Our method

Define what you are trying to understand and the period you are examining. A clear question makes it easier to recognize which information is relevant.

Look for the original source. Record the publication date, reporting period, units, definitions, and material limitations. Keep reported facts separate from estimates and your own calculations.

Every model leaves something out. State what must be true for a calculation or conclusion to hold, and examine a different set of inputs.

Connect the evidence to the conclusion. Include the strongest contrary evidence and the uncertainty that remains. An attractive asset and strong evidence are different questions; neither should be hidden inside an unexplained score.

A conclusion belongs to a particular body of evidence at a particular time. Record the developments or observations that would justify a fresh assessment.

Our editorial standard requires documented sources, visible assumptions, and appropriate review. Sensitive investment analysis requires a qualified independent reviewer before publication. The presence of a template does not establish that a review has occurred.

Material AI assistance is disclosed. Any review statement describes work that actually happened.

If you find a material error, send the page URL, the issue, and a supporting source. A correction should explain the original error, the change, and its effect on the conclusion.

Send us the page, the question, and the context. We welcome feedback and corrections.