MACD, Explained in Plain English

The three parts, in plain words

MACD stands for moving average convergence divergence, which sounds intimidating but describes something simple: how two moving averages of price pull apart and squeeze back together. It starts with two exponential moving averages, or EMAs. An EMA is just an average of recent closing prices that gives extra weight to the newest days, so it reacts faster than a plain average. MACD uses a fast one covering roughly 12 days and a slow one covering roughly 26 days.

The MACD line is the fast EMA minus the slow EMA. That is the whole formula. When the 12-day EMA sits above the 26-day EMA, recent prices have been running hotter than the longer trend, and the MACD line is positive. When recent prices cool off relative to the longer trend, the line shrinks or goes negative. The second part, the signal line, is a 9-day EMA of the MACD line itself, so it is a smoothed, slower-moving copy that trails behind. The third part, the histogram, is the gap between those two lines, drawn as bars above or below zero.

Each layer answers a different question. The MACD line asks whether short-term momentum is ahead of or behind the longer trend. The signal line asks what that momentum has averaged lately. And the histogram asks whether the gap between them is widening or narrowing, which is why people call it momentum of momentum: it does not measure the move itself, it measures whether the move is still gaining or starting to fade.

A worked example with invented numbers

Picture a stock that has been climbing and now trades near 100 dollars. Suppose the 12-day EMA sits at 102.40 and the 26-day EMA at 100.90. The MACD line is 102.40 minus 100.90, or plus 1.50. The positive reading tells you the recent stretch of closes has been strong enough to drag the fast average well above the slow one.

Now suppose the signal line, the smoothed copy, sits at 1.10 because it is still catching up to the recent surge. The histogram is 1.50 minus 1.10, a bar of plus 0.40. A positive and growing histogram says the gap between fast and slow momentum is still widening, which is what a healthy advance tends to look like in MACD terms.

Then the rally stalls and price drifts sideways for a week or two. Nothing dramatic happens to the stock, but the fast EMA slips to 101.60 while the slow EMA grinds up to 101.00. The MACD line is now plus 0.60. The signal line, being slower, has only eased to about 1.05. The histogram flips to roughly minus 0.45, and the MACD line has crossed below its signal line, even though the MACD line itself is still positive and price never had a scary down day. That is the appeal of the histogram and the crossover: they can flag fading thrust before it shows up as an obvious decline. It is also, as the next sections cover, exactly where the false alarms come from.

Crossovers and the zero line

The most watched MACD event is the signal-line crossover. When the MACD line crosses above its signal line, short-term momentum has picked up faster than its own recent average, which readers of the indicator treat as a bullish shift. When it crosses below, as in the worked example, momentum has slipped beneath its recent average, read as a bearish shift. Neither cross says anything about price targets or timing; it only says the spread between two averages changed direction recently.

The zero line is a different, slower checkpoint. The MACD line equals exactly zero when the 12-day EMA and the 26-day EMA are identical. Above zero, the fast average is above the slow one, which is one common definition of an uptrend; below zero, the reverse. A zero-line cross therefore marks the moment the two underlying averages themselves trade places, a much bigger-picture event than a wiggle in the spread.

A useful mental model is a hierarchy of speed. The histogram turns first, because it reacts to any narrowing of the gap. The signal-line crossover comes next, once the narrowing persists. The zero-line cross comes last and least often, once the whole relationship between fast and slow averages has inverted. Faster signals arrive earlier but are wrong more often; slower signals are more meaningful but show up well after the move began.

Why MACD lags, and why it whipsaws

Every ingredient in MACD is a moving average, and a moving average by construction contains old prices. The 26-day EMA still carries the influence of closes from weeks ago, so it cannot register a turn until enough new days pile up to outweigh the old ones. Stack an average on an average on a difference of averages, and the result is an indicator that confirms what already started rather than one that foresees anything. In a strong, persistent trend that lag is a tolerable price for smoothness.

Sideways markets are where the design breaks down. Imagine a stock chopping between 98 and 102 for six weeks with no direction. The two EMAs braid around each other near the middle of that range, the MACD line hovers close to zero, and tiny wobbles in price are enough to flip it back and forth across the signal line. You might see a bullish crossover at plus 0.20 reversed by a bearish one at plus 0.10 a few sessions later, then another flip the following week. Each cross looks identical, on the indicator, to the ones that precede real moves.

This is why experienced readers of MACD treat it as trend-following equipment and try not to lean on it when a chart is rangebound. The catch is that the indicator has no way to announce that conditions are choppy; the same lines and bars print either way. Deciding whether the environment suits the tool is a judgment MACD cannot make for you.

Divergence, briefly

Divergence is when price and MACD disagree. Say the stock from the earlier example pushes to a new high of 110, and MACD peaks at 1.50 on that move. Weeks later price grinds slightly higher to 112, but MACD only manages a peak of 0.90. Price made a higher high while momentum made a lower high. The second push covered more ground on the chart with less thrust behind it, the way a ball thrown upward keeps rising for a while even as it decelerates.

The honest caveat is that divergence is a condition, not a countdown. Strong trends routinely print two or three divergences in a row and keep going for weeks afterward. It is fair to read a divergence as a note that momentum is thinning; it is not fair to read it as a schedule for when anything reverses.

A complement, not a crystal ball

It helps to remember what MACD literally is: recent price history, averaged twice, subtracted, and averaged again. It contains no information that was not already in the chart, and nothing about the future. Simple systems that mechanically trade every crossover have been tested many times, and the common finding is unimpressive performance once losing whipsaws and trading costs are counted. The indicator earns its keep as context, a compact way to see whether momentum is building or fading, rather than as a standalone decision maker.

That is also why two people can act on the identical crossover and get opposite results. One risks a small, planned fraction of an account, defines an exit before entering, and lets the occasional whipsaw cost little. The other sizes the position on conviction and has no exit plan, so the same false signal does real damage. The crossover was the same; the risk decisions were not.

This is where the practical work lives. Deciding how many shares a given idea justifies, and whether the potential reward is worth the defined risk, are arithmetic questions that tools like a position size calculator and a risk-reward calculator can help you think through. MACD can describe the momentum backdrop while you do that math, but the math, not the indicator, is what determines what any single trade can cost you. That framing keeps MACD in its proper role: one input among several, describing the recent past, on charts where the future remains unwritten.