Public companies in the United States must publish an audited annual account of themselves. The requirement arrives through a regulatory filing rather than goodwill, and knowing exactly which document carries which information turns several hours of vague reading into about twenty focused minutes – plus it removes any reason to pay for a newsletter that summarises what you can read yourself for nothing.
Which document you’re actually holding
Companies send out glossy annual reports to shareholders because marketing departments enjoy doing that. Nothing legally requires the gloss. The obligation lives in the annual report on Form 10-K, filed with the regulator and available free through its electronic system and most companies’ investor-relations pages. Three related documents do most of the work: the 10-K for the year, a shorter unaudited quarterly equivalent filed after each quarter, and a brief filing used to announce significant events as they happen. Amendments appear with an “A” suffix, usually because something had to be corrected – always worth opening.
The 10-K is organised into numbered items, and their usefulness is wildly uneven. Roughly in order of value to someone building an opinion:
Audited financial statements and the notes. Balance sheet, income statement, cash flow statement, and the footnotes where the consequential accounting judgments get disclosed – how revenue gets recognised, what counts as a segment, stock-based pay assumptions, lease terms, contingent liabilities. Analysts read the notes first precisely because the statements themselves summarise them away. Each primary statement has its own explainer on this site: the income statement, the balance sheet, cash flow.
Management’s discussion and analysis. Management’s own explanation of why the numbers moved. This is where revenue growth gets attributed to volume, pricing or acquisitions, where margins are defended or explained away, where liquidity and capital commitments are quantified. It’s still written by the people being graded, so read it for causes they couldn’t conceal and discount the adjectives.
Risk factors. A long statutory list of terrible things that might happen. Mostly standardised; the signal hides in what’s specific and what changed since last year. New, detailed language around a customer concentration, a patent expiry, a funding requirement or a regulatory investigation means somebody inside decided that risk deserved words. Compare against competitors’ versions and boilerplate becomes obvious quickly.
Business overview and properties. What actually earns money, in the company’s own segmentation, plus where the physical assets are. Useful mostly as a reality check against how the shares are marketed.
Internal control disclosures. Whether management and the auditor believe reporting controls work. Terms worth learning to notice: a material weakness in internal control is a genuine warning sign, not a technicality.
Separate filing, separate purpose: the proxy statement covers executive pay, board composition and votes – the place to look if you want to know whether management is compensated for results or for attendance.
Reading between the documents
The shareholder letter attached to pretty annual reports rewards scepticism rather than avoidance. Everyone writes one; almost all of them describe innovation, resilience and opportunity while noting record execution. Its real value is longitudinal. Collect five years – they’re archived freely – and read them as an audit trail of promises: what did management say three years ago would matter, and did it happen? Executives who discussed a pending challenge candidly before it hurt results behave differently from those whose difficult years arrive as surprises caused exclusively by external conditions. That track record is one of the few genuine pieces of evidence available about management character, and it can’t be bought, only collected.
Two adjacent habits raise signal cheaply. First, check the auditor’s opinion. A clean, unmodified opinion means the accountants concluded the statements fairly present the position under the applicable framework; anything softer – modifications, emphasis paragraphs, explicit doubt about the entity’s ability to continue as a going concern – deserves your full attention well before you form any view on growth. Changes of auditor, or an auditor resignation, deserve reading up on. Second, use search instead of skimming. Opening a filing and looking for particular phrases beats agreeably scanning prose: going concern, material weakness, restatement, impairment, related party, segment, dilution, waiver, forbearance. Documents nobody reads tend not to hide their bad news well.
Skip press-release framing entirely. Earnings releases emphasise whichever figures suit the story and define “adjusted” in whatever way flatters – non-GAAP measures have legitimate uses and notorious abuses, and the reconciliation back to reported results is where the conversation happens. Anyone offering paid access to filings is selling formatting; regulators publish them, complete, for nothing, on the same day anyone else sees them.
Once you can locate the interesting parts without assistance, the interesting questions get sharper. If earnings rose, did cash rise alongside them, or is the difference sitting in receivables? Did shares outstanding increase while profit per share held, meaning growth was funded by issuing claims on itself? Does management’s stated strategy require spending that shows up anywhere in investing activities? Is the company’s disclosed exposure to one customer, supplier or geography consistent with the confident narrative in the letter? These are answerable in a sitting, from primary sources, by someone with no finance degree – and they beat almost every summary ever written about a company, including whatever you’d otherwise have subscribed to.
This is educational explanation of public disclosure documents, not investment advice or a recommendation to buy or sell any security; filings describe the past under accounting rules that involve judgment, summaries quoted anywhere can be stale or wrong, and even thorough reading cannot prevent losing money on a holding.