Financial statements

Definition

The three statements — income statement (accrual-based performance over a period), balance sheet (positions at an instant), and cash-flow statement (cash movements reconciling the two) — are the audited factual substrate of every fundamental thesis. The platform’s reading rule: the three statements are one system; every number in one has a counterpart in the others, and theses live in the RECONCILIATION (earnings vs cash, growth vs capital consumed), not in any single line.

How it works / structure

• The filing stack (engine-executable sources): 10-K (annual, audited), 10-Q (quarterly, reviewed), 8-K (material events between), proxy (compensation/governance) — all on EDGAR with structured XBRL data; the press release’s “adjusted” figures are marketing until reconciled to the filed GAAP numbers.
• Income statement: revenue recognition drives everything (accrual timing is where discretion lives — fa-earnings-quality); the margin cascade (gross → operating → net) locates where a business model earns and bleeds.
• Balance sheet: working capital (the operating cycle’s cash absorption), debt structure AND maturities (the wall that matters in stress), goodwill (acquisitions’ residue — impairment risk), and equity mechanics (share count changes reconcile to buybacks/dilution — event-buybacks).
• Cash flow statement: operating cash flow vs net income (the accrual gap — the single most information-dense reconciliation), capex (maintenance vs growth split is estimated, not stated), and financing flows (payouts, raises, debt).
• Footnotes and MD&A: where the bodies are buried — segment data, off-balance-sheet items, revenue-recognition policies, litigation.

When it applies

Every fundamental thesis (lens-fundamental builds here); the input layer for ratios (fa-ratio-analysis), quality screens (fa-earnings-quality), and valuation (fa-dcf-valuation); earnings-event grading (the print is measured against these baselines — event-earnings).

Risk profile & failure modes

• Adjusted-number capture: theses built on company-defined “adjusted EBITDA” inherit the company’s preferred exclusions; the platform reconciles to GAAP and treats the gap as information.
• Point-in-time illusion: balance sheets are quarter-end snapshots; window-dressing (repo 105-style) is documented history — trends across periods beat single snapshots.
• Restatement risk: filed numbers change; a thesis on aggressive accrual patterns should EXPECT restatements (fa-earnings-quality flags).
• Stale-data lag: filings arrive weeks after quarter-end; prices incorporate faster channels — the statements are for thesis QUALITY, not speed.

Evidence & limits

Filing requirements, timing, and content are SEC-regulated; XBRL structuring makes the data machine-readable (with known tagging noise). The statements’ informativeness for returns is the entire fundamental-analysis literature — mediated by the quality and valuation layers cited in their entries. Facts here; the return claims live downstream.

Falsifiable-thesis examples

Illustrations only, not signals:

• “X’s operating cash flow will cover its dividend and capex in each of the next four quarters (self-funding thesis)” — falsified by any uncovered quarter.
• “X’s inventory growth will fall back below revenue growth within two quarters (working-capital normalization)” — falsified by the filed ratios.

Cross-references

• The reading discipline: lens-fundamental
• Derived layers: fa-ratio-analysis, fa-earnings-quality, fa-dcf-valuation, fa-capital-allocation
• Event interface: event-earnings, fa-guidance-estimates