3 Day Low in ETF’s

Last Updated on November 17, 2020 by Oddmund Groette

Larry Connors has written a strategy called the double 7. This is a very simple strategy with just two rules. The fewer rules, the better, because less probability of curve fitting. This strategy is based on 7 days low (and high for exit), but in my testing, it seems to work very well on all time frames (in SPY and S&P 500). It takes advantage of the mean-reverting tendencies in the stock market.

I decided to test this myself but by changing the entry parameters somewhat to better fit my trading style:

1. The ETF must close on a 3 day low (low is lower than LOW the previous 3 days).

2. The (c-l)/(h-l) must be lower than 0.33 (IBS).

3. Entry is on the close.

The exit is also changed somewhat:

1. Exit is on close which is higher than the HIGH the previous 2 days.

2. OR a time stop of 6 days

3. Exit is on the close.

I changed the rules to have a lesser drawdown. Besides, this strategy does not pay off to hold for a long time. Either it turns around quite quickly, or you risk sitting unnecessary long and tie up capital. I also have a shorter time frame on exits to have lesser drawdown (but also lesser profits).

Then I did a scan to pick tradeable ETF’s. I simply use IBS to find the best ETF’s with the best mean-reverting tendencies: go long if IBS is below 0.33 and go short if IBS is above 0.8. Entry on close and exit the next days close. Very simple. I ranked the ETF’s and want to trade the 37 best ones going forward. I also included 3 bond ETFs just to have other instruments than stock ETFs to end up with 40. Is this curve fitting? Of course, an element of curve fitting is in there. But this is the way I have always done it and it seems to work. This strategy will certainly not work in a lot of ETF’s like USO, DBC and UUP for example. They are not mean-reverting. I also excluded all “ultra” (geared) and short ETF’s.

This is the equity curve for the whole strategy:

I have allocated one part of my portfolio for this strategy. It will maximum have 3 positions at any time. that means I have to pick ETFs randomly on days there are many potential fills.