I looked at the CCI indicator in a previous post exploring technical behavior of the VIX:
Based on yesterday's RSI system improvement using a conditional MA suggested by Clueless, I decided to test the same concept on the CCI. The results are very impressive and I actually favor the CCI model in lieu of the RSI model.
Both the CCI and RSI generate approx. the same number of trades.
The RSI produces about 18% more total gain than the CCI but, and this is big . . .
The max drawdown on the CCI system is 50% of the RSI system.
The largest losing trade and the average losing trade on the CCI are 25% of the RSI system.
Max drawdown on the CCI long trades is 25% of the RSI system.
Although the CCI system generates a lower overall return, it does so with considerably less drawdown risk using no stops. The CCI system as presented clearly favors the long side trades and a continuing area of research is improvement of the short side trades. This may be accomplished by varying the entries/exits triggers based on the open/close and a number of other variables.
The goal is to keep the system as simple as possible, while generating consistent returns.
Here's the TS2000i code to use for further testing. As noted in yesterday's post, there's some variation in verbiage when transposing to TS8 format.
Anyone caring to share those nuances, please submit as comment.
TS sets the range of the CCI at -100 to 100. Other platforms using the CCI indicator may use a default range of -150 to 150 or 0 to 100. Make sure to check the embedded CCI range before testing the settings noted here or your results may be substantially different.
Also as noted yesterday, further testing on a fractal time frame is required to confirm the robustness and reliability of the system.