Breadth-filtered commodity channel index entry and exit rules

Why a lone oscillator is incomplete

The commodity channel index is a leading overbought and oversold oscillator of price versus its typical range. In this procedure it is only the timing trigger inside a broader rule set. In a strong advance it can remain overbought, and in a decline it can remain oversold, so a lone extreme is not a complete entry or exit command.

On the illustrated Nasdaq 100 tracking series, commodity channel index extremes clustered near plus or minus 150 and the oscillator spent little time in the mid-range.

Market breadth names the regime first

Market breadth is a market-wide reading of participation. Here it is the balance of new highs and new lows, used to label a bullish or bearish regime. The high-low ratio is new highs divided by the sum of new highs and new lows, often smoothed, and is the bullish or bearish breadth reading in these rules.

A three-day simple moving average of that high-low ratio stayed near 0.7 to 0.9 during a strong advance and near 0.1 to 0.3 during a strong decline. That pattern supported using the ratio as a bullish versus bearish market-breadth regime reading.

A breadth-only mechanical trading system entered and exited from the momentum of that high-low ratio, taken as the difference of two exponential averages versus a signal line. It also used separate long and short ratio thresholds, a one-session entry delay, and same-session exits.

Combined entry rules

The combined long rule required an oversold commodity channel index reading plus either a high-low ratio above a bullish threshold or positive ratio momentum. Shorts were reserved for overbought oscillator conditions in a bearish regime.

After testing an overbought oscillator alert paired with negative ratio momentum, the short-entry rule was reduced to the overbought commodity channel index alert. Adding the momentum filter did not change the equity path.

Coded exits stay incomplete

A mechanical trading system is a fully specified procedure that states when to enter, exit, or abstain from coded conditions instead of discretionary judgment. Fully coded exits were still treated as incomplete. In a bullish regime the procedure wanted to wait until ratio momentum turned negative, but how far negative remained a judgment, and same-session discretionary exits were described as simpler than a mechanical close.

The coded long exit required the high-low ratio below an upper band, negative ratio momentum, and a still-slightly-overbought commodity channel index. The coded short exit required a bullish ratio reading plus an oversold oscillator.

The mechanical short-close rule failed to release a short during a strong late-1999 advance because market breadth stayed too bullish for the exit conditions to fire.

Range, trend, and the practical split

The write-up states that one mechanical trading system is unlikely to serve both trending and sideways markets equally. Protective stops plus choosing a procedure suited to range or trend conditions is the practical split.

Editorial reading

Editorial: the archive is taught here as a two-layer mechanical procedure, not as a claim that either layer is sufficient on its own. Market breadth names the regime first. The commodity channel index is allowed to time only the entries and exits that fit that regime. Standing aside is written as an explicit rule rather than left to the oscillator alone.