Gap Down Strategy In Stocks – Going Long
Last Updated on April 18, 2023
Gap down trading strategies in stocks are popular and perhaps for good reason.
Below is a gap down trading strategy in stocks:
- The stock must gap down. That means today’s HIGH must be lower than yesterdays LOW, and
- Today’s volume must be lower than the average over the last 50 days. This is important. A higher volume most probably means bad news and will likely lead to more selloff.
- If all the above-mentioned criteria are met, entry is on the next day if the stock falls 0.5 times Average True Range the last 50 days from the open. The reason I want to buy if the stocks fall the next day is that this strategy has no edge over the first 1-4 trading days. Actually, I have a potential twist to this strategy of going short. I’ll publish that later.
- The stocks close above the 10-day moving average, then exit at CLOSE, or
- The stock reaches a target of 7.5%. It’s better with a higher target, but a lower target locks in profits on “dead cat bounces” during bear markets. If the target is reached, the target price is exit (but if the stock gaps up higher than the target, the OPEN is exit).
- If neither 1 nor 2 are met, there is a time stop of 15 trading days at the CLOSE. The reason I have chosen 15 days is that stocks have performed well over this period. For a shorter period, it is much more erratic. In general, stocks fall a lot faster than it rises.
My sample is on the S&P 600 stocks that were in the index in January 2007.
Here is the result with 5000 USD per position:
Here is the equity curve with max 10 positions (on days with more fills positions are picked randomly) and NOT reinvested profits:
Annual returns are about 11%. 2804 trades in total, ie. a lot of potential trades are not taken.
Now, this strategy can also be traded with an entry on OPEN, ie. go long on the open in point 3 on the entry criteria and not wait for a drop from the open. The results are still very good with an annual return of 12% (higher return because of a lot more fills).
Do any readers have any inputs/suggestions?
Wow, great job! I am impressed with the speed of new strategies coming out of your blog.
I am just curious – when you say 5000 USD per position and 11% annual returns, what is your starting equity in this simulation? Also, do you have a max of 10 positions open at any given time or the potential for opening 10 new positions per day?
Hi James, yes, it’s my job so I spend many hours per day researching.
The software I’m using use 5000 per position and 50000 in equity as default. Max capacity is 10 OPEN positions at ANY time (of course, I can change that if I want).
I did some tests this morning on different stocks. This strategy works, as might be expected, a lot better in more thinly traded stocks. Very liquid stocks are not particularly good with this strategy. I tested with entry on the OPEN the day after the signal to avoid any dubious LOW quotes. The best stocks have an average volume between 100K to 1 mill.
out of curiosity i’ve back tested most of the strategies you published on this blog (hats off to you, very impressive algos) and could confirm most of your equity graphs.
however, i am struggling with this one, my implementation of the rules just gets terrible slaughtered in 2008 and 2011 😉
i wonder, did you select a sub-set of stocks from the sp600 ? or did you have a stop loss setting? i tried with yahoo data btw, could that be the reason for the difference?
thanks and congrats again on this blog!
Thanks for your nice words. I guess your parameters must be wrong? I didn’t get “slaughtered” in those bad years. My stocks were based on a selection from 2006 so no survivorship bias either.
i think one reason for the difference could be in the way i plot the profit/loss curve. i evaluate the p/l at each day, whereas it appears you evaluate the profit/loss at the moment where you close a trade. is that correct? cheers
Yes, you’re right. I only measure drawdown at close. The program does not measure drawdown your way, unfortunately. Actually, a big minus, and far from real trading.
thanks for the reply, cheers!