Three Ways to Modify CCI Calculation and Trading Signals
Summary
The article explores three changes to the Commodity Channel Index. First, it replaces the usual division by standard deviation with multiplication, then sets positive and negative signal levels separately from the observed frequency of indicator values. Second, it removes standard-deviation normalization and compares several averaging windows, including rectangular, triangular, flat-top, linear, power, and exponential weights. Third, it adds a reference price or average to the oscillator so a CCI-derived line can be displayed on the main price chart.
The article reports EURUSD hourly backtests over 2021 for the modified indicators, using entries based on crossings of adaptive levels and closing opposing positions. Results vary: the classical calculation is profitable in its comparison, while the other first-step variants lose; among the window tests, most are profitable and the exponential version loses. These are single-period, single-instrument tests, and the author notes that trigger levels and exit rules need further tuning. The reported results do not establish robustness or live performance.
Key ideas
- Replacing CCI's division by standard deviation with multiplication makes large deviations and large dispersion reinforce each other.
- Positive and negative indicator levels can be estimated separately from the empirical frequency of their values.
- The article compares CCI variants using several symmetric and asymmetric averaging windows.
- Adding a reference price or average lets a CCI-derived signal appear on the main chart.
- The reported EURUSD backtests are limited to one instrument and one test period, with exit and threshold tuning still open.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.