Repainting indicators at the tick level
Posted: Sat Oct 31, 2026 9:43 pm
Repainting indicators at the tick level
Some indicators change their past values as new ticks arrive. On a closed bar the signal looks perfect. On the live bar it flickers, appears, and disappears.
Traders see a signal on a historical chart and assume it would have been tradeable. In reality the signal only appeared after the fact, or it appeared and vanished within the bar.
Scalpers feel this more than anyone because they trade on short bars where the live candle is always partially formed. A signal that looks clear at the close may have shown the opposite a few seconds earlier.
The test is simple. Watch the indicator in real time for an hour and record when signals appear and whether they ever disappear. Any that change after appearing should be treated as unreliable.
Zigzag, some fractal tools, and a number of pivot-based indicators are well known for it. Plenty of free scripts that promise perfect entries rely on the same trick.
One common version is the indicator that uses future bars to place its signals, then redraws the arrows afterwards. Looking at the code or the description often reveals this. If an indicator promises to mark exact tops and bottoms, it almost certainly looks ahead.
A solid test is saving screenshots of live signals in real time for a week and comparing them with the same chart afterwards. Any signal that changed position or disappeared tells the story. Many popular free tools fail this test immediately.
Even an honest indicator can mislead on a forming bar. If the signal depends on the current candle's close, it is only final when the candle closes. Acting before that is acting on a draft.
Another quiet source of repainting is a data feed that revises recent bars after the fact. Anything built on those bars will shift slightly, which is another reason to judge an indicator only on what it showed in real time.
Has an indicator ever looked brilliant on history and failed in live trading because of repainting?
Some indicators change their past values as new ticks arrive. On a closed bar the signal looks perfect. On the live bar it flickers, appears, and disappears.
Traders see a signal on a historical chart and assume it would have been tradeable. In reality the signal only appeared after the fact, or it appeared and vanished within the bar.
Scalpers feel this more than anyone because they trade on short bars where the live candle is always partially formed. A signal that looks clear at the close may have shown the opposite a few seconds earlier.
The test is simple. Watch the indicator in real time for an hour and record when signals appear and whether they ever disappear. Any that change after appearing should be treated as unreliable.
Zigzag, some fractal tools, and a number of pivot-based indicators are well known for it. Plenty of free scripts that promise perfect entries rely on the same trick.
One common version is the indicator that uses future bars to place its signals, then redraws the arrows afterwards. Looking at the code or the description often reveals this. If an indicator promises to mark exact tops and bottoms, it almost certainly looks ahead.
A solid test is saving screenshots of live signals in real time for a week and comparing them with the same chart afterwards. Any signal that changed position or disappeared tells the story. Many popular free tools fail this test immediately.
Even an honest indicator can mislead on a forming bar. If the signal depends on the current candle's close, it is only final when the candle closes. Acting before that is acting on a draft.
Another quiet source of repainting is a data feed that revises recent bars after the fact. Anything built on those bars will shift slightly, which is another reason to judge an indicator only on what it showed in real time.
Has an indicator ever looked brilliant on history and failed in live trading because of repainting?