To analyze a company, in the dictionary's sense of the word, is to separate a complex whole into its parts so you can see its true nature and inner relationships. Applied to a firm, that means three questions: does the business earn more than it spends, is that advantage hard to copy, and do the people running it behave like stewards or like salesmen. You can answer all three from public records.
The framework below works for any firm, public or private, and for any reason you might invest time in one — as a shareholder, a job applicant, a supplier, or a partner. It asks for no special data access. It asks for patience, because the point is to build a cause-and-effect map before reaching for an opinion. The order matters: fundamentals first, moat second, management last. Most people run it backwards and fall for the pitch.
What do the fundamentals actually tell you?
Start with the accounts, not the story. A company's filings — the annual report, the audited statements, the notes — are its public record. Read the income statement to learn whether revenue is growing, flat, or shrinking. Read the balance sheet to learn who really owns the firm: how much debt sits against the assets, and when that debt comes due. Read the cash flow statement last, because it is the hardest to flatter.
Three checks do most of the work. First, compare profit to cash. Accounting profit rests on estimates; cash is a fact, and a wide, persistent gap between the two is a question worth pressing. Second, look at where the money goes. A firm that spends more on buybacks than on the plant, the product, or the people is telling you something about its own growth prospects. Third, read the notes. The footnotes carry the obligations that never reach the headline figures — leases, guarantees, litigation. The front of the report is written for you; the back is written about the company.
One caution. A single year proves little. Look for direction over several years, and compare the firm to others in its own industry, because a healthy margin for a grocer would be alarming for a software company.
Does the company have a moat — or just a moment?
A moat is whatever stops a competitor from doing the same thing at the same cost. The honest versions are narrow and checkable: a network that gets more valuable as more people join it, a switching cost that makes leaving painful, a cost position built on scale or location, a regulation that limits who may compete, or a brand that lets the firm charge more for the same goods.
The test is simple. If this company disappeared tomorrow, how long would it take a well-funded rival to rebuild what it had? Days suggest a moment. Years suggest a moat. Be especially wary of advantages that exist only in the management team's own description of itself. Vision is not a barrier to entry.
This is also where the cause-and-effect map earns its keep. Trace one concrete chain: a cost advantage lets the firm price lower, which holds volume, which funds the scale that keeps the cost advantage. If the loop does not close, what looked like a moat was a head start.
How do you judge management without meeting them?
You cannot interview your way to a view of management, and you should not try. Judge them by what they leave in writing. Read the last few years of shareholder letters side by side. Did the promises of year one appear in the results of year three? A letter that explains a missed target plainly is worth more than one that reframes every miss as a victory. For related coverage, see Regulatory Budgets: Did One-In,Two-Out Actually Cut Rules?.
Watch three specific things. Capital allocation: what did they do with the cash, and does the choice match the story? Compensation: is pay tied to measures the executives themselves chose, and do those measures move with the business or merely with the share price? And candor in bad years: the character of a management team shows up in the downturn, not the boom.
There is a government parallel worth borrowing. When a Measure Becomes a Target: Government Metrics and Gaming describes how any single number, once it determines rewards, gets managed rather than reported. The same logic applies to corporate scorecards. A management team paid on one metric will deliver that metric. Read the footnotes to find out what else moved. This connects to our earlier piece, When a Measure Becomes a Target: Government Metrics and Gaming.
What are the red flags worth memorizing?
Some warnings recur so often that they deserve a standing checklist. None of these items alone proves anything. A cluster of them is a strong reason to walk away.
- Revenue growth that consistently outruns cash collection.
- Frequent "adjusted" measures that exclude real, recurring costs.
- Auditor changes, late filings, or accounting opinions with qualifications.
- Executives selling shares while telling investors to buy.
- Acquisitions that arrive faster than the explanations for them.
- A strategy that changes shape every time the last one stops working.
Notice what is absent from this list: the product you personally like. Liking a product is a fine reason to start looking. It is not evidence about the firm. Plenty of beloved products have sat inside badly run companies, and plenty of dull ones inside excellent ones.
What this means for your own process
Our analysis reduces to a sequence. Build the financial picture from the filings. Test the claimed advantage against the disappearance test. Read management's own record against its own words. Then, and only then, form a view. If the view survives contact with the checklist, you have earned an opinion. If it does not, you have saved yourself time you would otherwise have spent learning the same lesson expensively.
Two habits make the framework stick. Write your reasoning down before you commit — a short memo stating what you expect the company to do, so future-you can check it. And set a review date. Companies change; a conclusion drawn once and never revisited is not analysis, it is a souvenir. For more of this publication's work in this register, see the rest of our analysis coverage, and our business news section for the shorter, faster-moving pieces.
None of this guarantees an outcome. It narrows the field of things that can go wrong that you failed to look at — which is, in the end, most of what careful analysis can offer anyone.
