Indicator

Fathom Walls

Liquidity walls defending repeated highs or lows.

At a glance

Identifier
liquidity-walls
Objects received
signal
Historical cost
1 compute unit per trade

What the indicator shows

  • A sell wall sits on a high tested several times, where aggressive buying fails to break the level; the Fathom chart marks it with an S.
  • A buy wall sits on a low tested several times, where aggressive selling fails to break it; the Fathom chart marks it with a B.
  • In both cases, passive orders absorb the aggression: the level is defended.

How to use it

  • A wall gives a level to trade for a bounce while it holds, or to watch for a breakout if it gives way.
  • On NQ, start with the looser settings of the examples, then tighten the criteria to keep only the clearest walls. These settings suit NQ only: on ES, they give many more walls. Stricter criteria do not guarantee that a level will hold.

How to read and draw it

  • A signal marks a timeframe-second bar that tests a wall: a sell wall at that bar’s high, a buy wall at its low.
  • The parameters set how demanding the detection is. Raising minimumBars, minPerBarVolume, minClusterVolume or minDeltaPercent gives fewer, stronger walls. Raising nearnessBars or tickBreakoutMargin makes it more tolerant: tests further apart, small breaks of the level. tickGrouping sets the thickness of the wall.
  • A signal carries one price, not the wall’s extent: draw it as a horizontal line at price from the signal bar (its id).
  • volume is the volume traded on the wall during its tests, sideVolume the part of it from the aggressive orders the wall absorbed, and cells a defense score: the higher, the more the level was defended.
  • The generic defaults are strict: on a ticker without its own defaults, they can give few walls, sometimes none over a whole session. The examples use settings looser than NQ’s defaults: minPerBarVolume: 10, minClusterVolume: 30, minDeltaPercent: 40 and tickGrouping: 4, the other parameters at their defaults. They suit NQ and do not carry over to another ticker: on ES, they give many more walls. The stricter the parameters, the fewer the walls.

When it updates

  • At most one signal per bar, sent final when the bar closes, that is on the first trade of a later bar.
  • Weekly reset: at the Sunday reopen (18:00 New York time), the detection starts over: the previous week no longer counts.

Subscribe

The same params work in real time, on historical data and in exports. Parameters you omit take the defaults of the instrument’s ticker.

Subscribe message
{
  "t": "subscribe",
  "sub": "liquidity-walls",
  "instrument": "NQ.front",
  "indicator": "liquidity-walls",
  "params": {
    "minClusterVolume": 30,
    "minDeltaPercent": 40,
    "minPerBarVolume": 10,
    "minimumBars": 2,
    "nearnessBars": 20,
    "tickBreakoutMargin": 1,
    "tickGrouping": 4,
    "timeframe": 60
  },
  "mode": "live",
  "from": "live"
}

These example parameters differ from the current NQ ticker defaults: minClusterVolume 30 (default 60), minDeltaPercent 40 (default 70), minPerBarVolume 10 (default 20), tickGrouping 4 (default 6). On NQ, this subscription therefore counts as a custom configuration; the Sandbox allows 1.

Parameters

As soon as a parameter differs from its default on the instrument’s ticker, the subscription counts as a custom configuration. To change a parameter, open a new subscription.

The Default column gives the generic default, used by the tickers without a default of their own, then each ticker that has its own value, such as CL 15. Ticker defaults are regularly updated, ticker by ticker: an update changes the results of the subscriptions and queries that rely on them. A subscription whose parameters are all at its instrument’s defaults is not a custom configuration. Only explicit parameters keep a computation identical over time: pass them to keep the same results.

ParameterTypeRoleDefaultAllowed valuesUnitDescription
minClusterVolumeintegerComputation400 · CL 15 · ES 300 · GC 30 · HG 15 · NG 60 · NQ 60 · RTY 15 · SI 15 · YM 15 · ZB 16000 to 10,000,000contractsMinimum volume of the wall as a whole, in contracts. Higher = fewer, heavier walls.
minDeltaPercentintegerComputation70 · CL 40 · RTY 40 · YM 550 to 100%How clear the absorption on the wall must be, in percent. Higher = fewer walls, with clearer absorption.
minPerBarVolumeintegerComputation100 · CL 5 · GC 10 · HG 5 · NG 20 · NQ 20 · RTY 5 · SI 5 · YM 5 · ZB 2000 to 1,000,000contractsMinimum volume for a test of the level to count, in contracts. Higher = fewer walls.
minimumBarsintegerComputation22 to 100—How many times the level must be tested. Higher = fewer walls.
nearnessBarsintegerComputation201 to 1,000barsHow far back, in bars, earlier tests of the level count. Higher = tests further apart can form a wall.
tickBreakoutMarginintegerComputation10 to 100ticksTolerance to small breaks of the level, in ticks. Higher = more tolerant to small breaks.
tickGroupingintegerComputation1 · ES 2 · GC 5 · NQ 6 · SI 3 · YM 21 to 100ticksThickness of the wall, in ticks. Higher = a thicker wall.
timeframeintegerComputation601 · 5 · 10 · 15 · 30 · 60 · 120 · 180 · 300 · 600 · 900 · 1800 · 3600 · 7200 · 10800 · 14400 · 86400secondsBar duration in seconds: from 1 second (1) to 1 day (86400). Bars are counted from the product's session open (18:00 New York time for every current product): 4-hour bars start at 18:00, 22:00, 02:00 and so on, and a 1-day bar covers the whole session.

Objects received

Each object keeps the same id from one update to the next. Its data is in the data field. Prices are in ticks: Prices and timestamps explains how to convert them.

An object sent in progress arrives with final: false, then again on every change, and each version fully replaces the previous one. Its description says when it ends: most objects end with a last version carrying final: true, and some are removed without one. An object sent final is sent once, with final: true.

Fields and units

Prices are in ticks (Prices and timestamps). Timestamps are nanoseconds since the Unix epoch, sent as strings: parse them as 64-bit integers. The fields below place the objects in time.

FieldUnitHow to read it
signal.idstring · <barStartNs>Start of the bar that tests the wall: draw a horizontal line at price from that bar.

signal sent final

Signal of the finished bar that starts at barStartNs.

id format: <barStartNs>.

FieldTypeUnitPresentDescription
cellsinteger—alwaysDefense score of the wall: the higher, the more the level was defended.
priceintegerticksalwaysHigh (sell wall) or low (buy wall) of the bar that tests the wall.
sidestring · buy | sell—alwayssell: sell wall at the high; buy: buy wall at the low.
sideVolumeintegercontractsalwaysPart of volume from the aggressive orders the wall absorbed: buying for a sell wall, selling for a buy wall.
volumeintegercontractsalwaysVolume traded on the wall during its tests, in contracts: the bars that tested it before, and this one.

Example messages

Messages as the WebSocket sends them to the subscription above, taken from a real NQ session. Historical data returns the same objects, with the same cursor, without the sub field.

The params of the subscription above are the current NQ ticker defaults, written out in full, except minClusterVolume 30, minDeltaPercent 40, minPerBarVolume 10, tickGrouping 4.

Final object
{
  "sub": "liquidity-walls",
  "cursor": "20720.13976.0",
  "t": "upsert",
  "id": "1790217900000000000",
  "final": true,
  "ts": "1790217960047941663",
  "data": {
    "price": 122988,
    "side": "sell",
    "volume": 186,
    "sideVolume": 162,
    "cells": 4
  }
}

JSON schemas

To validate or type your data: the JSON schemas of the parameters and of data, also available from GET /v1/indicators. The SDKs can generate types from them. Rules that compare two parameters, such as a maximum of at least its minimum, and the two decimals of decimal parameters, are only in the parameter descriptions: the API checks them too.

Parameters schema
JSON
{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://fathomcharts.com/schemas/liquidity-walls/params.json",
  "type": "object",
  "additionalProperties": false,
  "properties": {
    "minClusterVolume": {
      "description": "Minimum volume of the wall as a whole, in contracts. Higher = fewer, heavier walls.",
      "type": "integer",
      "default": 400,
      "minimum": 0,
      "maximum": 10000000
    },
    "minDeltaPercent": {
      "description": "How clear the absorption on the wall must be, in percent. Higher = fewer walls, with clearer absorption.",
      "type": "integer",
      "default": 70,
      "minimum": 0,
      "maximum": 100
    },
    "minPerBarVolume": {
      "description": "Minimum volume for a test of the level to count, in contracts. Higher = fewer walls.",
      "type": "integer",
      "default": 100,
      "minimum": 0,
      "maximum": 1000000
    },
    "minimumBars": {
      "description": "How many times the level must be tested. Higher = fewer walls.",
      "type": "integer",
      "default": 2,
      "minimum": 2,
      "maximum": 100
    },
    "nearnessBars": {
      "description": "How far back, in bars, earlier tests of the level count. Higher = tests further apart can form a wall.",
      "type": "integer",
      "default": 20,
      "minimum": 1,
      "maximum": 1000
    },
    "tickBreakoutMargin": {
      "description": "Tolerance to small breaks of the level, in ticks. Higher = more tolerant to small breaks.",
      "type": "integer",
      "default": 1,
      "minimum": 0,
      "maximum": 100
    },
    "tickGrouping": {
      "description": "Thickness of the wall, in ticks. Higher = a thicker wall.",
      "type": "integer",
      "default": 1,
      "minimum": 1,
      "maximum": 100
    },
    "timeframe": {
      "description": "Bar duration in seconds: from 1 second (`1`) to 1 day (`86400`). Bars are counted from the product's session open (18:00 New York time for every current product): 4-hour bars start at 18:00, 22:00, 02:00 and so on, and a 1-day bar covers the whole session.",
      "type": "integer",
      "default": 60,
      "enum": [
        1,
        5,
        10,
        15,
        30,
        60,
        120,
        180,
        300,
        600,
        900,
        1800,
        3600,
        7200,
        10800,
        14400,
        86400
      ]
    }
  }
}
data field schema
JSON
{
  "$schema": "https://json-schema.org/draft/2020-12/schema",
  "$id": "https://fathomcharts.com/schemas/liquidity-walls/data.json",
  "title": "signal",
  "type": "object",
  "additionalProperties": false,
  "properties": {
    "cells": {
      "description": "Defense score of the wall: the higher, the more the level was defended.",
      "type": "integer"
    },
    "price": {
      "description": "High (sell wall) or low (buy wall) of the bar that tests the wall.",
      "type": "integer"
    },
    "side": {
      "description": "`sell`: sell wall at the high; `buy`: buy wall at the low.",
      "type": "string",
      "enum": [
        "buy",
        "sell"
      ]
    },
    "sideVolume": {
      "description": "Part of `volume` from the aggressive orders the wall absorbed: buying for a sell wall, selling for a buy wall.",
      "type": "integer"
    },
    "volume": {
      "description": "Volume traded on the wall during its tests, in contracts: the bars that tested it before, and this one.",
      "type": "integer"
    }
  },
  "required": [
    "price",
    "side",
    "volume",
    "sideVolume",
    "cells"
  ]
}