The Coppock curve is a long-term momentum indicator used in technical analysis to identify the start of major advances in a stock index. It is calculated from monthly closing prices as a weighted moving average of the sum of two rates of change, and it is plotted as a line that swings above and below zero. The classic signal is an upturn in the line while it is below zero.
The indicator was devised by the American economist Edwin Coppock and published in Barron's in 1962. It was designed for long-term investors in broad indices such as the Dow Jones Industrial Average and the S&P 500, not for short-term trading, and it was intended to give buying signals only.
Origin of the indicator
Coppock is said to have developed the measure at the request of a church body that wanted a simple guide for long-term investment. According to the account usually given, he reasoned that a bear market resembles a bereavement and that investors need time to recover before they are willing to buy again. He asked clergy how long people normally take to get over a loss and was told 11 to 14 months, and he used those two figures as the look-back periods of his calculation. The story explains the unusual parameters, which were not chosen by statistical testing.
Formula
The calculation uses the rate of change of the index over two periods:
The curve is then:
- is the latest monthly close of the index and is the close months earlier.
- and are the percentage changes over 14 and 11 months.
- is a ten-month weighted moving average, in which the latest month has a weight of 10, the month before a weight of 9, and so on down to 1, with the total divided by 55.
Worked example
Suppose an index closes the month at 4,400. Fourteen months earlier it closed at 4,000 and eleven months earlier at 4,200. The 14-month rate of change is and the 11-month rate of change is . Their sum is 14.76.
The same sum is worked out for each of the nine earlier months, and the ten sums are averaged with the declining weights. The latest month therefore counts for , about 18 percent of the result, and the oldest for , under 2 percent. Because of the long look-back periods and the smoothing, about two years of monthly data are needed before the first value can be plotted.
How the curve is read
The original rule is simple. When the curve is below zero and turns upward, a long-term buying opportunity is indicated. The reading rests on the idea that momentum stops deteriorating before prices have recovered much, so an upturn from a negative level marks the early part of a new bull market. Coppock gave no selling rule, on the view that market tops form slowly and differ too much from one another.
Later users have added their own conventions. Some treat a downturn from above zero as a warning or an exit signal. Others apply the calculation to weekly or daily data, to single shares or to commodities, often with shorter periods. These are adaptations, and the evidence usually quoted for the indicator relates to the monthly version applied to a broad stock market index. In form the curve is an oscillator with no upper or lower bound, so its level cannot be compared across markets with different volatility.
Limitations
The curve is built from past prices and is smoothed twice, so it belongs with the lagging indicators. A signal arrives some months after the low, and part of the recovery has already happened by then. In a long decline the line can turn up during a temporary rally and then fall again, as happened to several momentum measures in the bear markets of the early 2000s.
Signals are also rare. A monthly indicator on one index may produce a handful of readings in a decade, which makes any backtesting statistically weak and exposes adjusted versions to overfitting. The 11- and 14-month periods have no theoretical basis beyond the story of their origin. Like other tools of market timing, the curve is generally compared with a simple policy of buy and hold, which it does not reliably beat once missed gains and transaction costs are counted. It is best understood as a description of long-term momentum, not as a forecast.