INV-04 · Thesis from public evidence

Write an investment thesis for one sector using only cited public sources. What would have to be true for the thesis to be right, and what range do you put on each assumption?

Public sources only: public company filings (https://www.sec.gov/edgar/search/), government statistics, and industry association data, each cited with source and date accessed. The True North metric standard for metric definitions (https://www.truenorth.guru/standard). The division-approved AI tool for USE: research help only, with every figure it supplies traced to a source before it counts.

Market-size headlines with no path to a customer. A model-supplied statistic with no source. Assumptions stated as single numbers. A thesis that is really a list of trends. Naming specific companies to buy.

A thesis memo: the sector and the claim in one sentence; three to five load-bearing assumptions, each with a low, base, and high value and its source; the evidence that would disprove the thesis; an AI log showing which model outputs were kept, changed, or discarded; and a statement that the memo analyzes a sector and recommends no security.

CT: separating evidence from narrative. CR: framing a thesis that could be wrong. CO: a peer argues the opposite thesis from the same sources. CM: a memo an investment committee could read in five minutes. CZ: who loses if the thesis is wrong and someone acts on it.

An analyst or economist reviews the assumptions table. Standing instruction: reject any figure without a source and date. School-supervised.

R1: one sector, one claim. R2: every figure cited. R3: comparator is the thesis without AI help — what did the model add? R4: assumption ranges and disproving evidence named. R5: survives a skeptic. R6: no security recommended; limits stated.

(a) A confident thesis built on uncited model output. (b) A careful thesis that ends in "buy these three stocks."

Lane A (economics, business, or finance capstone). No verified credit claim.