Insight Grades

Analyzes a company’s financial health across 7 areas, assigns A~F grades, detects anomalies, and classifies an overall profile.

Usage

from dartlab.engines.insight import analyze

result = analyze("005930")

# 10-area grades
result.grades()
# {
#   "performance": "A",
#   "profitability": "A",
#   "health": "A",
#   "cashflow": "B",
#   "governance": "A",
#   "risk": "B",
#   "opportunity": "A"
# }

# Summary
result.profile   # "premium"
result.summary   # "삼성전자는 실적, 수익성, 재무건전성 등 대부분..."
result.anomalies # [] (no anomalies)

analyze()

analyze(stockCode: str, company: Company = None) -> AnalysisResult | None

Parameter	Type	Description
`stockCode`	str	Stock code (6 digits)
`company`	Company (optional)	Company instance. Created internally if None

Returns None if data is insufficient.

Analysis Process

Build quarterly/annual financial time series
Calculate financial ratios
Auto-detect financial sector (6 signals: debt ratio, interest income, etc.)
Grade 7 areas
Detect anomalies (8 rules)
Calculate distress scorecard (4-axis composite)
Classify profile
Generate Korean-language summary

7 Analysis Areas

performance (Earnings Growth)

Analyzes revenue and operating income growth rates and volatility.

Grade	Criteria
A	Both revenue and operating income growing, low volatility
B	One growing or moderate volatility
C	Stagnant
D	Declining
F	Persistent decline

profitability (Profitability)

Analyzes operating margin, net margin, ROE, and ROA. Sector benchmark adjustments are applied.

Grade	Criteria
A	Operating margin 10%+, ROE 15%+
B	Operating margin 5%+, ROE 8%+
C	Operating margin positive, ROE positive
D	Operating margin or ROE negative
F	Both significantly negative

health (Financial Health)

Analyzes debt ratio and current ratio. Separate criteria applied for financial sector.

Grade	General Company Criteria
A	Debt ratio below 50%, current ratio above 200%
B	Debt ratio below 100%, current ratio above 150%
C	Debt ratio below 200%, current ratio above 100%
D	Debt ratio below 400%
F	Debt ratio above 400% or capital erosion

cashflow (Cash Flow)

Analyzes operating cash flow, free cash flow (FCF), FCF margin, and trends.

Grade	Criteria
A	Operating CF positive + FCF positive + improving trend
B	Operating CF positive + FCF positive
C	Operating CF positive, FCF negative
D	Operating CF negative
F	Persistently negative operating CF

governance (Governance)

Analyzes largest shareholder ownership ratio, audit opinion, and dividend status.

Grade	Criteria
A	Unqualified opinion + stable ownership + dividend
B	Unqualified opinion + partially met
C	Unqualified opinion, rest unmet
D	Audit opinion warning
F	Adverse opinion / Disclaimer of opinion

risk (Overall Risk)

Synthesizes risk flags from the other 6 areas.

opportunity (Investment Opportunity)

Synthesizes opportunity flags from the other 6 areas.

AnalysisResult

Field	Type	Description
`corpName`	str	Company name
`stockCode`	str	Stock code
`isFinancial`	bool	Whether financial sector
`performance`	InsightResult	Earnings growth
`profitability`	InsightResult	Profitability
`health`	InsightResult	Financial health
`cashflow`	InsightResult	Cash flow
`governance`	InsightResult	Governance
`risk`	InsightResult	Overall risk
`opportunity`	InsightResult	Investment opportunity
`anomalies`	list[Anomaly]	List of anomalies
`distress`	DistressResult	Distress prediction scorecard
`summary`	str	Korean-language summary
`profile`	str	Company profile

grades() Method

result.grades() -> dict[str, str]

Returns grades for all 7 areas as a dict.

InsightResult

Analysis result for each area.

Field	Type	Description
`grade`	str	Grade (A/B/C/D/F, N=no data)
`summary`	str	Korean summary
`details`	list[str]	Detailed analysis items
`risks`	list[Flag]	Risk flags
`opportunities`	list[Flag]	Opportunity flags

Anomaly Detection

Detects financial anomaly signals using 8 rules.

Category	Detection Target
`earningsQuality`	Operating income increasing but operating CF decreasing
`workingCapital`	Rapid increase in trade receivables / inventories
`balanceSheetShift`	Debt/borrowings/equity changed by more than +/-50%, capital erosion
`cashBurn`	Sharp cash decline, operating CF deficit + financing CF positive (debt-dependent)
`marginDivergence`	Operating margin changed by +/-5%p, non-operating income sharp change
`financialSector`	Financial sector debt ratio sharp change, net income sharp decline
`trendDeterioration`	Consecutive net losses, operating CF losses, ICR below 1, rising debt ratio
`cccDeterioration`	Cash conversion cycle expanding 3+ consecutive years

for a in result.anomalies:
    print(f"[{a.severity}] {a.category}: {a.text}")
    # [warning] earningsQuality: 영업이익은 증가했으나 영업CF는 감소

Anomaly

Field	Type	Description
`severity`	str	“danger”, “warning”, “info”
`category`	str	Anomaly type
`text`	str	Detailed description
`value`	float (optional)	Numeric value

Distress Prediction Scorecard

The distress field provides a comprehensive distress prediction report with evidence-based formatting.

result = analyze("005930")
print(result.distress)
# === 부실 예측 스코어카드 ===
# 종합: safe (2.1/100) | 신용등급: AAA (투자적격 최상위)
# ...

DistressResult

Field	Type	Description
`level`	str	safe/watch/warning/danger/critical
`overall`	float	Composite score (0~100)
`creditGrade`	str	S&P-mapped credit grade (AAA~D)
`creditDescription`	str	Grade description in Korean
`axes`	list[DistressAxis]	4 weighted axes with model details
`cashRunwayMonths`	float?	Estimated months of cash runway
`liquidityAlert`	str?	Liquidity alert level
`riskFactors`	list[str]	Structured risk factors
`modelCount`	int	Number of models used
`dataQuality`	str	충분/보통/부족

4 Axes

Axis	Weight	Models
정량 분석	40%	Ohlson O-Score, Altman Z”-Score, Altman Z-Score
이익 품질	20%	Beneish M-Score, Sloan Accrual, Piotroski F-Score
추세 분석	30%	trendDeterioration, cccDeterioration anomalies
감사 위험	10%	audit/governance anomalies

Credit Grade Mapping

Overall Score	Grade	Description
5 미만	AAA	투자적격 최상위
10 미만	AA	투자적격 상위
15 미만	A	투자적격
25 미만	BBB	투자적격 하한
35 미만	BB	투기등급
50 미만	B	투기등급 하위
65 미만	CCC	상당한 부실 위험
80 미만	CC	부실 임박
90 미만	C	부도 직전
90 이상	D	부도 수준

ModelScore

Each quantitative/quality model produces a ModelScore:

Field	Type	Description
`name`	str	Model name (e.g., “Ohlson O-Score”)
`rawValue`	float	Raw computed value
`displayValue`	str	Human-readable value (e.g., “P(부도) 0.5%“)
`zone`	str	safe / gray / distress
`interpretation`	str	Evidence-based interpretation
`reference`	str	Academic reference

DistressAxis

Field	Type	Description
`name`	str	Axis name
`score`	float	Normalized score (0~100)
`weight`	float	Weight in composite (0.0~1.0)
`models`	list[ModelScore]	Individual model results
`summary`	str	Axis-level summary

Company Profiles

Classifies companies into one of 6 profiles based on grade combinations.

Profile	Condition
`premium`	High average grade + low risk
`growth`	Performance / profitability / opportunity all top-tier
`stable`	Health / risk / profitability all stable
`caution`	Risk D/F or health F
`distress`	Most areas low-grade
`mixed`	Does not match above conditions

Examples

from dartlab.engines.insight import analyze

result = analyze("005930")

# Check grades
for area, grade in result.grades().items():
    print(f"{area}: {grade}")

# Profitability details
print(result.profitability.summary)
for d in result.profitability.details:
    print(f"  - {d}")

# Risk flags
for flag in result.risk.risks:
    print(f"[{flag.level}] {flag.category}: {flag.text}")

# Anomalies
if result.anomalies:
    print("=== Anomalies Detected ===")
    for a in result.anomalies:
        print(f"  [{a.severity}] {a.text}")

# Summary
print(f"Profile: {result.profile}")
print(result.summary)