Bulk Volume Classification

Install and import#

bash

npm install fintech-algorithms

import { bulkVolumeClassification } from "fintech-algorithms/market-microstructure/trade-classification/bulk-volume-classification";

Signature#

bulkVolumeClassification(data, config)

Assigns a *fraction* of each volume bar to buyers rather than signing individual trades. Designed for aggregated data where trade-level signing is impossible, and the input to VPIN.

Parameters#

Name	Type	Notes
`data`	`ClassificationInput`	Trades with prices, and where the rule needs them, the prevailing quotes.
`config`	`ClassificationConfig`	Tolerances and tie-breaking rules. Trades exactly at a quote or unchanged in price are the cases where classifiers differ, so the tie rule is part of the contract.

Returns#

{ classifications, summary, diagnostics }

Per-trade sign with the rule that produced it, plus counts of the unclassifiable cases. Bulk volume classification disagrees with its siblings on exactly those, and hiding them would hide the disagreement.

Errors#

When required quote data is absent for a rule that needs it — reported per trade rather than thrown

Complexity: time O(n), space O(n).

Worked example#

executed Captured by running this function on the input its own test provides. Real output of real code — but not asserted against a published figure.

Input#

data

{
  "starting_price": 100,
  "bars": [
    {
      "id": "B001",
      "sequence": 1,
      "event_time_ms": 1000,
      "close": 100.05,
      "volume": 1000,
      "status": "valid"
    },
    {
      "id": "B002",
      "sequence": 2,
      "event_time_ms": 61000,
      "close": 100.05,
      "volume": 1125,
      "status": "valid"
    },
    {
      "id": "B003",
      "sequence": 3,
      "event_time_ms": 121000,
      "close": 100.025,
      "volume": 1250,
      "status": "valid"
    }
  ]
}

config

{
  "sigma": 0.05,
  "degrees_of_freedom": 1
}

Call#

bulkVolumeClassification(data, config)

Returns#

object with 17 fields: state, method, trace, total_valid, classified_count, unknown_count, buy_count, sell_count, …

{
  "state": "ok",
  "method": "bulk-volume-classification",
  "trace": [
    {
      "id": "B001",
      "sequence": 1,
      "event_time_ms": 1000,
      "price": 100.05,
      "volume": 1000,
      "sign": 1,
      "side": "buy-leaning",
      "reason": "student-t-volume-split",
      "price_change": 0.05,
      "z_score": 1,
      "buy_fraction": 0.75,
      "buy_volume": 750,
      "sell_volume": 250,
      "cumulative_signed_volume": 500
    },
    {
      "id": "B002",
      "sequence": 2,
      "event_time_ms": 61000,
      "price": 100.05,
      "volume": 1125,
      "sign": 0,
      "side": "balanced",
      "reason": "student-t-volume-split",
      "price_change": 0,
      "z_score": 0,
      "buy_fraction": 0.5,
      "buy_volume": 562.5,
      "sell_volume": 562.5,
      "cumulative_signed_volume": 500
    },
    {
      "id": "B003",
      "sequence": 3,
      "event_time_ms": 121000,
      "price": 100.025,
      "volume": 1250,
      "sign": -1,
      "side": "sell-leaning",
      "reason": "student-t-volume-split",
      "price_change": -0.025,
      "z_score": -0.5,
      "buy_fraction": 0.3524163823,
      "buy_volume": 440.520477937,
      "sell_volume": 809.479522063,
      "cumulative_signed_volume": 131.040955874
    }
  ],
  "total_valid": 60,
  "classified_count": 60,
  "unknown_count": 0,
  "buy_count": 24,
  "sell_count": 24,
  "buy_volume": 47435.5504890151,
  "sell_volume": 41439.4495109849,
  "unknown_volume": 0,
  "signed_volume": 5996.1009780303,
  "total_volume": 88875,
  "coverage": 1
}

Showing 14 of 17 fields.

Other exports#

This module also exports tickTest, quoteTest, leeReady, studentTCdf, runTopic. Every module additionally exports run as an alias of its primary function, and a meta object carrying its catalog id, domain, family, shape and article URL.

Diagrams#

Calculation flow#

Bulk Volume Classification Calculation Flow

flowchart LR
    A["Final synthetic records"] --> B["Order by event time and sequence"]
    B --> C["Apply status and condition filters"]
    C --> D["Enforce availability and causal reference"]
    D --> E["Run Bulk Volume Classification"]
    E --> F["Sign or fractional volume"]
    F --> G["Reason code and trace"]
    G --> H["Coverage and imbalance"]
    H --> I["Sensitivity comparison"]

How it works#

This page states the contract — how to call it correctly. The article explains the concept: why it works, and where it breaks.

Read the article →

References#

Discerning Information from Trade Data — David Easley, Marcos Lopez de Prado, and Maureen O'Hara
Nasdaq TotalView-ITCH 5.0 Specification — Nasdaq
Daily TAQ Client Specification — NYSE / ICE Data Services

Bulk Volume Classification

Install and import#

Signature#

Parameters#

Returns#

Errors#

Worked example#

Input#

Call#

Returns#

Other exports#

Diagrams#

Calculation flow#

How it works#

References#

The rest of the Trade Classification family#