How to read a 10-K for a migration signal

An annual report is an argument, and it is arranged in the order that makes the argument persuasive. If you read it front to back you will meet the strategy narrative before you meet the numbers that constrain it, which is exactly backwards for judging whether a business is genuinely moving up the value chain. The fix is a different reading order: establish where the money actually is, find out what the lawyers are worried about, only then read what management says it is doing, and finish by checking whether the cash flow statement agrees. Four passes, and the fourth is the one that settles most arguments.

Pass one: the segment footnote

Start in the financial statements, at the segment disclosure. It is the only audited place where revenue and operating profit are split along the lines management genuinely uses to run the company, and it is where a mix shift becomes visible before any narrative admits to it. Read at least five years of it side by side, and look for three things: whether the higher-value segment’s share of revenue is rising, whether its profitability is diverging from the rest, and whether the segment definitions themselves have changed.

That third point deserves attention. Reporting segments follow how the chief operating decision maker reviews results, so when internal reporting is reorganised, the external segments change too. Sometimes that reflects a real change in how the business is run. Sometimes it makes an unflattering line disappear into a larger one. Either way, check whether prior periods were restated on the new basis — a mix trend computed across a definition change is not a trend, it is an artefact.

Pass two: the risk factors

Risk factors are drafted to limit liability, not to sell a story, which makes them the most literal prose in the document. For a migration read they are unusually informative, because a company transitioning into higher-value work has to disclose the specific hazards of that transition. Look for language about qualification timelines with customers, permits and consents not yet obtained, dependence on a small number of customers for the new product line, single-source inputs, and intellectual property that is licensed rather than owned.

Two patterns are worth flagging. New risk factors appearing around a specific facility, approval or customer programme tend to indicate that the programme has become material enough to require disclosure — a useful, slightly grudging confirmation that something real is happening. Conversely, a strategy section describing a transformation that leaves no trace at all in the risk factors is a mismatch worth explaining.

Pass three: the business and MD&A narrative

Only now read the description of the business and management’s discussion. By this point you know what the segments say and what the lawyers flagged, so the narrative can be read as a set of claims to be checked rather than as information to be absorbed. The useful discipline is to extract each specific, falsifiable claim — a target mix, a capacity date, a named programme — and write it down with the date it was made.

That list becomes the input to two other things we do. It feeds the say/do credibility ledger, because next year’s report will either honour these claims or quietly drop them. And it populates the first rung of the triangulation chain, which then asks whether each claim has been approved, funded, permitted and protected, or is still sitting where it started.

Pass four: the cash flow statement and the property footnote

The narrative describes intent; the cash flow statement records behaviour. Capital expenditure is reported as a single figure, and there is no reporting category that separates growth investment from replacement, so the split management quotes is a characterisation rather than a disclosure. What you can do is look at the relationship over time. Capital spending that runs persistently above depreciation suggests the asset base is being expanded rather than maintained; spending that tracks depreciation year after year suggests a business standing still, whatever the strategy section says.

The property, plant and equipment footnote adds resolution. Construction in progress is capital already committed to something not yet running, so a rising balance is a forward indicator that the narrative has not caught up with. Research spending is worth the same treatment: consistent research investment across a cycle is evidence of manufactured differentiation, while research that gets cut whenever cash is tight tells you differentiation is being rented.

A worked checklist

Run these in order, and record the page or note number for each answer rather than the answer alone.

• Five-year segment table. Has the higher-value segment’s revenue share risen, and has its margin diverged from the rest of the group?
• Segment definitions. Did they change during the period, and were prior years restated on the new basis?
• Risk factors, year over year. Which are new, and do the new ones correspond to a named programme, facility or customer?
• Customer concentration. Is any single customer large enough to require disclosure, and is the new higher-value revenue more concentrated than the old?
• Claims list. Every specific, dated, falsifiable statement about the transition, copied verbatim.
• Last year’s claims. Which ones were repeated, which were quietly restated, and which vanished?
• Capex versus depreciation. Sustained above, roughly equal, or below?
• Construction in progress. Rising, and attributable to anything named in the narrative?
• Research spending across the cycle. Steady, or cut when cash was tight?
• Backlog and long-term agreements. Disclosed, and moving in the direction the story requires?

Ten answers with citations is a defensible starting read. Ten answers from memory is a summary of the company’s own marketing.

Where this method runs out

A 10-K is an annual snapshot of a company that changes quarterly, so the picture is always somewhat stale — quarterly filings and current reports fill the gaps, and transcripts carry the specific commitments that annual reports sand down. More importantly, disclosure quality varies enormously. Some companies report three richly detailed segments; others report one, which makes a mix shift almost unreadable from the outside and forces the whole analysis onto weaker evidence.

The approach is also confined to what has already been filed. It is good at establishing whether a transition is underway and poorly suited to anticipating one that has not yet reached the page. And it is US-specific in its vocabulary: the same logic travels, but the document names, item structure and disclosure thresholds differ market by market, which is part of why our coverage is sequenced rather than applied uniformly everywhere at once.

Finally, none of this produces a valuation. It tells you whether a business is climbing. What that climb is worth, and whether the market has already paid for it, is a separate question.

The bottom line

Read the segment footnote first, the risk factors second, the narrative third and the cash flow statement last, and an annual report stops being a persuasive document and becomes a checkable one. Every claim ends up attached to a number, a date, and a place in the filing where you found it — which is the standard the rest of our method is built on, and the reason every read in the research book carries its citations rather than its conclusions alone. The vocabulary used above is defined term by term in the glossary.