattern that with 2 gap days in a row with the prior day gap being unfilled, today's gap will fill...? If so, I just did a test on this for the ES and it wasn't very promising. I excluded gaps of 2,50 points and less. 2 Gap Up in a row - 58 % chance the gap on day 2 will fill. 2 Gap Down in a row - 60 % chance the gap on day 2 will fill. I included 2,00 points leeway for a gap fill. I did this very fast just now, so I stand corrected if I have done any errors. Will look more into this.
Just a word of advice. These aren’t the type of patterns you can use in a robust model. I’ve beat these to death years ago. Narrow your testing down to time-series momentum/trend and reversion analysis. In my years of research I’ve found momentum outperforms reversion, though I still analyze reversion for comparison data. @nooby_mcnoob posted this a while back. I think this emphasizes the idea very well.
Your test is correct, lately the rule hasn't been working with a great probability. I would say in the last 6 months or so, its predictive ability got very questionable...
That's how it seems to be in the market. Patterns come and go. I just checked the 3 Day Gap Rule on up gaps where you have 3 gap up in a row. I had 22 days in my sample. Using 2,5 point leeway, I had 64 % of those gaps fill on the same day. I bet if one used an additional filter one could arrive at more interesting results as looking at gaps in isolation might be futile...
For instance, I might require the market to be below it's 20 MA. For the 2nd Day Gap Rule, this yields 24 days in my sample. Using 2,5 point leeway we have a fill of 71% or 17/24! An improvement indeed.
Can you attached a chart of what you mean as a gap for example in ES, and how does it close. For example, if I am assuming market opens with a gap up meaning it opens higher, and then does again the 2nd day, you expect the 3rd day for it to go down? Or does it go down the 2nd day?
The rule is mostly for ETFs and using the 4pm cash close, not the future's value. The future is good only for guessing before the open how big of a gap we are getting, but you have to use the 4 pm value, not the future's close value. So instead of ES use the 9:30 - 4:00 SPX chart. For example on the 16th (Thursday) the SPX opened up like 12 points and never looked back during the day. In fact that gap is still open. The next day there was no gap, so no rule. Now the following trading day (Monday) there was another upgap about 3-4 points and it promptly came back and closed before returned to continue the rally. The rule calls for consecutive days and this wasn't that, but if I am a bit flexible, we can stretch it. The more days pass between the gaps, the less the probability of the rule working. Last Friday we had another upgap of 5 SPX points and that one also didn't close during the day. So tomorrow if we have an upgap (looking at the futures we are expecting a 4-5 points one) that would call for a closing of that upgap according to the rule. We shall see... IIRC a few months back we had 3 consecutive upgaps unfilled on the same day, what was unheared of a couple of years back. I guess it is just showing the strength of the bullmarket...
FWIW, I tested this on ES futures using the 09:30-16:00 US Equities RTH session. In my experience, prior day closes are very significant in the ES market.
How to trade gaps? talk about trading index commodity currency futures. Asian market - it gaps when there is something major happening during recent US session. gap could be rather huge. Then simply trade US session to improve odd of success. Eur market eg Eurex products - it hardly gap even though it has long closing hours. Even if there is gap, the gap is usually quite insignificant. for other products like CME products, there wouldn't be any gap at all. US market - there wouldn't be gap as there is no break between eur and US session