Why order matters
Most stock research goes wrong not because the investor looked at the wrong data, but because they looked at it in the wrong order. The typical path is: hear about a company, look at the chart, decide it looks cheap or looks like it's running, then go find reasons. By the time the fundamentals get examined, the conclusion is already set and the research becomes a search for confirmation.
The order below is deliberately inconvenient. It puts the business first and the price fourth, which means you spend the first half of the process without any idea whether you want to buy. That is the point.
1. What does this business actually do?
Write one sentence, without jargon, describing how the company turns activity into money. If you cannot, you do not yet know enough to own it.
This sounds trivial and it is not. "Cloud infrastructure" is not an answer; "rents computing capacity by the hour to companies that would rather not run their own data centers, billed monthly" is. The test is whether your sentence would let someone else predict what would hurt the company.
Where to find it: the Business section (Item 1) of the most recent 10-K, and the company's own investor-relations page. Skip the marketing site — it describes the product, not the business model.
While you are there, note two things: who the customers are (a few large ones, or millions of small ones — this determines how quickly revenue can vanish) and what the company must keep spending on to stay where it is.
2. Does it make money, and is that improving?
Three years of numbers is enough for a first pass. You are looking for direction, not precision.
- Revenue — growing, flat, or shrinking? A single good year means little.
- Gross margin — what fraction of each dollar of sales survives the direct cost of delivering it. Stable or rising margins usually mean pricing power. Falling margins in a growing company often mean growth is being bought.
- Operating income — profit from the actual business, before financing and tax effects.
- Free cash flow — cash left after the spending required to keep operating. This is harder to flatter than earnings, which is why it is worth more attention than earnings.
A company can grow revenue for years while consuming cash. That is not automatically bad — it is how most infrastructure gets built — but it means the company depends on someone continuing to fund it, and that dependency is a risk you are underwriting.
3. What am I being asked to pay?
Valuation is not a verdict. It is a statement of what the market currently expects.
The price-to-earnings ratio is the usual starting point, and its usual misuse is comparing it across industries. A P/E of 30 is unremarkable for software and alarming for a utility. Two comparisons are worth more than an absolute number:
- Against the company's own history. Is it expensive relative to where it has traded?
- Against direct competitors. Not "the market" — companies with similar economics.
When earnings are negative or erratic, P/E returns nothing useful. Price-to-sales or enterprise value to free cash flow will carry more signal.
The question to hold in mind is not "is this cheap?" but "what has to happen for this price to make sense, and do I believe that?" A high multiple is a forecast of growth. You are deciding whether to accept the forecast.
4. What is the price doing?
Now, and only now, the chart — and for a limited purpose. Technical analysis will not tell you what a company is worth. It tells you what other people have been willing to pay recently, which is genuinely useful context and nothing more.
Three readings cover most of the value:
- Position in the 52-week range. Near the low, near the high, or mid-range. This is context, not a signal: cheap stocks get cheaper and strong stocks keep running, and the range tells you nothing about which is happening.
- Trend, via a long moving average. Whether price has generally been above or below its 200-day average tells you the direction of the prevailing opinion.
- Volume on large moves. A big move on heavy volume reflects broad participation. The same move on thin volume often reverses.
Resist reading more than this into a chart on a first pass. Pattern-matching on price is where confirmation bias does its best work.
5. Who else is buying or selling?
Two disclosures are public, free, and consistently underused.
Insider transactions (SEC Form 4). Executives and directors must report their trades within two business days. Buying is more informative than selling — there are many innocent reasons to sell and comparatively few to buy. Look for clusters rather than single trades. (What insider selling actually means covers how to read these properly.)
Institutional ownership (SEC Form 13F). Large managers report quarterly holdings. The data is delayed by up to 45 days, so it is history, not news. What it is good for is noticing a sustained change in direction across several quarters.
Neither is a recommendation. Both are people with more information than you, making decisions you can observe.
6. What would tell me I was wrong?
This is the step that gets skipped, and it is the one that does the most work.
Before you buy, write down the specific, observable conditions under which you would conclude your reasoning was mistaken — the invalidation conditions. Not a price — a fact about the business. "If gross margin falls below 40% for two consecutive quarters." "If the largest customer does not renew." "If free cash flow stays negative through next fiscal year."
Two things make this valuable. It forces you to state your thesis precisely enough to be falsifiable, which frequently reveals that you did not have one. And it gives you a pre-committed exit made while you were still thinking clearly, rather than one made in the middle of a drawdown.
A price stop is a risk-management tool and worth having. It is not the same thing. A stock can fall 30% while your thesis remains entirely intact, and it can rise 30% while your thesis quietly breaks.
Putting it together
The output of this process is not a buy or a sell. It is a written position: what the business does, what you are paying, what you expect, and what would change your mind. That document is the asset. The trade is downstream of it.
Two honest caveats. This framework will not tell you what a stock will do — nothing will, and treat confidently otherwise as a warning sign about the source. And it is a first pass: it filters out the obvious mistakes and surfaces the questions worth more work. That is a lower bar than "know the answer," and it is a realistic one.
If the six steps sound like a lot of reading, that is because they are. Doing them badly and quickly is worse than not doing them: it produces the confidence without the information.