How to Analyze a Stock

The three questions that actually matter

Most investors begin with a stock screener. They filter for low P/E or high revenue growth, skim a few headlines, and call it research. What they've done is screening — a way to find companies worth investigating. It is not analysis.

Analysis is the work of building a mental model: understanding why a business earns the returns it does, what could undermine those returns, and whether the stock price gives you an edge or takes one away. Three questions do most of the heavy lifting.

These aren't sequential steps. You run all three simultaneously. A business can be high-quality and overpriced, or mediocre and cheap enough to work. The questions force you to separate the business from the stock. Most retail investors never make that separation cleanly, which is why they can be right about the company and still lose money.

Read the business before you read the financials

The most skipped step in stock research is understanding the actual business before opening a spreadsheet. Not the stock — the business. What does the company sell? Who buys it? Why do customers keep buying it, and how does the company make more money as it grows?

These questions are not answered by a P/E ratio.

Consider two companies with identical 15% revenue growth. The first sells enterprise compliance software on annual subscriptions. The second manufactures polypropylene. Both look the same on a revenue growth chart. The economics are nothing alike.

The software company's revenue is contractual. Once an enterprise deploys the platform across its legal and finance teams, switching costs — data migration, retraining, change management — make cancellation genuinely expensive. Gross margins run 75–80%. When the company adds a new feature, every existing customer gets it at near-zero incremental cost. The business gets structurally more valuable as it gets larger.

The polypropylene manufacturer lives in a different world. Raw material costs move with crude oil. New industry capacity can come online within 18–24 months of an investment decision, flooding the market and compressing margins. There's no lock-in — customers buy from whoever offers the lowest price on a given delivery date. Growth requires capital, and returns arrive only when the market isn't oversupplied. The income statement will tell you neither of those things.

A useful cross-check: look at customer concentration data in the 10-K (usually in the "Concentrations" note to the financial statements) and the cost structure breakdown on the income statement. High gross margins with a stable customer base and low capital intensity is the pattern underlying most durable compounders. You can see it before you build a single model. See also: How to Read a 10-K.

Five numbers that tell the story

There are roughly 40 financial ratios you could track for any stock. Most are noise for a first-pass analysis. Five metrics carry enough signal to tell you whether a business is getting stronger or weaker — and whether the current valuation makes sense. The key discipline with all five: look at the trend, not just the level.

Revenue growth rate — but the trajectory, not the number. A 20% growth rate in the most recent quarter means nothing in isolation. Was it 15% the quarter before? 25% the quarter before that? The direction matters more than the magnitude. Decelerating growth in a high-multiple stock is one of the most reliable signals that estimate cuts are coming. The market prices in continuation; when continuation breaks, corrections are fast and severe.

Warning sign: three consecutive quarters of slowing growth while management guidance stays flat. That divergence between reported results and forward guidance rarely resolves in the bull's favor.

Gross margin trend — expanding or compressing? Gross margin is the percentage of revenue left after direct production costs. It tells you whether the company has pricing power and whether scale is generating leverage. A software company at 75% gross margin that was at 78% the prior year is showing a compression signal worth investigating. It could mean pricing pressure from competition, rising cloud infrastructure costs, or a deliberate mix shift toward lower-margin services — each of which has a different investment implication.

Warning sign: gross margins falling in a business whose pitch is pricing power. One quarter is noise. Two or three consecutive quarters is a pattern the earnings call should address directly.

Free cash flow conversion. Net income is an accounting number. Free cash flow is the cash the company actually has to reinvest, pay down debt, or return to shareholders. Some businesses convert earnings to cash at 90%+; others — particularly capital-intensive ones or companies that extend credit to customers — convert at 50% or below. The calculation is simple: free cash flow divided by net income. Over 80% consistently is healthy. Under 60% consistently warrants an explanation. Sometimes it's benign (an inventory build ahead of strong demand). Sometimes receivables are climbing because customers are stretching payment terms — a warning sign that end-market demand is weaker than the income statement suggests. See also: Free Cash Flow Explained.

Net debt to EBITDA. This captures balance sheet risk. Net debt (total debt minus cash) divided by EBITDA (operating profit before depreciation and amortization). At 1x or below, the company has light leverage and real financial flexibility. At 4x–5x, a meaningful slowdown in business performance creates refinancing risk. At 6x or above, you're essentially owning a leveraged position on the business surviving its debt load — and the stock behaves accordingly. The most important context: whether this number is rising or falling year over year, and whether the business is cyclical. A 4x leverage ratio in a stable recurring-revenue business is different from the same ratio in an industrial company that just hit peak demand.

Forward P/E versus the stock's own historical average. The absolute level of a P/E multiple is nearly meaningless without context. A 22x P/E is cheap for a software company growing at 25% and expensive for a slow-growth industrial. The useful comparison is the stock's current forward P/E versus its own five-year average. A stock trading at a 30% discount to its historical multiple has something negative priced in — worth understanding why. A stock at a 40% premium to its historical multiple needs a clear story for why the next five years justify a higher valuation than any recent period. That story exists sometimes. Often it doesn't, and the premium compresses.

Build the bear case first

Most retail investors build the bull case and then search for evidence that it's correct. This is the wrong order. It's also, as a psychological matter, very hard to reverse once you've already spent an afternoon getting excited about a stock.

Before you buy, you need to be able to state the bear case clearly enough that a skeptic would call it fair. Not "the stock might go down if things don't work out." The specific scenario where your central assumption is wrong — the competitor who wins the enterprise contract, the margin structure that compresses when the company tries to expand internationally, the technology shift that makes the core product defensible for three more years but not ten.

Three patterns appear most often when retail investors miss the bear case.

The bear case is not pessimism. It is the specific fact pattern that would make you wrong. If you can't name it concretely, you don't have a thesis — you have a feeling. Feelings do not have a position size.

The price question

Every stock price is an implicit forecast. The current price represents the market's aggregate judgment of what the business is worth based on future cash flows, growth rates, and risk. Your job is not to calculate intrinsic value in some absolute sense. It's to understand what the current price is already assuming — and whether that assumption is realistic given the company's own track record.

A stock trading at 35x forward earnings when the market average is 18x is pricing in something specific: higher-than-average growth, lower-than-average risk, or both. The question is whether the company's history supports that premium. Look at actual revenue and earnings growth over the last four or five years. If the stock is at a 40% premium to its own historical average multiple, accelerating growth needs to be the story — not continuation of current trends.

Putting it together: the one-paragraph test

There's a practical test for whether your analysis is complete: write your thesis in one paragraph, in plain English. No bullet points. No jargon you'd have to define. A paragraph you could read aloud to someone who has never analyzed a stock and have them understand the bet.

Here's what a bad paragraph looks like:

That paragraph contains zero usable information. It doesn't say what the business does, why it earns the returns it does, what could go wrong, or what the current price is already assuming. It is a feeling wearing the costume of a thesis.

Here's what a good paragraph looks like:

The second paragraph has a business model, a retention mechanism, key financial trends, a valuation anchor, and a specific falsifiable condition with a checkable metric. That's the shape of a complete thesis. If any of those elements is still fuzzy, the work isn't finished. See also: How to Read an Earnings Report.