OPEN-SOURCE SCRIPT
Updated Adaptive Friction Filter (AFF) [QuantAlgo]
🟢 Overview
The Adaptive Friction Filter (AFF) identifies trending market conditions by applying a physics-inspired friction model to price movement. Rather than smoothing price through fixed averaging, it introduces a dynamic noise threshold derived from recent market volatility, which means price must generate enough force to overcome this threshold before the filter moves at all. Once breached, the filter closes the gap at a configurable rate, producing a step-like trend line that holds steady through noise and responds decisively to genuine directional moves. This allows traders to distinguish between meaningful trend continuation and low-conviction chop across any instrument or timeframe.
🟢 How It Works
The AFF's core methodology is built around a two-stage mechanism: a volatility-derived friction threshold that gates filter movement, and a catch-up scalar that governs how much of the gap the filter closes on each bar once that threshold is exceeded.
First, the friction threshold is computed as the simple moving average of absolute bar-to-bar price changes over the configured lookback window, scaled by the friction coefficient. This makes the threshold inherently self-adjusting; it widens during volatile conditions and contracts during quiet ones, without requiring any manual recalibration:
Pine Script®
Next, the raw displacement between current price and the filter's last position is evaluated as force. The filter only advances if this force exceeds the friction threshold. When it does, the filter moves toward price by a fraction of the gap governed by the catch-up scalar, rather than closing the full distance immediately, producing a controlled and progressive response:
Pine Script®
Trend direction is then resolved by comparing the current filter value to its prior bar value. The direction state persists when the filter is flat, so no transition is registered on bars where the filter does not move:
Pine Script®
Finally, the filter is rendered as two overlapping plots at the same value: a step-line that traces the filter's path and a circle overlay positioned at each bar's filter value. The circles serve a visual purpose, reinforcing the current filter level at each step and making it easier to read the filter's position at a glance, particularly during flat periods where the step-line alone can be harder to track. Together they produce a dotted step appearance that improves legibility across different chart zoom levels and timeframes.
🟢 Signal Interpretation
▶ Bullish Trend (AFF Line Rising with Bullish Colour): When price generates enough upward force to exceed the friction threshold, the filter begins stepping higher and the line shifts to the bullish colour. The step-line rendering makes the transition visually clear; flat segments indicate bars where force was insufficient to move the filter, while upward steps reflect bars where it was. The bullish trend state persists until force in the downward direction is large enough to push the filter lower, at which point trend direction flips and the line shifts to the bearish colour.
▶ Bearish Trend (AFF Line Declining with Bearish Colour): When price generates enough downward force to exceed the friction threshold, the filter begins stepping lower and shifts to the bearish colour. As with the bullish state, the filter holds its last value on bars where force is insufficient to breach the threshold, and the direction state remains unchanged on those bars. A full reversal back to bullish requires upward force to exceed the friction threshold and push the filter higher, at which point trend direction flips and the colour transitions accordingly.
🟢 Features
▶ Preconfigured Presets: Three parameter sets cover a range of trading styles and timeframes. "Default" delivers balanced noise filtering for swing trading on 4-hour and daily charts. "Fast Response" lowers the friction threshold and accelerates the catch-up rate for intraday and scalping use on 5-minute to 1-hour charts, producing earlier filter movement in response to smaller price displacements. "Smooth Trend" raises the threshold and slows the catch-up rate for position trading on daily and weekly charts, requiring larger price displacements relative to the average noise level before the filter advances.
▶ Built-in Alerts: Three alert conditions support automated monitoring of trend transitions. "Bullish Trend Signal" fires on the first bar trend direction flips from bearish to bullish. "Bearish Trend Signal" fires on the first bar trend direction flips from bullish to bearish. "Any Trend Change" triggers on either transition for traders who want a single unified alert regardless of direction. All alerts include the exchange, ticker, and timeframe in the message for immediate context.
▶ Visual Customisation: Six colour presets, Classic, Aqua, Cosmic, Cyber, Neon, and Custom, provide coordinated bullish and bearish colour pairings suited to different chart themes and personal preferences. Selecting Custom exposes independent colour pickers for full manual control over both states. Optional bar colouring tints price candles with the active trend colour using a configurable transparency level, and optional background colouring extends the trend state tint across the full chart pane at a separately configurable transparency.
The Adaptive Friction Filter (AFF) identifies trending market conditions by applying a physics-inspired friction model to price movement. Rather than smoothing price through fixed averaging, it introduces a dynamic noise threshold derived from recent market volatility, which means price must generate enough force to overcome this threshold before the filter moves at all. Once breached, the filter closes the gap at a configurable rate, producing a step-like trend line that holds steady through noise and responds decisively to genuine directional moves. This allows traders to distinguish between meaningful trend continuation and low-conviction chop across any instrument or timeframe.
🟢 How It Works
The AFF's core methodology is built around a two-stage mechanism: a volatility-derived friction threshold that gates filter movement, and a catch-up scalar that governs how much of the gap the filter closes on each bar once that threshold is exceeded.
First, the friction threshold is computed as the simple moving average of absolute bar-to-bar price changes over the configured lookback window, scaled by the friction coefficient. This makes the threshold inherently self-adjusting; it widens during volatile conditions and contracts during quiet ones, without requiring any manual recalibration:
Next, the raw displacement between current price and the filter's last position is evaluated as force. The filter only advances if this force exceeds the friction threshold. When it does, the filter moves toward price by a fraction of the gap governed by the catch-up scalar, rather than closing the full distance immediately, producing a controlled and progressive response:
Trend direction is then resolved by comparing the current filter value to its prior bar value. The direction state persists when the filter is flat, so no transition is registered on bars where the filter does not move:
Finally, the filter is rendered as two overlapping plots at the same value: a step-line that traces the filter's path and a circle overlay positioned at each bar's filter value. The circles serve a visual purpose, reinforcing the current filter level at each step and making it easier to read the filter's position at a glance, particularly during flat periods where the step-line alone can be harder to track. Together they produce a dotted step appearance that improves legibility across different chart zoom levels and timeframes.
🟢 Signal Interpretation
▶ Bullish Trend (AFF Line Rising with Bullish Colour): When price generates enough upward force to exceed the friction threshold, the filter begins stepping higher and the line shifts to the bullish colour. The step-line rendering makes the transition visually clear; flat segments indicate bars where force was insufficient to move the filter, while upward steps reflect bars where it was. The bullish trend state persists until force in the downward direction is large enough to push the filter lower, at which point trend direction flips and the line shifts to the bearish colour.
▶ Bearish Trend (AFF Line Declining with Bearish Colour): When price generates enough downward force to exceed the friction threshold, the filter begins stepping lower and shifts to the bearish colour. As with the bullish state, the filter holds its last value on bars where force is insufficient to breach the threshold, and the direction state remains unchanged on those bars. A full reversal back to bullish requires upward force to exceed the friction threshold and push the filter higher, at which point trend direction flips and the colour transitions accordingly.
🟢 Features
▶ Preconfigured Presets: Three parameter sets cover a range of trading styles and timeframes. "Default" delivers balanced noise filtering for swing trading on 4-hour and daily charts. "Fast Response" lowers the friction threshold and accelerates the catch-up rate for intraday and scalping use on 5-minute to 1-hour charts, producing earlier filter movement in response to smaller price displacements. "Smooth Trend" raises the threshold and slows the catch-up rate for position trading on daily and weekly charts, requiring larger price displacements relative to the average noise level before the filter advances.
▶ Built-in Alerts: Three alert conditions support automated monitoring of trend transitions. "Bullish Trend Signal" fires on the first bar trend direction flips from bearish to bullish. "Bearish Trend Signal" fires on the first bar trend direction flips from bullish to bearish. "Any Trend Change" triggers on either transition for traders who want a single unified alert regardless of direction. All alerts include the exchange, ticker, and timeframe in the message for immediate context.
▶ Visual Customisation: Six colour presets, Classic, Aqua, Cosmic, Cyber, Neon, and Custom, provide coordinated bullish and bearish colour pairings suited to different chart themes and personal preferences. Selecting Custom exposes independent colour pickers for full manual control over both states. Optional bar colouring tints price candles with the active trend colour using a configurable transparency level, and optional background colouring extends the trend state tint across the full chart pane at a separately configurable transparency.
Release Notes
Added a toggleable neutral state that flags bars where the AFF line stays flat, meaning price failed to move beyond the friction band in either direction, and renders them in a customisable grey instead of the bullish or bearish trend colour.Open-source script
In true TradingView spirit, the creator of this script has made it open-source, so that traders can review and verify its functionality. Kudos to the author! While you can use it for free, remember that republishing the code is subject to our House Rules.
🟢 Access our best trading & investing tools: joinquantalgo.com/plans
🟢 Free Special Edition indicators: joinquantalgo.com/special-editions
🟢 Discord: discord.gg/hJaFyGEAyX
🟢 Free Special Edition indicators: joinquantalgo.com/special-editions
🟢 Discord: discord.gg/hJaFyGEAyX
Disclaimer
The information and publications are not meant to be, and do not constitute, financial, investment, trading, or other types of advice or recommendations supplied or endorsed by TradingView. Read more in the Terms of Use.
Open-source script
In true TradingView spirit, the creator of this script has made it open-source, so that traders can review and verify its functionality. Kudos to the author! While you can use it for free, remember that republishing the code is subject to our House Rules.
🟢 Access our best trading & investing tools: joinquantalgo.com/plans
🟢 Free Special Edition indicators: joinquantalgo.com/special-editions
🟢 Discord: discord.gg/hJaFyGEAyX
🟢 Free Special Edition indicators: joinquantalgo.com/special-editions
🟢 Discord: discord.gg/hJaFyGEAyX
Disclaimer
The information and publications are not meant to be, and do not constitute, financial, investment, trading, or other types of advice or recommendations supplied or endorsed by TradingView. Read more in the Terms of Use.