Here are the recent Bear Catcher Combined signals as taken from GTR1 using GTR1 Helper 4.5. The complete set of signals may be found by running GTR1 with the BCC url: GTR1 Backtester
Launch the url, take the option for Detailed Report and check the box for Signal Values. Finally, click “Run Backtest” and then download the report.
For the NHNL signal, the look back period is 253 days. Before I try looking at different look backs, I am wondering if anyone else has checked in NHNL works better with a shorter look back?
I’ve wondered the same. Is that an option on GTR1? I wouldn’t think so, as it would require a lot of computation to build that data. But I’d be very interested in the results of adjusting the lookback parameter.
I’ve also wondered about the efficacy of using NHNL on the Russell 2000. I don’t have data for that universe, but thought that it might break down prior to the larger-cap indices.
NH/NL lookback period (2013) found “a minor peak at 150 and a major at 240, and a fairly orderly degradation in either direction from the peak. The standard 52 week lookback ends up being nearly the best to use.” But that was almost 10 years ago, and so an update would be interesting.
Barchart.com has many shorter NHNL time frames on many indexes. For example there is the 3-Month Hi-Lo Nasdaq ($M3DQ) 3-Month Hi-Lo Nasdaq Price - Barchart.com . Unlike stockcharts you can even use a weighted moving average, but there is no backtesting.
If you use advanced search when looking for a symbol, you can search “1-month hi-lo”, “6-month hi-lo” or even “5-day hi-lo” and then chart with different types of moving averages. There have many different indexes and sectors. Very interesting data but without backtesting I’m not sure what to do with it. I couldn’t find a useful strategy from it but perhaps others can.
I looked at the period 20131118 to present for ^GSPC and 19260104 to present for ^S5T. A lookback of about 240 days is in the area of peak CAGR. In both cases, the CAGR is much less using NH/NL but MDD is significantly better.