Dechow-Dichev Accrual Quality

Install and import#

bash

npm install fintech-algorithms

import { dechowDichevAccrualQuality } from "fintech-algorithms/fundamental-analysis-and-valuation/quality-and-distress/dechow-dichev-accrual-quality";

Signature#

dechowDichevAccrualQuality(data)

Runs the firm-specific Dechow-Dichev regression of scaled working-capital accruals on lagged, current and leading operating cash flow, and reports the residual standard deviation as the accrual-quality measure.

Parameters#

Name	Type	Notes
`data`	`{ observations: { period: string; cfo_scaled: number; working_capital_accrual_scaled: number }[] }`	`observations` must be an array of at least seven period records ordered in time. Each record needs a unique non-empty `period` label, a `cfo_scaled` value and a `working_capital_accrual_scaled` value. The first and last records are used only as the lag and lead for their neighbours, so the regression is fitted on the interior periods.

Returns#

{ state: string; method: string; coefficients: { intercept: number; cfo_t_minus_1: number; cfo_t: number; cfo_t_plus_1: number }; estimation_periods: string[]; fitted_accruals: number[]; residuals: number[]; accrual_quality: number; interpretation: string; available_after_period: string }

coefficients holds the fitted intercept and the three cash-flow slopes. estimation_periods lists the interior period labels actually regressed, with fitted_accruals and residuals aligned to it. accrual_quality is the residual standard deviation using degrees of freedom equal to the observation count less the four coefficients, so a larger value means lower quality, which interpretation restates. available_after_period is the label of the final observation, since the last interior fit needs its lead. method is dechow-dichev-2002-firm-specific and state is calculated.

Errors#

When data is not a plain object — throws TypeError
When cfo_scaled or working_capital_accrual_scaled is missing or not a finite number — throws TypeError
When observations is not an array, holds fewer than seven records, or contains a non-object entry — throws RangeError
When a period label is empty or repeated — throws RangeError
When the regression design matrix is singular — throws RangeError

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

{
  "observations": [
    {
      "period": "Y1",
      "cfo_scaled": 0.082,
      "working_capital_accrual_scaled": 0.011
    },
    {
      "period": "Y2",
      "cfo_scaled": 0.064,
      "working_capital_accrual_scaled": 0.018
    },
    {
      "period": "Y3",
      "cfo_scaled": 0.091,
      "working_capital_accrual_scaled": -0.006
    }
  ]
}

Call#

dechowDichevAccrualQuality(data)

Returns#

object with 9 fields: state, method, coefficients, estimation_periods, fitted_accruals, residuals, accrual_quality, interpretation, …

{
  "state": "calculated",
  "method": "dechow-dichev-2002-firm-specific",
  "coefficients": {
    "intercept": 0.03644529821051595,
    "cfo_t_minus_1": 0.04149503236651636,
    "cfo_t": -0.5247590858810153,
    "cfo_t_plus_1": 0.06807435790812336
  },
  "estimation_periods": ["Y2", "Y3", "Y4", "Y5", "Y6", "Y7"],
  "fitted_accruals": [
    0.012458075937824543,
    -0.0054526017115175875,
    0.022705476699814174,
    -0.011870859874287619,
    0.00953492500736294,
    -0.0031794130798601925
  ],
  "residuals": [
    0.005541924062175456,
    -0.0005473982884824126,
    -0.0017054766998141728,
    -0.0001291401257123813,
    -0.0005349250073629404,
    0.00017941307986019247
  ],
  "accrual_quality": 0.003108868359945476,
  "interpretation": "higher residual dispersion means lower accrual quality",
  "available_after_period": "Y10"
}

Other exports#

This module also exports calculate, altmanZScore, piotroskiFScore, beneishMScore, sloanAccrualMeasure, ohlsonOScore, zmijewskiXScore, springateSScore, tafflerZScore, fulmerHScore, groverGScore, dechowFScoreForMisstatementRisk, modifiedJonesDiscretionaryAccrualModel. 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.