Why signals turn out false
Some signals never play out. That is not a bug but a property of trend indicators. Where false triggers come from and how many there really are.
What counts as a false signal
A signal is usually called false when price did not move the stated way but reversed and reached the stop level. Formally the indicator did its job: the filter line really did turn. The move simply did not follow.
A signal is not a forecast. Range Filter records a regime change that has already happened on closed candles; it does not predict the next ones. No indicator built on past prices can know whether a move will continue or stall.
Where they come from
- Ranging markets. While price drifts inside a narrow range it keeps crossing the channel boundaries in both directions. Every such exit is a formal reversal, that is, a new signal. Flat markets produce by far the most false triggers.
- A sensitive setting. We use a sampling period of 20 bars and a range multiplier of 3.5 — noticeably faster than the common 50 and 3.0 from TradingView. Sensitivity buys an earlier entry and pays for it with extra signals.
- Lower timeframes. On 5m and 15m noise makes up a larger share of the move than on 4h and 24h, so the share of empty triggers there is higher.
- News and outside events. The indicator only looks at closing prices. A report, a rate decision or a large liquidation does not exist for it — it sees the aftermath, not the cause.
How many there really are
We ran the indicator across the ten catalogue pairs on the 15m, 1h and 4h timeframes, 1000 candles per series, which produced 810 closed signals. Of those, 58.3% closed at the stop level, 41.7% reached at least TP1, and 26.4% closed positive.
So most individual triggers end in the red, and that is a normal picture for trend logic: its point is not the share of winning entries but whether the winners bring more than the losers give away. Check that ratio on your own settings and horizon rather than taking it on trust.
How to reduce their share
- 1Check the higher timeframe. Skipping signals that go against the higher trend is the cheapest filter there is.
- 2Do not take signals inside a tight range. If the band around the filter line has narrowed and price has flipped back and forth several times, the market is not trending.
- 3Move to a higher timeframe if there are too many triggers: 4h produces several times fewer than 15m.
- 4Always set a stop. A false signal with a limited loss is a working situation; a false signal without a stop is how accounts are lost.