← Files Bigdata.comARCHIVED FILE
skills/bigdata-earnings-quality-screen/scripts/earnings_quality.py
20.7 KB · Oct 5, 2026 · 12:03 UTC
#!/usr/bin/env python3
"""
Earnings Quality Analysis and Beneish M-Score Calculator
This module provides comprehensive earnings quality assessment including:
- Operating cash flow to net income ratio
- Accrual ratio and quality metrics
- Working capital efficiency (DSO, DIO, DPO, Cash Conversion Cycle)
- Beneish M-Score for earnings manipulation detection
- Red flag identification with explanations
The Beneish M-Score uses 8 financial variables to detect potential
earnings manipulation. An M-Score > -1.78 suggests higher probability
of manipulation.
Author: Equity Analyst Skill
"""
from typing import Dict, List, Tuple, Optional
from dataclasses import dataclass
import math
@dataclass
class FinancialData:
"""
Financial data required for earnings quality analysis.
All values should be for the same fiscal period unless noted.
"""
# Income Statement
net_income: float
revenue: float
gross_profit: float
operating_income: float
cogs: float # Cost of goods sold
sga_expense: float # SG&A expenses
depreciation: float
# Balance Sheet - Current Period
total_assets: float
current_assets: float
cash: float
receivables: float
inventory: float
ppe_net: float # Property, plant & equipment (net)
current_liabilities: float
payables: float
long_term_debt: float
total_liabilities: float
# Balance Sheet - Prior Period
prior_total_assets: float
prior_receivables: float
prior_inventory: float
prior_ppe_gross: float
prior_current_assets: float
prior_current_liabilities: float
prior_long_term_debt: float
# Cash Flow Statement
operating_cash_flow: float
# Prior Period Income Statement
prior_revenue: float
prior_gross_profit: float
prior_sga_expense: float
prior_depreciation: float
@dataclass
class QualityMetrics:
"""Container for earnings quality metrics."""
ocf_ni_ratio: float # Operating CF / Net Income
accrual_ratio: float # (NI - OCF) / Avg Total Assets
days_sales_outstanding: float
days_inventory_outstanding: float
days_payable_outstanding: float
cash_conversion_cycle: float
quality_score: str # High/Medium/Low
@dataclass
class MScoreResult:
"""Container for Beneish M-Score results."""
dsri: float # Days Sales in Receivables Index
gmi: float # Gross Margin Index
aqi: float # Asset Quality Index
sgi: float # Sales Growth Index
depi: float # Depreciation Index
sgai: float # SG&A Index
lvgi: float # Leverage Index
tata: float # Total Accruals to Total Assets
m_score: float
manipulation_probability: str
def calculate_ocf_ni_ratio(
operating_cash_flow: float,
net_income: float
) -> Tuple[float, str]:
"""
Calculate Operating Cash Flow to Net Income ratio.
A healthy company should generate operating cash flow at least
equal to its reported net income. Consistently low ratios may
indicate aggressive revenue recognition or expense capitalization.
Args:
operating_cash_flow: Cash from operations
net_income: Reported net income
Returns:
Tuple of (ratio, interpretation)
"""
if net_income == 0:
return (float('inf') if operating_cash_flow > 0 else 0, "Net income is zero")
ratio = operating_cash_flow / net_income
if ratio >= 1.0:
interpretation = "Healthy: OCF exceeds or equals net income"
elif ratio >= 0.8:
interpretation = "Acceptable: OCF slightly below net income"
elif ratio >= 0.5:
interpretation = "Caution: Significant gap between OCF and NI"
else:
interpretation = "Warning: OCF much lower than net income"
return ratio, interpretation
def calculate_accrual_ratio(
net_income: float,
operating_cash_flow: float,
total_assets: float,
prior_total_assets: float
) -> Tuple[float, str]:
"""
Calculate the accrual ratio.
Accrual Ratio = (Net Income - Operating Cash Flow) / Average Total Assets
High accrual ratios indicate earnings are driven more by accounting
accruals than cash generation, which may be less sustainable.
Args:
net_income: Reported net income
operating_cash_flow: Cash from operations
total_assets: Current period total assets
prior_total_assets: Prior period total assets
Returns:
Tuple of (ratio, interpretation)
"""
avg_assets = (total_assets + prior_total_assets) / 2
if avg_assets == 0:
return 0, "Cannot calculate: zero assets"
accruals = net_income - operating_cash_flow
ratio = accruals / avg_assets
if ratio < 0:
interpretation = "Good: Cash earnings exceed accrual earnings"
elif ratio < 0.05:
interpretation = "Acceptable: Low accrual component"
elif ratio < 0.10:
interpretation = "Caution: Moderate accrual component"
else:
interpretation = "Warning: High accrual component in earnings"
return ratio, interpretation
def calculate_working_capital_metrics(
receivables: float,
inventory: float,
payables: float,
revenue: float,
cogs: float
) -> Dict[str, float]:
"""
Calculate working capital efficiency metrics.
DSO = (Receivables / Revenue) * 365
DIO = (Inventory / COGS) * 365
DPO = (Payables / COGS) * 365
CCC = DSO + DIO - DPO
Args:
receivables: Accounts receivable
inventory: Inventory balance
payables: Accounts payable
revenue: Annual revenue
cogs: Cost of goods sold
Returns:
Dictionary with DSO, DIO, DPO, and CCC
"""
dso = (receivables / revenue * 365) if revenue > 0 else 0
dio = (inventory / cogs * 365) if cogs > 0 else 0
dpo = (payables / cogs * 365) if cogs > 0 else 0
ccc = dso + dio - dpo
return {
'dso': dso,
'dio': dio,
'dpo': dpo,
'ccc': ccc
}
def calculate_beneish_mscore(data: FinancialData) -> MScoreResult:
"""
Calculate the Beneish M-Score for earnings manipulation detection.
The model uses 8 financial ratios to compute a score where:
- M-Score > -1.78 suggests higher probability of manipulation
- M-Score < -1.78 suggests lower probability of manipulation
The 8 variables are:
1. DSRI - Days Sales in Receivables Index
2. GMI - Gross Margin Index
3. AQI - Asset Quality Index
4. SGI - Sales Growth Index
5. DEPI - Depreciation Index
6. SGAI - SG&A Index
7. LVGI - Leverage Index
8. TATA - Total Accruals to Total Assets
Args:
data: FinancialData with current and prior period data
Returns:
MScoreResult with component scores and final M-Score
"""
# 1. Days Sales in Receivables Index (DSRI)
# DSRI = (Receivables_t / Sales_t) / (Receivables_t-1 / Sales_t-1)
dsr_current = data.receivables / data.revenue if data.revenue > 0 else 0
dsr_prior = data.prior_receivables / data.prior_revenue if data.prior_revenue > 0 else 0
dsri = dsr_current / dsr_prior if dsr_prior > 0 else 1.0
# 2. Gross Margin Index (GMI)
# GMI = Gross_Margin_t-1 / Gross_Margin_t
gm_current = data.gross_profit / data.revenue if data.revenue > 0 else 0
gm_prior = data.prior_gross_profit / data.prior_revenue if data.prior_revenue > 0 else 0
gmi = gm_prior / gm_current if gm_current > 0 else 1.0
# 3. Asset Quality Index (AQI)
# AQI = [1 - (CA_t + PPE_t) / TA_t] / [1 - (CA_t-1 + PPE_t-1) / TA_t-1]
aq_current = 1 - (data.current_assets + data.ppe_net) / data.total_assets if data.total_assets > 0 else 0
aq_prior = 1 - (data.prior_current_assets + data.prior_ppe_gross) / data.prior_total_assets if data.prior_total_assets > 0 else 0
aqi = aq_current / aq_prior if aq_prior > 0 else 1.0
# 4. Sales Growth Index (SGI)
# SGI = Sales_t / Sales_t-1
sgi = data.revenue / data.prior_revenue if data.prior_revenue > 0 else 1.0
# 5. Depreciation Index (DEPI)
# DEPI = (Dep_t-1 / (Dep_t-1 + PPE_t-1)) / (Dep_t / (Dep_t + PPE_t))
dep_rate_current = data.depreciation / (data.depreciation + data.ppe_net) if (data.depreciation + data.ppe_net) > 0 else 0
dep_rate_prior = data.prior_depreciation / (data.prior_depreciation + data.prior_ppe_gross) if (data.prior_depreciation + data.prior_ppe_gross) > 0 else 0
depi = dep_rate_prior / dep_rate_current if dep_rate_current > 0 else 1.0
# 6. SG&A Index (SGAI)
# SGAI = (SGA_t / Sales_t) / (SGA_t-1 / Sales_t-1)
sga_ratio_current = data.sga_expense / data.revenue if data.revenue > 0 else 0
sga_ratio_prior = data.prior_sga_expense / data.prior_revenue if data.prior_revenue > 0 else 0
sgai = sga_ratio_current / sga_ratio_prior if sga_ratio_prior > 0 else 1.0
# 7. Leverage Index (LVGI)
# LVGI = [(CL_t + LTD_t) / TA_t] / [(CL_t-1 + LTD_t-1) / TA_t-1]
lev_current = (data.current_liabilities + data.long_term_debt) / data.total_assets if data.total_assets > 0 else 0
lev_prior = (data.prior_current_liabilities + data.prior_long_term_debt) / data.prior_total_assets if data.prior_total_assets > 0 else 0
lvgi = lev_current / lev_prior if lev_prior > 0 else 1.0
# 8. Total Accruals to Total Assets (TATA)
# TATA = (NI - CFO) / TA
tata = (data.net_income - data.operating_cash_flow) / data.total_assets if data.total_assets > 0 else 0
# Calculate M-Score using Beneish (1999) coefficients
m_score = (
-4.84 +
0.920 * dsri +
0.528 * gmi +
0.404 * aqi +
0.892 * sgi +
0.115 * depi +
-0.172 * sgai +
4.679 * tata +
-0.327 * lvgi
)
# Determine manipulation probability
if m_score > -1.78:
probability = "HIGH - Likely manipulator"
elif m_score > -2.22:
probability = "MODERATE - Grey zone, needs further analysis"
else:
probability = "LOW - Unlikely manipulator"
return MScoreResult(
dsri=dsri,
gmi=gmi,
aqi=aqi,
sgi=sgi,
depi=depi,
sgai=sgai,
lvgi=lvgi,
tata=tata,
m_score=m_score,
manipulation_probability=probability
)
def identify_red_flags(
data: FinancialData,
quality_metrics: QualityMetrics,
mscore: MScoreResult
) -> List[Dict[str, str]]:
"""
Identify specific red flags in the financial data.
Args:
data: Financial data
quality_metrics: Calculated quality metrics
mscore: Beneish M-Score results
Returns:
List of red flag dictionaries with 'flag' and 'explanation' keys
"""
red_flags = []
# OCF/NI ratio check
if quality_metrics.ocf_ni_ratio < 0.5:
red_flags.append({
'flag': 'Low OCF/NI Ratio',
'explanation': f'Operating cash flow is only {quality_metrics.ocf_ni_ratio:.1%} of net income. '
'This suggests earnings quality issues - income may not be converting to cash.'
})
# Negative operating cash flow with positive net income
if data.operating_cash_flow < 0 and data.net_income > 0:
red_flags.append({
'flag': 'Negative OCF with Positive Net Income',
'explanation': 'Company reports profit but is burning cash from operations. '
'This is a significant warning sign.'
})
# High accrual ratio
if quality_metrics.accrual_ratio > 0.10:
red_flags.append({
'flag': 'High Accrual Ratio',
'explanation': f'Accrual ratio of {quality_metrics.accrual_ratio:.1%} indicates earnings '
'are heavily driven by accounting accruals rather than cash.'
})
# DSO increasing significantly
if mscore.dsri > 1.3:
red_flags.append({
'flag': 'Receivables Growing Faster Than Sales',
'explanation': f'DSRI of {mscore.dsri:.2f} suggests receivables are growing faster than revenue. '
'Could indicate aggressive revenue recognition or collection issues.'
})
# Gross margin declining
if mscore.gmi > 1.2:
red_flags.append({
'flag': 'Declining Gross Margins',
'explanation': f'GMI of {mscore.gmi:.2f} indicates gross margin compression. '
'Companies with declining margins may be tempted to manipulate earnings.'
})
# Asset quality deteriorating
if mscore.aqi > 1.3:
red_flags.append({
'flag': 'Deteriorating Asset Quality',
'explanation': f'AQI of {mscore.aqi:.2f} suggests increasing proportion of intangible assets. '
'May indicate aggressive capitalization of costs.'
})
# Very high sales growth (can incentivize manipulation)
if mscore.sgi > 1.5:
red_flags.append({
'flag': 'Unusually High Sales Growth',
'explanation': f'SGI of {mscore.sgi:.2f} (>{mscore.sgi*100-100:.0f}% growth). '
'Rapid growth can mask underlying issues and create pressure to maintain momentum.'
})
# High total accruals
if mscore.tata > 0.05:
red_flags.append({
'flag': 'High Total Accruals',
'explanation': f'TATA of {mscore.tata:.2%} indicates high accrual component. '
'Accruals above 5% of assets warrant scrutiny.'
})
# High leverage growth
if mscore.lvgi > 1.3:
red_flags.append({
'flag': 'Rapidly Increasing Leverage',
'explanation': f'LVGI of {mscore.lvgi:.2f} indicates significant increase in debt. '
'Rising leverage combined with other red flags is concerning.'
})
# M-Score itself
if mscore.m_score > -1.78:
red_flags.append({
'flag': 'Elevated M-Score',
'explanation': f'M-Score of {mscore.m_score:.2f} exceeds -1.78 threshold. '
'Statistical model suggests higher probability of earnings manipulation.'
})
# Cash conversion cycle
if quality_metrics.cash_conversion_cycle > 120:
red_flags.append({
'flag': 'Long Cash Conversion Cycle',
'explanation': f'CCC of {quality_metrics.cash_conversion_cycle:.0f} days is quite long. '
'Company takes extended time to convert investments to cash.'
})
return red_flags
def run_quality_analysis(data: FinancialData) -> Dict:
"""
Run comprehensive earnings quality analysis.
Args:
data: FinancialData with all required inputs
Returns:
Dictionary with quality_metrics, m_score_result, and red_flags
"""
# Calculate OCF/NI ratio
ocf_ni, ocf_interpretation = calculate_ocf_ni_ratio(
data.operating_cash_flow,
data.net_income
)
# Calculate accrual ratio
accrual, accrual_interpretation = calculate_accrual_ratio(
data.net_income,
data.operating_cash_flow,
data.total_assets,
data.prior_total_assets
)
# Calculate working capital metrics
wc_metrics = calculate_working_capital_metrics(
data.receivables,
data.inventory,
data.payables,
data.revenue,
data.cogs
)
# Determine overall quality score
if ocf_ni >= 0.9 and accrual < 0.05:
quality_score = "HIGH"
elif ocf_ni >= 0.7 and accrual < 0.10:
quality_score = "MEDIUM"
else:
quality_score = "LOW"
quality_metrics = QualityMetrics(
ocf_ni_ratio=ocf_ni,
accrual_ratio=accrual,
days_sales_outstanding=wc_metrics['dso'],
days_inventory_outstanding=wc_metrics['dio'],
days_payable_outstanding=wc_metrics['dpo'],
cash_conversion_cycle=wc_metrics['ccc'],
quality_score=quality_score
)
# Calculate M-Score
mscore_result = calculate_beneish_mscore(data)
# Identify red flags
red_flags = identify_red_flags(data, quality_metrics, mscore_result)
return {
'quality_metrics': quality_metrics,
'm_score_result': mscore_result,
'red_flags': red_flags,
'ocf_interpretation': ocf_interpretation,
'accrual_interpretation': accrual_interpretation
}
def format_quality_report(analysis: Dict, company_name: str = "Company") -> str:
"""
Format the quality analysis as a readable report.
Args:
analysis: Output from run_quality_analysis
company_name: Name of the company for the report header
Returns:
Formatted string report
"""
qm = analysis['quality_metrics']
ms = analysis['m_score_result']
lines = []
lines.append("=" * 70)
lines.append(f"EARNINGS QUALITY ANALYSIS: {company_name}")
lines.append("=" * 70)
lines.append("\n1. CASH FLOW QUALITY")
lines.append("-" * 40)
lines.append(f" OCF/Net Income Ratio: {qm.ocf_ni_ratio:.2f}x")
lines.append(f" --> {analysis['ocf_interpretation']}")
lines.append(f" Accrual Ratio: {qm.accrual_ratio:.2%}")
lines.append(f" --> {analysis['accrual_interpretation']}")
lines.append(f" Overall Quality Score: {qm.quality_score}")
lines.append("\n2. WORKING CAPITAL EFFICIENCY")
lines.append("-" * 40)
lines.append(f" Days Sales Outstanding (DSO): {qm.days_sales_outstanding:.0f} days")
lines.append(f" Days Inventory Outstanding (DIO): {qm.days_inventory_outstanding:.0f} days")
lines.append(f" Days Payable Outstanding (DPO): {qm.days_payable_outstanding:.0f} days")
lines.append(f" Cash Conversion Cycle: {qm.cash_conversion_cycle:.0f} days")
lines.append("\n3. BENEISH M-SCORE ANALYSIS")
lines.append("-" * 40)
lines.append(" Component Scores:")
lines.append(f" DSRI (Receivables Index): {ms.dsri:.3f}")
lines.append(f" GMI (Gross Margin Index): {ms.gmi:.3f}")
lines.append(f" AQI (Asset Quality Index): {ms.aqi:.3f}")
lines.append(f" SGI (Sales Growth Index): {ms.sgi:.3f}")
lines.append(f" DEPI (Depreciation Index): {ms.depi:.3f}")
lines.append(f" SGAI (SG&A Index): {ms.sgai:.3f}")
lines.append(f" LVGI (Leverage Index): {ms.lvgi:.3f}")
lines.append(f" TATA (Total Accruals/Assets): {ms.tata:.3f}")
lines.append(f"\n M-SCORE: {ms.m_score:.2f}")
lines.append(f" Threshold: -1.78 (higher = more likely manipulation)")
lines.append(f" Assessment: {ms.manipulation_probability}")
lines.append("\n4. RED FLAGS IDENTIFIED")
lines.append("-" * 40)
if analysis['red_flags']:
for i, flag in enumerate(analysis['red_flags'], 1):
lines.append(f"\n [{i}] {flag['flag']}")
lines.append(f" {flag['explanation']}")
else:
lines.append(" No significant red flags identified.")
lines.append("\n" + "=" * 70)
return "\n".join(lines)
if __name__ == "__main__":
# Example: Analyze a hypothetical company's earnings quality
print("=" * 70)
print("EARNINGS QUALITY ANALYSIS - EXAMPLE")
print("=" * 70)
# Create sample financial data
sample_data = FinancialData(
# Current period income statement
net_income=500,
revenue=10000,
gross_profit=4000,
operating_income=800,
cogs=6000,
sga_expense=3000,
depreciation=200,
# Current period balance sheet
total_assets=15000,
current_assets=5000,
cash=1000,
receivables=1500,
inventory=2000,
ppe_net=8000,
current_liabilities=3000,
payables=1200,
long_term_debt=4000,
total_liabilities=7000,
# Prior period balance sheet
prior_total_assets=13500,
prior_receivables=1200,
prior_inventory=1800,
prior_ppe_gross=7500,
prior_current_assets=4500,
prior_current_liabilities=2700,
prior_long_term_debt=3500,
# Cash flow
operating_cash_flow=400, # OCF lower than NI - potential issue
# Prior period income statement
prior_revenue=9000,
prior_gross_profit=3800,
prior_sga_expense=2700,
prior_depreciation=180
)
# Run analysis
analysis = run_quality_analysis(sample_data)
# Print formatted report
report = format_quality_report(analysis, "Sample Corp")
print(report)
# Also show raw metrics
print("\n" + "=" * 70)
print("RAW METRICS OUTPUT")
print("-" * 40)
qm = analysis['quality_metrics']
print(f"Quality Metrics Dict:")
print(f" ocf_ni_ratio: {qm.ocf_ni_ratio:.3f}")
print(f" accrual_ratio: {qm.accrual_ratio:.3f}")
print(f" dso: {qm.days_sales_outstanding:.1f}")
print(f" dio: {qm.days_inventory_outstanding:.1f}")
print(f" dpo: {qm.days_payable_outstanding:.1f}")
print(f" ccc: {qm.cash_conversion_cycle:.1f}")
print(f" quality_score: {qm.quality_score}")
ms = analysis['m_score_result']
print(f"\nM-Score: {ms.m_score:.2f}")
print(f"Red Flags Count: {len(analysis['red_flags'])}")
SHA-256: e5fe09d46ad59ab887ef27fe862687a891e3659e563c656c7d7394056d395abc