Coppock Curve (COPPOCK)

Summary

Coppock Curve: Edwin S. "Sedge" Coppock's long-term momentum oscillator (Barron's, originally published as the "Trendex Model"), computed as a weighted moving average of the sum of two rates of change.

Unbounded; positive turns from below zero are the signal the indicator was designed for (long-term buying opportunities on monthly index data).

Formula

COPPOCK = WMA(ROC(optInROC1Period) + ROC(optInROC2Period), optInWMAPeriod)

Each ROC carries ROC's own zero guard — a zero price optInROC*Period bars back yields 0.0 for that term, never an infinity. The two ROCs are summed, not averaged: every published definition sums them. (Tulip's copp averages, so it reads at exactly half this amplitude — a clean 2.0x ratio against Tulip is Tulip's variant, not a defect.)

The formula is symmetric in the two ROC periods and the lookback keys off their max, so optInROC1Period > optInROC2Period is accepted rather than rejected.

The classic defaults are 11/14/10 on monthly data. Wikipedia's daily-scale variant (231/294-bar ROC, 210-bar WMA) is a parameter choice reachable through this API, not a competing formula.

Inputs

• inReal — Source price/value series (canonically a monthly close)

Outputs

• outReal — Coppock Curve value

Parameters

Parameter	Type	Default	Accepted values	Description
optInWMAPeriod	integer	10	1–100000	Smoothing period for the ROC sum
optInROC1Period	integer	11	1–100000	Short rate-of-change period
optInROC2Period	integer	14	1–100000	Long rate-of-change period

Notes

• The single fused pass is bit-identical to running ROC + ROC into WMA.
• First output at max(optInROC1Period, optInROC2Period) + optInWMAPeriod - 1. Not start-dependent: each output depends only on its finite trailing window.

Properties

Numerical Stability: Start-Independent

☐ Overlap Input
✅ Independent Y-Axis i
☐ Candlestick
☐ Can Output NaN or ±Inf
☐ Identity at Period 1

Implementation

TA-Lib Definition: coppock.c · coppock.yaml

Native	File
C	ta_COPPOCK.c
Rust	coppock.rs
Java	Core_COPPOCK.java

TA-Lib is also available for Python, R and more using a wrapper.