A paper portfolio can be graded two ways. The shallow way asks: did it go up? The useful way asks: was the reasoning any good, and did I update honestly when reality disagreed? This page states the questions I'm actually testing — each one falsifiable, each one designed so that being wrong teaches me something specific.
Question 1 — Do "turnaround" stories actually re-rate, or does the market stay skeptical longer than I expect?
Two of my eight positions are explicit turnaround bets — Nike and Starbucks — and I bought McDonald's deliberately as the mature comp: a stable operator to measure them against. That gives me a small natural experiment already sitting in the portfolio. If the turnaround theses are right, NKE and SBUX should outperform the steady comp as the operational fixes show up in the numbers. If the market is right to stay skeptical, they won't — and the comp will quietly win.
What being wrong looks like: the turnaround names lag MCD over the tracking window even after the "fixes" are supposedly in place. That would tell me I mistook a good story for a good setup — a specific, nameable error.
Question 2 — How does having real money on the line change my judgment?
Writing my first update, I noticed the hardest part wasn't the research — it was placing the trades. Once money is committed, the goal quietly shifts from "understand the company" to "don't lose money," and that pressure changes how I read the same facts. So the second question is behavioral: does ownership make me hold losers too long, sell winners too early, or read news more defensively than I did before I bought?
How I'll test it: the trade journal records the reasoning at the moment of each decision. Reviewing those entries later — against what actually happened — turns my own behavior into a dataset I can inspect for the classic biases instead of just reading about them.
Question 3 — Is my tilt toward consumer brands an edge or a bias?
Five of my eight positions are Consumer Discretionary — companies whose products I encounter and understand directly. There's a respectable case for that (know what you own) and a well-documented trap (familiarity bias: confusing "I recognize this brand" with "I have an analytical edge here"). I don't yet know which one I'm doing, so the question is whether my consumer picks actually outperform my out-of-comfort-zone picks (Visa, Apple) — or whether the comfort is just comfort.
The rule the whole project runs on
Every position has to have a one-sentence thesis and a named risk before it goes on the board — no picks first, reasoning after. That rule is what makes the Thesis Scorecard possible: each claim is written down in advance, so it can later be graded honestly as held, broke, or still open. The point isn't a perfect record. It's a legible one.
Track the answers as they come in: the Thesis Scorecard grades each claim, the Trade Journal is the decision record, and the performance tracker is the raw data behind Question 1.