Non-Repaint TradingView Indicators: What It Means and How to Test Them

In short
What non-repaint TradingView indicators really promise, why the label is a claim and not a guarantee, and three ways to test any script for yourself.
A non-repaint indicator is one whose past signals stay exactly where they first appeared. The arrow printed on Tuesday's candle is still on Tuesday's candle a week later, so the history you are reading is the history you would have traded. That is the whole promise. The important part: "non-repaint" is a claim typed into a description by whoever published the script, not a certificate anyone checked - and you can verify it yourself in about ten minutes.
What the label is actually promising
It promises agreement between past and present. If an indicator repaints, the backtest you read and the signals you would have received are two different things, and the difference always flatters the backtest. A non-repainting script gives up something in exchange: it cannot commit to a signal until the candle is finished, so it is always one candle late. That lateness is not a flaw to be fixed. It is the honest price of a signal that will still be there tomorrow.
Three ways to test the claim
- Screenshot, reload, compare. Capture the last ten signals on a bitcoin daily chart, refresh the page, and look again. Anything that shifted or disappeared has just answered the question.
- Walk it through bar replay. Rewind a few weeks and step forward candle by candle. A clean script shows each signal at the moment the history says it appeared.
- Watch one live candle. Keep an eye on an ether chart through a full candle. If the signal flickers on and off while the candle forms and settles differently at the close, it repaints intrabar.
What it still cannot give you
Honesty is not accuracy. A non-repainting moving-average cross is perfectly truthful about when it fired and will still be wrong plenty of the time. It also says nothing about costs: fees and slippage apply to every signal you act on, and on a thin token your own order moves the price in a way no chart shows, while bitcoin and ether on a large exchange absorb it quietly.
Test before you trust, and treat any script promising instant signals with no lag as repainting until it proves otherwise. The mechanics of why scripts repaint are in what repainting in TradingView is, and the wider set of ways a test can mislead you is in common backtesting mistakes.