How to build an investment thesis you can test

Separate the fact from your interpretation#

Suppose a fictional manufacturer reports that its service revenue grew while equipment sales fell. That is an observation. “Recurring service income may make earnings less dependent on new equipment orders” is an interpretation to test.

You would still need to examine service margins, customer retention, contracts and the size of that activity. One growing segment does not automatically change the economics of the whole company.

Use a table or separate sentences so estimates and opinions cannot quietly become reported facts.

Write the mechanism#

Explain how the proposed change could reach profit, cash and each share. For the fictional manufacturer, more service work might improve revenue stability, but only if it earns useful margins and does not require disproportionate investment.

Then consider financing and dilution. A successful expansion can create value for the company while producing a smaller benefit per share if substantial new equity is required. See revenue per share and company debt checks.

Include the price you are evaluating#

A business thesis is incomplete as an investment case without valuation. Record the date, price source, relevant financial period and assumptions used in the estimate.

Do not call a company undervalued solely because your favourable scenario exceeds the market price. Show a weaker case and explain which assumptions drive the difference. The valuation walkthrough demonstrates an explicit model.

Build a one-page record#

You can keep this in a document or spreadsheet. It does not depend on a paid research platform or an elaborate scoring system.

Make the counter-case specific#

“The stock could go down” is a possible outcome, not a competing business explanation. For the service example, a stronger counter-case is that service growth reflects temporary repairs and will fade as the installed equipment base shrinks.

That counter-case suggests evidence to inspect: the installed base, recurring contract revenue and demand for new equipment. It helps you investigate instead of merely defending the original idea.

Update without erasing the old reasoning#

After a new report, append the date, new facts and how they affect the thesis. Say whether the evidence strengthens, weakens or complicates the view. If the original mechanism no longer holds, do not quietly replace it with an unrelated reason to remain interested.

A correct share-price prediction is not proof of a well-supported thesis, and a short-term loss does not prove every observation was wrong. Review the reasoning against the evidence that was available, as well as the eventual outcome.