Range Filtered Trend Signals indicator

Category: Indicators By: Iván González Created: July 15, 2025, 9:25 AM
July 15, 2025, 9:25 AM
Indicators
10 Comments

1. Introduction

The Range Filtered Trend Signals indicator (by AlgoAlpha) is a composite tool designed to help traders identify both tentative “in-range” price movements and decisive trend changes. By combining three powerful elements

  1. Kalman filter to smooth price noise,
  2. SuperTrend algorithm to gauge directional bias on the filtered line, and
  3. volatility-based range filter to spot mild breakouts—
    this indicator offers clear visual cues for entries and exits on any market or timeframe.

Why combine these components?

  • Noise reduction: The Kalman filter adaptively smooths price data, preserving real movements while suppressing random fluctuations.
  • Trend bias: SuperTrend applied to the smoothed line defines when the market is trending up or down.
  • Range filtering: Dynamic bands based on recent volatility highlight genuine momentum moves without overreacting to micro-noise.

Together, they produce two distinct signal types:

  • “In-range” breakouts (dots when price crosses a volatility band but trend has not yet shifted), and
  • Trend-change alerts (colored fills when the filtered line flips relative to the SuperTrend).

2. Input Parameters

Parameter Default Description Recommended Range
kalmanAlpha 0.01 Measurement noise coefficient; smaller values yield smoother filtering but slower response. 0.005 – 0.05
kalmanBeta 0.10 Process noise coefficient; controls adaptability of the error covariance update. 0.05 – 0.2
kalmanPeriod 77 Period factor in the Kalman update; higher values emphasize longer-term smoothing. 50 – 150
dev 1.20 Multiplier for the range bands; determines band width relative to recent volatility. 1.0 – 2.0
supertrendFactor 0.70 ATR multiplier for SuperTrend; larger values produce wider SuperTrend bands (less frequent flips). 0.5 – 3.0
supertrendAtrPeriod 7 ATR look-back for SuperTrend; controls sensitivity to true range fluctuations. 5 – 14

Tip: Start with the default values, then tweak one group at a time (Kalman vs. SuperTrend vs. range factor) to match your preferred balance of responsiveness vs. noise suppression.

3. Internal Logic

3.1 Kalman Filter Smoothing

The Kalman filter produces a smoothed price line K by iteratively updating:

  1. Estimate update:
    – v3 = kalmanAlpha * kalmanPeriod
    – gain = v2 / (v2 + v3)
    – v1 = v1_prev + gain * (close – v1_prev)
  2. Error covariance update:
    – v2 = (1 – gain) * v2 + kalmanBeta / kalmanPeriod

Here, v1 is the estimated true price, and v2 is its error variance. Smaller kalmanAlpha/larger kalmanPeriod yield heavier smoothing; a higher kalmanBeta adapts variance faster when volatility changes.

3.2 SuperTrend on the Filtered Line

Using the smoothed series K as the “price,” a standard SuperTrend computation defines two bands:

  • UpperBand = K + supertrendFactor × ATR(supertrendAtrPeriod)
  • LowerBand = K – supertrendFactor × ATR(supertrendAtrPeriod)

Logic ensures each band only moves in a favorable direction (Upper never decreases unless broken, Lower never increases unless broken). The SuperTrend line itself alternates between the latest Upper and Lower band depending on whether K is above or below the prior band—yielding a clean up/down bias on the Kalman-filtered price.

3.3 Volatility-Based Range Filter

A Weighted Average of the daily range (high – low) over 200 bars defines vola—a smoothed measure of recent volatility. Two dynamic range bands are then set:

  • UpperRange = K + dev × vola
  • LowerRange = K – dev × vola

These bands capture typical short-term swings around the Kalman line.

3.4 Signal Generation

The indicator produces two types of visual signals:

  1. In-Range Breakouts:
    • When the current close crosses above UpperRange or below LowerRangeand the SuperTrend line has not just flipped (isupertrend = isupertrend[1]), a small dot is plotted at the breakout point. This flags a potential pullback or continuation within the prevailing trend.
  2. Trend-Change Alerts:
    • When the SuperTrend line flips (i.e., its value differs from the prior bar) and K has crossed through the old band, the chart background (or the Kalman line) is filled in green for an up-flip or red for a down-flip. This highlights a more decisive shift in market direction.

4. Interpretation of Signals

  1. In-Range Breakouts (Dots):
    • Purpose: Catch smaller, intra-trend opportunities without waiting for a full trend flip.
    • How to trade: Look for price retests of the breakout level or close-of-bar confirmation before entering in the direction of the predominant trend.
  2. Trend-Change Alerts (Fills):
    • Purpose: Identify robust reversals when noise has been removed and volatility bands confirm momentum beyond typical ranges.
    • How to trade: Consider these as primary entry/exit points—enter on the next bar after the fill appears, or wait for a retest of the SuperTrend line for a more conservative approach.

Example Scenario: On an hourly chart of EUR/USD, price may cross the upper range band (dot signal) three times, indicating small pullbacks within an uptrend. Eventually, the Kalman-filtered line pierces the SuperTrend band and the background turns green—a high-conviction signal to establish a new long position.

5. Configuration and Advanced Adjustments

  • Adapting to Timeframes:
    • Shorter timeframes (e.g., 5 min) often require a smaller kalmanPeriod (e.g., 50) and slightly higher dev (e.g., 1.5) to avoid false dots.
    • Longer charts (daily/weekly) benefit from a larger kalmanPeriod (100–150) and modest supertrendFactor (1.5–2.0) for stable swings.
  • Optimizing Kalman vs. SuperTrend:
    • Increase kalmanBeta if the filter is too sluggish to adapt after sudden volatility spikes.
    • Raise supertrendAtrPeriod if you want the trend filter to ignore small whipsaws.
  • Volatility Regime Tuning:
    • In trending, low-volatility markets, reduce dev toward 1.0 to tighten bands and capture earlier in-range moves.
    • In choppy/high-volatility periods, raise dev toward 2.0 to avoid over-signaling.
  • Combining with Other Tools:
    • Use higher-timeframe Simple Moving Averages or RSI to confirm the direction flagged by this indicator.
    • Overlay support/resistance zones to validate Swing entries on dot signals.

6. Full ProBuilder Code

//-----------------------------------------------------//
//PRC_Range Filtered Trend Signals | indicator
//16.06.2025
//Iván González @ www.prorealcode.com
//Sharing ProRealTime knowledge
//translated from tradingview code
//-----------------------------------------------------//
// --- Parámetros de Entrada ---
//-----------------------------------------------------//
// Grupo Filtro Kalman
kalmanAlpha = 0.01
kalmanBeta = 0.1
kalmanPeriod = 77
dev = 1.2
// Grupo Supertrend
supertrendFactor = 0.7
supertrendAtrPeriod = 7
//-----------------------------------------------------//
// --- Lógica del Filtro Kalman ---
//-----------------------------------------------------//
ONCE v2 = 1.0
IF BarIndex <= 1 THEN
   v1 = close[1]
else
   v3 = kalmanAlpha * kalmanPeriod
   v5 = v1
   v4 = v2 / (v2 + v3)
   v1 = v5 + v4 * (close - v5)
   v2 = (1 - v4) * v2 + kalmanBeta / kalmanPeriod
ENDIF
k = v1 // Línea del Filtro Kalman
//-----------------------------------------------------//
// --- Lógica del Supertrend ---
//-----------------------------------------------------//
src=k
atr=averagetruerange[supertrendAtrPeriod]

upperband=src+supertrendFactor*atr
lowerband=src-supertrendFactor*atr

if barindex<=supertrendAtrPeriod then
   
   upperband=src
   lowerband=src
   direction=1
   isupertrend=src
   
else
   
   if lowerband>lowerband[1] or k[1]<lowerband[1] then
      lowerband=lowerband
   else
      lowerband=lowerband[1]
   endif
   
   if upperband<upperband[1] or k[1]>upperband[1] then
      upperband=upperband
   else
      upperband=upperband[1]
   endif
   
   if isupertrend[1]=upperband[1] then
      if k>upperband then
         direction=-1
      else
         direction=1
      endif
   else
      if k<lowerband then
         direction=1
      else
         direction=-1
      endif
   endif
   
   if direction=-1 then
      isupertrend=lowerband
   else
      isupertrend=upperband
   endif
   
endif
//-----------------------------------------------------//
// --- Lógica Principal del Indicador ---
//-----------------------------------------------------//
vola = WeightedAverage[200](high - low)

// Bandas del Filtro de Rango
upper = k + vola * dev
lower = k - vola * dev

midbody = (close + open) / 2
//-----------------------------------------------------//
// --- Visualización y Dibujo ---
//-----------------------------------------------------//
rk=124
gk=124
bk=124
ak=255

if close crosses over upper and isupertrend=isupertrend[1] then
   r=124
   g=124
   b=124
   a1=255
   a2=255
   inrange=1
elsif close crosses under lower and isupertrend=isupertrend[1] then
   r=124
   g=124
   b=124
   a1=255
   a2=255
   inrange=1
elsif isupertrend<>isupertrend[1] and k<isupertrend then
   r=255
   g=0
   b=0
   a1=255
   a2=0
   inrange=0
elsif isupertrend<>isupertrend[1] and k>isupertrend then
   r=0
   g=255
   b=0
   a1=0
   a2=255
   inrange=0
endif
if inrange=1 then
   ak=0
endif
//-----------------------------------------------------//
colorbetween(k,midbody,r,g,b,ak*0.1)
colorbetween(lower,upper,124,124,124,30*inrange)
//-----------------------------------------------------//
return k as "K" style(line,3)coloured(rk,gk,bk,ak), upper as "Upper" coloured(r,g,b,a1)style(point,2), lower as "Lower" coloured(r,g,b,a2)style(point,2)

Download
Filename: PRC_Range-Filtered-Trend-Sign.itf
Downloads: 109
Iván González Master
I usually let my code do the talking, which explains why my bio is as empty as a newly created file. Bio to be initialized...
Author’s Profile

Comments

Logo Logo
Loading...