What Happens After a MACD Crossover?
15,698 bullish MACD signal-line crossovers across 197 US large-cap stocks, 2018 to 2026, measured over the following 35 trading days against the same measurement taken from every trading day in those stocks.
The Test Bench › MACD studies
We tested the MACD signal-line crossover across US large-cap stocks in a series of event studies. Each one measures what happened in the 35 trading days after a bullish crossover and compares it with ordinary trading days in the same stocks. The series works through three questions in order: what follows a MACD crossover at all, whether the comparison it is measured against is fair, and whether the sharpness of the crossover itself carries extra information.
15,698 bullish MACD signal-line crossovers across 197 US large-cap stocks, 2018 to 2026, measured over the following 35 trading days against the same measurement taken from every trading day in those stocks.
Study 002 compared MACD crossovers with every trading day. A reader asked if that was fair, so we re-ran the test only against days with similar trend and volatility, with a calendar-block bootstrap interval on each crossover-minus-comparison difference.
Not every MACD crossover crosses by the same amount. We measured the size of the one-day histogram move, scaled it by the share price, and tested whether sharper crossovers were followed by larger 35-day returns — with one threshold fixed in advance and applied once to data it had not seen.
Every study in the series shares the same core design:
What changed between them is the comparison. Study 002 measured crossovers against every trading day in the same stocks. A reader asked whether that was a fair benchmark, so Study 003 re-ran the test against days carrying the same trend and volatility labels. Study 004 keeps that comparison and asks a further question of the crossovers themselves.
Each study reports a median outcome with an uncertainty range around it, and says which of its findings were explored in the data and which were tested with a rule fixed beforehand. The two carry different weight and the studies keep them apart.
Not on its own, in the sample tested. Study 002 found the median return after a bullish crossover close to the median after an ordinary trading day, and Study 003 found the same once trend and volatility were held level.
The MACD histogram is the distance between the MACD line and its signal line, and it passes through zero at a bullish crossover. Sharpness is that one-day move through zero, divided by the share price.
The histogram is measured in the same units as the price, so a fixed threshold would mostly sort stocks by how expensive they are. Dividing by the closing price puts every stock on the same relative scale.
A crossover is single when no other crossover in the same stock falls within the measurement window either side of it, and clustered when at least one does.
Most of the analysis is exploratory and measured on the same data that suggested it. Where a study runs a holdout test, it fixes the rule on one part of the period and applies it once to another, and labels that result separately.
These are historical analyses. They describe what happened in the sample tested and do not predict future returns or recommend buying or selling anything.
Rebuild any of these
Every study here was run with the same event-study tool the application ships. Build a MACD crossover event, set the horizon, add conditions and compare what follows against ordinary trading days.
Create a free account and run your own event studies → or see every Test Bench study →