What “non-repainting” really means — and how to check any indicator yourself

“Non-repainting” is the most abused word in indicator marketing. Almost everyone claims it; far fewer mean it. Here's what it actually means, why it decides whether a backtest is worth anything, and how to check it yourself on any script in a few minutes.

What repainting actually is

An indicator repaints when it changes what it already showed you. A signal appears, you act on it, and later — as new bars arrive — that signal moves, weakens, or quietly disappears from the chart's history. Scroll back a week and everything looks perfect, because the tool has rewritten the past to match what actually happened.

The problem is simple: a repainting indicator makes decisions using information that wasn't available at the time it pretends to have decided. On a historical chart that's invisible. In live trading it's the gap between the flawless backtest you were sold and the very different experience in your account.

Why it matters more than almost anything else

Every performance claim rests on this. If an indicator repaints, its backtest is fiction — the arrows sit exactly where hindsight put them, so the win rate is guaranteed to look great and guaranteed to be unrepeatable. This is why we treat non-repainting as non-negotiable, and why it's one of the five questions worth asking before you buy any tool (see how to spot a curve-fit indicator).

The flavors worth knowing

“Repainting” covers a few different behaviors. You don't need to be an engineer to recognize them:

• Intrabar repainting: a signal flickers on and off while the current candle is still forming, and only “settles” when the bar closes. If you act mid-bar, you're acting on a signal that may vanish.
• Historical vs live mismatch: the signal logic behaves one way on closed historical bars and another way in real time — so the backtest and live trading are literally not the same strategy.
• Higher-timeframe lookahead: the indicator pulls in higher-timeframe data that wasn't actually finalized yet at that point in history — a subtle peek into the future that inflates results.

How to check any indicator yourself

You don't have to trust the marketing. On TradingView you can test it directly:

• The bar-replay test. Open Bar Replay, step forward one candle at a time, and watch the current bar. If a signal appears mid-bar and is still there after the bar closes — and doesn't move afterwards — that's a good sign. If it appears and later disappears, it repaints.
• The refresh test. Note exactly where the last few signals sit, then reload the chart. If any signal shifts to a different bar or price after the reload, the history was being rewritten.
• The live-vs-history test. Watch a signal form in real time, screenshot it, and compare it to the same spot a day later. It should be identical.

A tool that's genuinely non-repainting has nothing to hide from any of these. A tool that dodges the question usually has a reason.

Non-repainting isn't a nice-to-have or a marketing bullet — it's the precondition for every other number meaning anything. Verify it first. Everything else comes after.