Skip to content
MT5MT4soonTradingViewsooncTradersoonNinjaTradersoonPythonsoon
Backtesting & statistics

Non-repainting indicator

Also called no-repaint indicator, non-repaint indicator, non-repainting arrow indicator, indicator that does not repaint

A non-repainting indicator never changes a signal or value on a candle that has already closed. The arrows and lines in its chart history are the same ones you would have seen live at the time.

A chart full of well-placed arrows proves nothing until you know those arrows were there at the time. If they were added, moved or deleted afterwards, any test built on that history overstates the result, and you end up sizing real risk on trades that never existed.

What stays fixed and what is allowed to move

Most indicators recalculate on the candle that is still forming, because its close is not known yet. That is normal. Repainting means something different: the indicator changes or deletes a value on a candle that has already closed.

A third behaviour sits in between. Many reversal tools need a few candles to the right of a swing high or low before they can confirm it, and then they draw the arrow back on the swing candle. That arrow never moves, but you could not have traded it where it sits.

Non-repainting, repainting and back-plotted arrows compared on the same candles
Only the top row shows on the chart what you could actually trade at the time. The forming candle on the right may flicker in all three rows, and that is normal.
BehaviourSignal on a closed candleWhat the chart history showsCan you trade it as drawn?
Non-repaintingFixed for goodWhat you saw liveYes, from the close of the signal candle
RepaintingCan move or vanishA cleaned-up versionNo
Back-plotted (delayed)Fixed once drawnArrow sits earlier than it appearedOnly from the candle where it first appeared, at that candle's price

Which common indicators repaint and which do not

You can often judge a tool from its formula before you test it. Use this rule: if the formula uses candles to the right of the signal, it repaints or back-plots.

IndicatorBehaviourWhy
Moving average, RSI, MACD read on closed candlesFixedEach value uses only prices up to and including that candle's close
FractalsBack-plotted, confirmed 2 candles laterA swing point needs 2 closed candles after it before the arrow can be drawn
ZigZagRepaints its last legThe latest swing point moves each time price makes a new extreme
Centered or TMA bandsRepaintThe centred calculation uses candles that come after the one being plotted
Any multi-timeframe versionCan repaint until the higher-timeframe candle closesEach smaller candle reads a higher-timeframe value that is still forming

What an erased arrow costs: a worked example

The chart and figures in this section are an illustrative example, not real trades.

You run a reversal arrow indicator on a 1-hour chart. A sell arrow prints on a small down candle that closes at 1.1021, and you sell. Price keeps rising to 1.1062, which is 41 pips against you. Your stop is hit on the way up and the trade closes at a loss. The indicator then deletes the first arrow and draws a new one at the high. The chart below shows both arrows.

A repainting sell arrow: where it printed live and where history shows itIllustration
Your live sell at 1.10211.1021Live arrow, later erasedArrow left in historyBack-plotted arrow appears h

Anyone opening that chart later sees one well-timed sell at the top. Your losing trade left no trace.

The same chart helps picture the back-plotted case. A tool that needs 3 candles to confirm a swing would draw its arrow on the candle with the 1.1062 high, but only after the third candle to its right has closed at 1.1025. The arrow sits at the top, while the first price you could have sold at was 37 pips lower.

Now scale it up. The history shows 20 arrows: 14 reached the target and 6 hit the stop. You risk $100 per trade to make $100. Your own live log of the same weeks holds 30 arrows, because 10 losing ones were erased afterwards. To keep the sums simple, this example counts only arrows deleted outright, with no replacement drawn.

Chart history versus live log for the same repainting arrows
Winners shown in history14 × $100= $1,400
Losers shown in history6 × $100= $600
Result the history suggests$1,400 - $600= $800
Erased losers you took live10 × $100= $1,000
Live result ($800 less $1,000)$200 loss

The history says you are up $800. Your account is down $200. The whole gap comes from signals you can no longer see.

How to check whether an indicator repaints

No platform has a "repaint" switch. You find it by watching, and the method is the same on any trading platform.

  • The indicator's inputs. Look for options such as "signal on candle close", "confirmation bars" or a higher-timeframe source. Reading a higher-timeframe candle before it has closed is a common cause of repainting.
  • Strategy tester replay. Step forward one candle at a time and watch whether older arrows shift.
  • The alert log. An alert set to fire at the candle close is time-stamped and cannot be withdrawn, so it records what the indicator said live.

The alert log gives you a check you can copy:

  1. Switch on alerts, set them to fire once per candle, at the close, and let the indicator run for a full session.
  2. The next day, count the alerts and count the arrows on the chart for the same hours.
  3. Match each alert time to an arrow on that exact candle.

If a candle-close alert has no arrow on its candle, a signal was erased. If an arrow sits on a candle earlier than its alert, the tool back-plots. An alert that fired while the candle was still forming and has no arrow only shows a flicker, not a repaint.

Matching candle-close alerts to chart arrows to spot erased, back-plotted and flickering signals
Illustrative times. Only a candle-close alert with no arrow on its candle proves a signal was erased.

Checks before you trade an arrow or signal

  • Act on closed candles only. If a signal flickers on and off while the candle forms, wait for the close and enter on the next open.
  • If the tool needs 3 candles to confirm a swing, test your entry 3 candles after the arrow, at the price that was available then.
  • If your live count and the history count differ by even one signal, discard any backtest built on that history.
  • Treat "no repaint" in a description as a claim until you see a test with counts: how many signals were checked and how many changed.
  • Repeat the check after changing settings or timeframe. Turning on a multi-timeframe filter can introduce repainting in a tool that was clean before.

FAQ

Does a non-repainting indicator give more winning trades?

No. It only makes the history honest. A fixed signal can still lose, and a strategy built on it still needs its own testing.

Does the Supertrend indicator repaint?

The standard calculation leaves the line on closed candles alone. The flip on the forming candle can appear and disappear before the close, and versions that read a higher timeframe can change candles that are already closed on your chart.

Will a backtest of a trading robot (automated strategy) expose a repainting indicator?

Usually yes. A tester feeds candles one at a time, so the robot reads each signal as it stood at that moment. That is why tester results for such tools often look much worse than the arrows on the chart. It exposes the problem only if the robot reads the indicator on each new candle as it arrives. Scrolling back through a finished chart does not count as a test.

Related terms