Claim, Evidence, Caveat: Structuring the Argument

A research write-up that opens with methodology and works its way toward a conclusion at the end makes the reader do the work of figuring out what the actual finding is. Flipping the order — claim first, then evidence, then caveat — puts the reader in a much better position: they know from sentence one what's being argued, so everything that follows can be evaluated against it instead of held in suspense.

The claim is one or two plain sentences stating what was found, in terms someone could act on: "a signal built from analyst estimate revisions predicts one-month-ahead returns, concentrated in mid-cap names." Not "we investigate whether analyst revisions have predictive power" — that's a research question, not a claim, and it tells the reader nothing about the answer. The evidence is the specific support: the backtest results, the statistical significance, the robustness checks, presented in enough detail that a skeptical reader could see how the claim was arrived at rather than being asked to take it on faith. The caveat is the honest statement of where the claim is weakest — a short sample period, a result sensitive to one parameter choice, a strategy that hasn't been tested through a full market cycle, transaction costs that were estimated rather than measured. Putting the caveat last isn't burying it; it's the natural place for it once the claim and evidence have given the reader context to understand what the caveat actually limits.

The structure works because it matches how a skeptical reader actually evaluates a claim: they want to know what's being asserted, then whether it's supported, then what would make them doubt it. Writing in that order does the reader's evaluation work for them instead of forcing them to reconstruct it from a narrative that reveals the conclusion only at the end.

The same structure scales down to a single paragraph or up to a full report — a slide can state a claim, show one supporting chart, and end with a caveat bullet, while a twenty-page note can devote a section to each. What stays constant is the order: claim first so the reader knows what to evaluate, evidence second so they can judge whether it holds up, caveat last so they know exactly how far to trust it once they've seen the support behind it.