MACD is an acronym for Moving Average Convergence Divergence. It is a mathematical indicator used by some financial traders to predict future price movements of stocks, commodities and other financial instruments. MACD was originally developed by Gerald Appel.

The MACD is constructed from two EMAs, or exponential moving averages, derived from the historical price movement of the asset being studied. In a traditional average, all data is treated equally. When calculating an EMA, some data is given a higher weight, or importance, than other data. In the case of MACD, the more recent the data, the more importance it is given. An EMA of six, for example, averages the last six values while giving more weight to the most recent ones.

MACD subtracts the longer average from the shorter average, and the result is plotted on a chart or graph. MACD typically uses 26- and 12-day EMAs, meaning it looks at the last 26 and 12 days worth of data. The resultant chart oscillates around zero, without preset bounds in either upper or lower directions. In addition to the difference between the averages, MACD values are themselves averaged to produce a center, or trigger line.

MACD attempts to measure both price trend and momentum, where momentum can be thought of as the strength of the trend. If MACD is greater than zero, it means the short-term average is higher than the long-term average, suggesting the financial instrument is trending upwards. Similarly, if MACD is less than zero it suggests the instrument is trending downwards. The steeper the slope of the MACD plot, the more violently the price is moving and, therefore, the stronger the momentum.

Bullish and bearish signals can be generated by crossovers or divergences. A divergence occurs when the MACD indicates a move in one direction while the asset price is moving in the other. A crossover is simply the MACD moving over or under the zero (neutral) point, or alternately, crossing its own trigger line. A MACD crossing from positive to negative would be a bearish signal, while a cross from negative to positive is a bullish signal.

Because MACD uses backward-looking data, it is by definition a lagging indicator. By using shorter averages, the MACD lag can be reduced, but never eliminated. Conversely, lengthening the averages used in the calculation dampens the movements of MACD and increases the lag. While MACD is most often used with data on a daily frequency, it can also be used with weekly, monthly or even yearly data.