Earnings Quality Reviewer

intent

assess whether reported earnings are reliable enough to support valuation conviction. use this skill when you have structured financial statements (income statement, balance sheet, cash flow) and want to surface accrual risk, working capital stress, leverage concerns, and accounting red flags before running a valuation. the skill takes no external data feeds, only what you supply.

inputs

structured financial statements (required):

• income statement: revenue, gross profit, operating profit, finance cost, tax, net income, one-off and non-recurring items when available.
• balance sheet: cash, receivables, inventory, current assets, debt, payables, current liabilities, equity, total assets.
• cash flow statement: operating cash flow, capital expenditure, free cash flow, financing cash flows, material working capital movements.
• ratios or raw fields to compute: current ratio, quick ratio, debt-to-equity, interest coverage, receivables turnover, inventory turnover, days sales outstanding, cash conversion cycle.

optional notes (improve confidence):

• accounting policy changes.
• related-party transactions and balances.
• debt maturity schedule and covenant terms.
• contingencies and legal exposures.
• one-off items, impairments, restructuring charges, asset sales, FX effects, tax adjustments.
• auditor concerns or restatements.

metadata:

• ticker or entity identifier.
• statement as-of date.
• reporting period (annual, quarterly).
• fiscal year end (for seasonality context).

edge cases to flag:

• stale statements (>90 days old relative to valuation date).
• missing cash flow statement or operating cash flow.
• missing balance sheet line items (e.g., no inventory disclosed, no debt detail).
• no notes section supplied.
• negative or zero operating cash flow.
• empty result sets (e.g., no receivables, no one-offs disclosed).

procedure

1. validate input completeness (explicit gate step).

   • input: income statement, balance sheet, cash flow statement, optional notes.
   • output: pass/fail gate. if fail, document gaps and continue at reduced confidence.
   • check: is income statement present with revenue, net income, and gross profit or cost of goods sold? is balance sheet present with current assets, current liabilities, and equity? is cash flow statement present with operating cash flow and capital expenditure or is free cash flow derivable?
   • gate rule: if any of the three statements is missing entirely, flag as "low confidence" but continue with available data. if notes are missing, cap confidence at "medium".

2. assess statement freshness and data quality (explicit inputs/outputs).

   • input: statement as-of date, valuation date.
   • output: freshness flag (current, stale, very stale) and list of missing or incomplete line items.
   • calculate: days between statement date and valuation date. flag if >90 days.
   • enumerate gaps: which key income statement lines are missing (e.g., operating profit, tax rate)? which balance sheet items are missing (e.g., debt maturity, inventory breakdown)? is cash flow from operations stated, or must it be backed into from net income and working capital changes?

3. compute or collect key ratios (explicit inputs/outputs).

   • input: income statement, balance sheet, cash flow statement line items.
   • output: calculated or sourced ratios for steps 4-6.
   • compute (if raw data available): current ratio (current assets / current liabilities), quick ratio ((current assets - inventory) / current liabilities), debt-to-equity (total debt / total equity), net debt-to-ebitda, interest coverage (ebit / interest expense), receivables turnover (revenue / average receivables), inventory turnover (cogs / average inventory), days sales outstanding (365 / receivables turnover), days inventory outstanding (365 / inventory turnover), cash conversion cycle (dso + dio - dpo, days payable outstanding = 365 / (cogs / average payables)).
   • if ratios are pre-supplied, validate against raw data.

4. examine cash conversion and accrual risk (explicit inputs/outputs).

   • input: net income, operating cash flow, free cash flow (if available), accruals (non-cash items), depreciation, amortization, stock-based compensation, deferred tax, changes in working capital.
   • output: cash-conversion score (strong/moderate/weak) and list of accrual concerns.
   • calculate: operating cash flow / net income ratio across available periods. if <0.8 or >1.3, flag as potential concern. calculate free cash flow / net income ratio.
   • flag patterns: net income growing but operating cash flow flat or declining. margin expansion without corresponding cash flow support. large non-cash items (depreciation, amortization, stock comp, deferred tax) that inflate earnings relative to cash. unusual working capital swings (e.g., receivables or inventory growing faster than revenue).
   • capitalization concern: are operating expenses being capitalized (e.g., as inventory or intangible assets) rather than expensed? review balance sheet composition for this.

5. assess working capital and liquidity pressure (explicit inputs/outputs).

   • input: current assets, current liabilities, receivables, inventory, payables, operating cash flow, cash balance.
   • output: working capital pressure (low/moderate/high) and list of liquidity signals.
   • calculate: current ratio trend. if <1.0 or declining sharply, flag. cash conversion cycle trend. if lengthening, flag strain on liquidity.
   • flag signals: receivables or inventory growing faster than revenue (extension risk). payables declining or static while inventory/receivables rise (working capital trap). cash balance shrinking. operating cash flow insufficient to cover capital expenditure plus debt service (refinancing risk).

6. examine leverage, debt maturity, and refinancing risk (explicit inputs/outputs).

   • input: total debt (short-term and long-term), equity, ebit/ebitda, interest expense, cash flow from financing, debt maturity schedule (if available).
   • output: leverage tier (low/moderate/high) and refinancing-risk flag.
   • calculate: gross debt-to-ebitda, net debt-to-ebitda (if cash available), interest coverage (ebit / interest), debt service ratio (operating cash flow / (interest + short-term debt due)).
   • flag concerns: debt-to-ebitda >4x. interest coverage <2x. material debt due in next 12 months relative to operating cash flow and cash balance. rising interest rates or refinancing spreads affecting cost of debt. covenant risk (if notes disclose covenants). negative or insufficient operating cash flow to service debt.

7. identify and categorize one-off and non-recurring items (explicit inputs/outputs).

   • input: optional notes on one-offs, income statement detail on operating vs. non-operating items, tax notes, FX gains/losses.
   • output: adjusted earnings (one-off items excluded), materiality of adjustments, list of one-offs by type.
   • extract from notes: asset sales and gains/losses, impairments, restructuring charges, legal settlements, FX effects, tax adjustments (e.g., deferred tax impacts from rate changes), stock-based compensation (if not already in operating expense), acquisition-related costs.
   • quantify: sum of one-offs as % of net income. if >20%, flag as material. if >50%, flag as high uncertainty in normalized earnings.
   • categorize: recurring vs. non-recurring. separate pre-tax and tax effects.

8. flag governance, accounting, and auditor red flags (explicit inputs/outputs).

   • input: notes on accounting policy changes, related-party transactions, auditor concerns, restatements, management estimates and assumptions, revenue recognition policy.
   • output: list of red flags by severity (critical/moderate/minor), summary of accounting quality signals.
   • flag: related-party transactions or balances material to results (>5% of net income or assets). accounting policy changes year-over-year without clear business justification. auditor qualifications, going-concern warnings, or restatements in last 3 years. unusually aggressive revenue recognition, capitalization, or reserve policies. management estimates with high sensitivity to assumptions (e.g., useful lives, reserve adequacy, allowance for doubtful accounts). unexplained note gaps (e.g., no debt maturity schedule, no contingency disclosure).

9. synthesize earnings quality score and confidence rating (explicit inputs/outputs).

   • input: outputs from steps 1-8 (freshness, accruals, working capital, leverage, one-offs, red flags).
   • output: earnings quality tier (high/medium/low), valuation confidence impact (positive/neutral/negative), summary narrative.
   • scoring logic: high tier requires fresh statements, complete data, strong cash conversion, limited accrual concerns, no material red flags, and notes available. medium tier allows incomplete notes or moderate concerns but no critical red flags. low tier indicates stale statements, missing data, significant accrual or leverage risk, or material red flags.
   • confidence impact: if high tier, earnings support valuation with lower discount rate. if medium tier, use baseline discount rate and sensitivity testing. if low tier, apply valuation haircut or require additional due diligence before conviction.

10. compose handoff bundle (explicit inputs/outputs).

    • input: all outputs from steps 1-9.
    • output: structured handoff bundle with exact markers (see output contract below).
    • format: JSON or YAML key-value pairs. include ticker, as_of_date, earnings_quality_tier, and all risk flags listed in output contract.

decision points

• if statement freshness fails (>90 days old): mark as stale; cap confidence at medium unless statements are otherwise strong. if >180 days old, cap at low confidence.
• if data gate fails (income statement, balance sheet, or cash flow missing entirely): continue with available data but explicitly list gaps and reduce confidence by one tier. if all three statements missing, stop and return "insufficient data" error.
• if operating cash flow is negative or not available: flag high accrual/quality risk. if available for multiple periods, check trend. if trending negative while net income is positive, escalate red flag to critical.
• if cash conversion ratio <0.8 for two or more consecutive periods: flag as accrual concern (earnings not cash-backed). if also accompanied by working capital deterioration, escalate to high leverage/refinancing risk.
• if current ratio <1.0 or debt-to-ebitda >4x: flag high liquidity or leverage risk. if both present, cap confidence at low unless other factors mitigate (e.g., strong ebitda growth).
• if notes are absent or incomplete: cap confidence at medium, even if financial statements are strong. if critical notes (debt maturity, related parties, one-offs) are missing and would materially affect interpretation, flag as data gap.
• if one-off items >20% of net income: mark as high earnings volatility. if >50%, note that normalized earnings are highly uncertain and require separate adjusted earnings calculation.
• if related-party transactions or management estimates appear aggressive: flag as governance concern. if accompanied by auditor qualification or prior restatement, escalate to critical red flag.
• if earnings quality is low but valuation is urgent: return medium confidence with strong caveats and recommend post-close verification or earn-out protection.

output contract

exact output sections (in this order):

1. Input Quality

   • statement freshness (current/stale/very stale, days old).
   • completeness of income statement, balance sheet, cash flow statement (present/partial/absent).
   • availability of notes (available/partial/absent).
   • list of missing or incomplete line items.

2. Earnings Quality Score

   • tier: high / medium / low.
   • basis: summary of key drivers (cash conversion, accrual patterns, leverage, red flags).

3. Cash Conversion and Accruals

   • operating cash flow / net income ratio (if available, multi-period trend).
   • free cash flow / net income ratio (if available).
   • accrual concerns: list of patterns or items flagged (e.g., "receivables growing 15% while revenue flat", "non-cash items 25% of net income").
   • capitalization concerns (if any).

4. Balance Sheet and Working Capital

   • current ratio trend.
   • cash conversion cycle (if computable).
   • working capital pressure: low / moderate / high.
   • liquidity signals: list of trends or concerns (e.g., "cash balance down 30%", "payables shrinking").

5. One-Offs and Accounting Red Flags

   • one-off items: list by type, amount, % of net income.
   • adjusted earnings (one-off items excluded).
   • governance and accounting red flags: list by severity (critical/moderate/minor).

6. Valuation Confidence Impact

   • confidence tier: high / medium / low.
   • rationale: how earnings quality affects discount rate, terminal growth rate, or valuation haircut.
   • recommendation: proceed as planned / use sensitivity testing / require additional due diligence.

7. Handoff Bundle

   • JSON or YAML block with exact markers:
     • ticker: entity identifier.
     • as_of_date: statement date (YYYY-MM-DD).
     • earnings_quality_tier: high / medium / low.
     • cash_conversion: strong / moderate / weak (or n/a if data unavailable).
     • accrual_risk: low / moderate / high.
     • working_capital_pressure: low / moderate / high.
     • leverage_refinancing_risk: low / moderate / high.
     • one_off_adjustments: materiality tier (low / moderate / high) and count of items.
     • red_flags: list of critical and moderate flags (empty list if none).
     • valuation_confidence_impact: positive / neutral / negative (and discount-rate or haircut delta, if applicable).
     • data_gaps: list of missing statements or critical notes (empty list if none).

data format:

• all numeric outputs (ratios, %, days) rounded to 1 decimal place.
• all date outputs in YYYY-MM-DD format.
• all lists and enumerations as bullet points or JSON arrays.
• all confidence/tier outputs capitalized and consistent (High/Medium/Low).

outcome signal

the skill has worked when:

• you receive a handoff bundle with all 10 required markers populated.
• earnings quality tier and valuation confidence impact are explicitly stated.
• data gaps, red flags, and accrual concerns are enumerated and ranked by severity.
• you can pass the handoff bundle directly to a valuation model step with confidence anchored to the tier (high confidence = lower discount rate or haircut; medium = baseline + sensitivity; low = halt or haircut).
• the output explains the difference between reported earnings and cash earnings, flagging where one is materially higher and why.
• you know whether you can trust the balance sheet and cash flow statement to inform working capital and capital structure assumptions in the valuation.