The Model Portfolio

Most research ends in a maybe.

The analyst walks you through the filings, builds the model, shows you the bull case and the bear case, and then leaves. Compelling risk-reward. Worth watching. One to keep on the radar. You finish reading and you still do not know what the author actually concluded.

I think that is a quiet form of cowardice. If the research cannot survive a decision, it was not research. It was content.

So every Research Dive here ends the same way: with a verdict. Enter, watch, or pass. And the verdict does not evaporate when the piece is published. It gets written down, with a date and a price on it, in the model portfolio.

Why a model book

Let me be precise about what this is, because precision here is the whole point.

The model portfolio is a paper book, and labeled as one. Tracked positions with dated entry prices, current values, and weights, run on a notional base of $100,000. It is not a record of real trades. My liquidity and my research run on different calendars, and the second should not wait for the first. Where I really do act in a name, I disclose that separately, in the edition, as my own action. The two are never blurred.

A model book is easy to fake, which is why most of them cannot be trusted. Entries appear after the stock has already moved. Losers quietly vanish. The record starts whenever the record looks good.

This one runs on three rules.

Every entry is priced on the day its dive publishes. Never before, never backfilled. The date on the entry is the date you could have read the reasoning.

Nothing vanishes. A wrong call stays on the sheet next to its arithmetic, and when the scoreboard moves against me, you read about it here, with the numbers of exactly how wrong.

Cash is a position. When I am mostly uninvested, the sheet says so, and the idle cash earns the T-bill rate the way an honest model book should. The whole thing is benchmarked against the S&P 500 from inception.

How positions are sized

Conviction is earned, not assumed.

Positions start small. A starter weight means the thesis is real but the catalyst has not proven itself, and I am paying a small tuition to sit in the front row while the dates arrive. As a thesis confirms, the position earns the right to grow. Concentration is the end state, never the starting point.

So the book opens flatter and wider than you might expect, with a healthy cash reserve. That is not indecision. That is the structure working as intended.

What paid subscribers get

The full book, live. Every position with its entry date, cost price, current value, unrealized profit or loss, and weight. Realized results kept separately when positions close. The benchmark line. And every dive from here forward ends with what the verdict did to the book.

Here is the live model book, every entry dated and priced on publish day, measured against the index from inception: