Repainting vs non-repainting: how to actually tell

“Non-repainting” has become a selling point, slapped on indicators the way “artisan” gets slapped on bread. Everyone claims it. Far fewer tools actually deliver it, and almost nobody checks. The claim is easy to make and, it turns out, easy to verify, so there’s no excuse for taking it on faith. Here’s how to find out for yourself.

Don’t read the code, watch the behaviour

You can try to audit an indicator by reading its source, and if you’re fluent that’s useful. But most people can’t, and even if you can, it’s easy to miss a subtle leak. The behavioural test needs no code-reading at all and it catches everything, because it tests the only thing that matters: do the past signals stay put.

The principle is simple. A non-repainting indicator commits to its signals when a bar closes and never touches them again. A repainting one goes back and edits finished bars. So all you have to do is record where the signals are, wait, and check whether they moved.

The forward test

The cleanest version takes a little patience. Put the indicator on a live chart. As bars close, write down or screenshot exactly where each signal appears, on which candle, at what price. Let real time pass, a session or a day. Then look again at those same now-old bars.

If every signal is exactly where it was when the bar closed, the indicator is honest. If any signal has shifted, vanished, or appeared on a bar that had nothing on it before, it repaints, and the historical chart you’re looking at is not what you’d have lived through. That’s the whole test. Reality versus memory, and you’re checking whether the memory got edited.

The faster version

If you don’t want to wait, there’s a quicker proxy. Load the indicator and let the current bar develop without refreshing. Watch whether signals on the live, forming bar appear and then disappear before the bar closes. That mid-bar flicker by itself is the harmless kind, it just means “wait for the close.” The thing to really watch is what happens to the bar right after it closes. Does the signal lock in and stay, or does it adjust a moment later once the next bar starts feeding the calculation? If closed bars keep changing, that’s the bad kind.

You can also reload the chart and compare the recent closed bars before and after the reload. A reload forces the indicator to recompute history from scratch. If the recomputed history disagrees with what was on screen a second ago, it’s building its past with information it only has in hindsight. Honest tools reload identically. Repainting ones don’t.

Why bother

Because the difference is everything when money’s involved. A repainting indicator shows you a past full of perfect entries you’d never actually have taken, which means any confidence you’ve built from looking at its history is confidence in a thing that didn’t happen. Five minutes of watching whether the old signals stay put tells you whether you’re holding a real tool or a flattering illusion. The claim on the label costs nothing to make. The test costs you one session of attention, and it’s the only thing that settles it.

The useful version

Non-repainting is not a badge. It is a timing choice.

A non-repainting signal usually waits until the evidence is complete. That can mean a later entry, a worse price, or fewer signals. A repainting or developing signal can be earlier, but it must be treated as provisional.

The live test

Watch one setup form in real time and write down:

• when the marker first appears;
• whether it disappears before close;
• what remains after the bar closes;
• what remains after reload.

That gives you the truth. Screenshots of old charts do not.

The right question is not “which is better?” It is “which timing contract does my trading plan require?” Scalping may accept provisional information. A backtested swing strategy usually cannot.