Moving averages on very short charts react to noise

[Amanhasnoname]

A 20-period moving average on a one-minute chart covers twenty minutes. On a tick chart, it might cover seconds. In either case, the line moves with every flicker, and crossovers happen constantly.

Signals from such averages produce frequent trades, many of which are noise. Each costs a spread, and the win rate rarely compensates.

Some traders smooth the output with longer periods or use averages as context rather than triggers. The line serves as a rough trend gauge, and entries come from other signals.

Different types behave differently. Exponential averages react faster and produce more whipsaws. Simple ones lag but are smoother.

Testing the average as a filter, such as only taking trades in the direction of price versus the line, shows if it adds value. Often it improves the result by removing counter-trend trades.

Counting how many crossover signals appear in a single hour on a one-minute chart is usually enough to show how noisy they are.

How are moving averages used on short charts, as triggers or filters?