You have a ticker, an app with a buy button, and maybe an hour. The hard part is not finding numbers, since a P/E ratio takes ten seconds to look up. The hard part is the order, because each step decides whether the next one is worth your time. Here is a sequence you can actually follow.

Start With the Business, Not the Chart

Before any figure, answer two questions. What does this company sell, and who hands over the money for it?

Try saying it out loud in two sentences. An airline sells seats to travelers and gets paid when a ticket is booked. A software company charges businesses a monthly fee per employee. If you cannot do that, stop here. Every number later in the process only means something once you know what it is measuring.

Watch for companies that look like one business and earn like another. Some restaurant chains own very few of their restaurants and collect fees from the franchisees who own the rest. The logo tells you none of this. The filing does.

Where the Free Data Is, and What to Read First

Public companies file their results with the Securities and Exchange Commission, and every filing is free to read on EDGAR at sec.gov/edgar. No account, no subscription. The company's own investor relations page carries the same documents, usually with the earnings call beside them.

Two forms do most of the work: the 10-K is the annual report, and the 10-Q covers one quarter. The largest companies file the 10-K within 60 days of their year end, and smaller ones get up to 90 days.

Start with two parts of the 10-K. Item 1, Business, is the company describing itself in ordinary language: products, customers, competitors, employee count. Item 1A, Risk Factors, is what could go wrong, written to protect the company from later claims that it hid something. Much of it is boilerplate that would fit any company. Read for the specific ones instead: a single customer worth a quarter of revenue, a patent expiring, a lawsuit, a debt payment coming due.

One year of results tells you almost nothing. The direction over several years tells you a lot. Three lines are worth tracking.

Revenue says whether the business is growing. Margins, meaning profit as a percentage of revenue, say whether that growth costs more each year to produce. Share count says whether your slice of the company is shrinking or growing.

The third gets skipped, and it should not. Suppose revenue rises from $2.0 billion to $2.4 billion over four years, which reads as healthy 20% growth. If the share count went from 200 million to 260 million in the same stretch, revenue per share actually fell, from $10.00 to $9.23. New shares were issued, and every existing owner owns less.

The income statement in a 10-K usually shows three years side by side, so two filings give you six years of history in ten minutes.

Then Look at What the Company Owes

Now open the balance sheet and find two figures: total debt and cash. A company with $1 billion of debt and $3 billion of cash sits in a very different spot from one with those numbers reversed, even if the profits look identical.

Debt is not the problem by itself. Not being able to carry it is. One quick check is interest coverage, which is operating income divided by interest expense. At $800 million of operating income against $100 million of interest, the company covers its interest eight times over. At $120 million of operating income, it barely covers it once, and one bad year becomes a crisis.

The notes to the financial statements list when the debt comes due. Debt maturing in seven years is a different animal from debt maturing next March. Whether a company can pay this month's bills and whether it survives the decade are separate questions, which is the difference between liquidity and solvency.

Adjusted Numbers Are Not Reported Numbers

Companies publish two sets of figures. The official ones follow GAAP, the accounting rules every US public company has to use. Next to them sit adjusted figures, where management removes items it considers unusual.

Sometimes that is fair. A factory fire or a one-time legal settlement really does distort a year. Sometimes it is not. A cost excluded as unusual every year for six years running is not unusual, it is the business. Stock paid to employees is the usual example. It is real pay, it dilutes existing shareholders, and it is stripped out of adjusted profit all the time.

You do not have to guess which is happening. When a company publishes an adjusted number, SEC rules require it to show a reconciliation back to the closest official figure. That table is the list of what was taken out. Read it, then decide which version you believe.

Valuation Comes Last, and Always Against Something

Only now does the price matter. Valuation goes last because it is a judgment about everything above it, not a shortcut past it. A P/E ratio you look up before you know what the company sells is a number with nowhere to sit.

Whatever measure you use, use it against a reference point. Two work well: the company's own multiple over the past five or ten years, and the multiples of three or four direct competitors. "Twenty-two times earnings" means nothing on its own. "Twenty-two times earnings, against a ten-year average near 15 and rivals around 17" is a question worth chasing down.

Analyst price targets are not a shortcut either. A target is one person's forecast of where a stock might trade in roughly a year, built on assumptions you usually cannot see. Analysts covering the same company disagree with each other, sometimes by a lot, and they revise their targets as prices move. Treat a target as an opinion with reasoning behind it, and go looking for the reasoning.

Write Down What Would Have to Be True

Finish by writing two short paragraphs to yourself, dated.

The first is what has to happen for this to work out. Be specific enough to check later. "Revenue keeps growing around 10% a year and the operating margin stays near 25%" can be tested against next year's 10-K. "The company is well run" cannot.

The second is what would tell you that you were wrong. Name it now, while you have nothing at stake and no reason to argue. Two straight years of falling margins. Losing the customer who accounts for a third of revenue. Written down beforehand, these are tests. Invented afterwards, they turn into excuses.

Four Ways This Goes Wrong

Even a careful process bends in predictable directions.

Reading only the bullish case. If everyone you read already owns the stock, you have collected agreement, not evidence. Go find the strongest argument against it and see whether it survives the risk factors.

Mistaking a familiar product for a good business. You might love the coffee. That tells you nothing about margins, debt, or the price of the shares. Familiarity feels like knowledge.

Anchoring on the first price you saw. If you first noticed a stock at $80, then $52 feels cheap and $95 feels expensive. Both feelings come from the $80, which was just the price on the day you looked.

Treating a falling price as a bargain. A price falls for a reason, and sometimes the reason is that the business got worse faster than the price did. A lower number is only good news if the earnings behind it hold up. Work out why it fell before you decide it is on sale.