Repainting, and how to check whether an indicator does it

A repainting indicator changes its own history. Its chart looks excellent and could not have been traded.

Repainting is when an indicator's past values change after the fact. A mark appears on a bar, and later — once more bars have printed — it moves, changes, or disappears entirely.

The result is a chart of decisions that were never actually available at the time. It will look far better than any tradeable version of the same idea, which is precisely why this matters when you are evaluating anything.

How it happens

Intentionally, by design. Some indicators legitimately require future bars. A swing-high marker that needs three bars either side cannot confirm a high until three bars later. A centred moving average is shifted backwards by half its period by definition. These are not dishonest — but the mark is placed on a bar it could not have been placed on live.

Through recalculation on live data. An indicator computing on every tick may display one thing mid-bar and settle somewhere else at the close. Perfectly normal, and it means what you saw at 10:42 is not what the chart shows afterwards.

Through higher-timeframe references. An indicator on a 5-minute chart pulling values from the hourly chart will show the hourly bar's final value across all twelve 5-minute bars in history — including the ones that printed before the hourly bar closed.

Deliberately, to sell something. It happens. A backtest that quietly uses future information produces results nothing real can match.

Three checks

1. Bar replay

Use your platform's replay function and step forward one bar at a time. Watch whether marks appear where they eventually sit, or whether they shuffle as new bars arrive. This is the definitive test and it takes about fifteen minutes.

2. Screenshot and reload

Capture the chart during the session. Reload the same chart the next day and compare. Any difference in the historical marks is repainting.

3. Read the logic

Ask what the indicator needs in order to confirm a mark. If the answer involves anything after the marked bar — a subsequent high, a close n bars later, a completed higher-timeframe bar — it cannot have marked that bar in real time.

Not every case is disqualifying

An indicator that repaints intrabar and settles at the close is usable if you act on closed bars. An indicator whose history rewrites itself days later is not usable for anything. The distinction is whether the final value was knowable at the moment you would have acted.