Stock Score History, Migration, and Change Attribution

Install and import#

bash

npm install fintech-algorithms

import { stockScoreHistoryMigrationAndChangeAttribution } from "fintech-algorithms/fundamental-analysis-and-valuation/integrated-equity-scoring/stock-score-history-migration-and-change-attribution";

Signature#

stockScoreHistoryMigrationAndChangeAttribution(data)

Compares the two most recent score snapshots, reports the overall change and the band migration between them, and attributes the move to the component that shifted furthest.

Parameters#

Name	Type	Notes
`data`	`{ history: Array<{ as_of: string; overall_score: number; band: string; components: Record<string, number> }> }`	`history` needs at least two snapshots in strictly increasing `as_of` order, each a `YYYY-MM-DD` date. Every row needs a `components` object, and the last two rows must each carry an `overall_score` and a `band`; only those two are compared, and every component score read from them must be between 0 and 100.

Returns#

{ state: string; method: string; previous_as_of: string; latest_as_of: string; previous_score: number; latest_score: number; score_change: number; prior_band: string; latest_band: string; migration: string; migration_direction: number; component_deltas: Record<string, number>; top_driver: string; clock_policy: string }

score_change is the latest overall score minus the previous one and sets state to improved, deteriorated or unchanged. component_deltas covers the sorted union of both snapshots' component keys and top_driver names the key with the largest absolute delta, ties going to the later key name. migration is unchanged or the two bands joined by an arrow, and migration_direction is their rank difference on the abstain, watch, eligible, strong ladder.

Errors#

When history is not a list of at least two points, or its dates are not strictly increasing — throws RangeError
When a history row lacks a components object — throws TypeError
When a history as_of is not a YYYY-MM-DD string — throws TypeError
When a compared overall_score or component score is below 0 or above 100 — throws RangeError

Complexity: time O(n + k log k), space O(n + k).

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

{
  "history": [
    {
      "as_of": "2024-12-31",
      "overall_score": 62,
      "band": "watch",
      "components": {
        "health": 60,
        "quality": 64,
        "resilience": 62
      }
    },
    {
      "as_of": "2025-03-31",
      "overall_score": 70,
      "band": "eligible",
      "components": {
        "health": 72,
        "quality": 64,
        "resilience": 68
      }
    },
    {
      "as_of": "2025-06-30",
      "overall_score": 78,
      "band": "strong",
      "components": {
        "health": 80,
        "quality": 70,
        "resilience": 74
      }
    }
  ]
}

Call#

stockScoreHistoryMigrationAndChangeAttribution(data)

Returns#

object with 14 fields: state, method, previous_as_of, latest_as_of, previous_score, latest_score, score_change, prior_band, …

{
  "state": "improved",
  "method": "point-in-time-score-history-change-attribution",
  "previous_as_of": "2025-03-31",
  "latest_as_of": "2025-06-30",
  "previous_score": 70,
  "latest_score": 78,
  "score_change": 8,
  "prior_band": "eligible",
  "latest_band": "strong",
  "migration": "eligible -> strong",
  "migration_direction": 1,
  "component_deltas": {
    "health": 8,
    "quality": 6,
    "resilience": 6
  },
  "top_driver": "health",
  "clock_policy": "only knowledge-available snapshots are comparable"
}

Other exports#

This module also exports calculate, pointInTimeStockScoringInputAssembly, stockScoringPeerCohortResolver, fundamentalMetricDirectionAndPeerNormalization, modelApplicabilityAndVariantRouter, accountingFinancialHealthComposite, earningsQualityComposite, dividendSafetyScore, balanceSheetResilienceScore, distressModelEnsemble, crossModelConflictAndDoubleCountingResolver, overallExplainableStockScore, scoreConfidenceMissingDataPenaltyAndAbstention, marketWideStockScreeningAndRanking. 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.