A moving average crossover buys when a faster moving average crosses above a slower one and sells when it crosses below. The 50 and 200 day averages produce the golden cross and death cross of the financial press; shorter pairs are used on intraday charts. It is the first strategy most traders learn.
The claim is that the crossover identifies the start of a trend early enough to ride it.
We tested moving average crossovers across many pairs of periods and timeframes on 17 years of data with real trading costs, and found no edge. The crossovers arrive late in every move, so the strategy buys after the rise and sells after the fall, and in ranging markets it is whipsawed repeatedly. The occasional long trend it catches does not pay for the losses in between.
The reason is that averages lag by construction, and the lag that filters out noise also filters out the beginning of the move the strategy is trying to catch.
If you trade it anyway, know the win rate is low, the profitable periods are trends, and the losing periods are everything else. Count the whipsaws.