Every constraint that protects you also forecloses something. This one is worth stating plainly rather than leaving for you to discover.
The model sleeve runs inside a drawdown envelope. Exposure comes down before that envelope is threatened — not after. That single constraint is what makes the record survivable: it is designed to compound without ever frightening you out of the seat at precisely the moment staying in it matters most.
Most managers stop there, because it sounds like nothing but virtue. It isn't.
A position capable of returning five times your money is, almost without exception, a position that falls forty percent somewhere along the way. Inside a ten percent envelope, that position cannot be carried — not because the thesis is wrong, but because holding it would put the envelope at risk. So the engine does not carry it.
That is a real cost, and it is paid in the returns you never see. You should be able to check the claim rather than take it on faith, so the constraint is visible in the record itself:
10%
Drawdown Envelope
Worst observed 4.58% · 5.42 points of headroom
20
Median Hold, Days
The envelope, not the thesis, is what usually ends a position.
106
Longest Hold, Days
Also the single best position in the book.
The longest hold produced the largest gain. That is one position out of 16 — an illustration, not a rule, and it would be dishonest to present it as evidence of one. But it is a clean picture of what the envelope forecloses: return that requires both time and tolerance for decline along the way.
A mandate governs a sleeve.
It does not govern a net worth.
This is the reason the research exists at all. The constraint that protects the managed sleeve is not a constraint on your thinking, your horizon, or the rest of your balance sheet — and the ideas it rules out do not stop being interesting simply because one mandate cannot hold them.
Inside the mandate
The Core Sleeve
Drawdown-capped and exposure-managed. Positions are sized so that being wrong is survivable, and cut when the envelope demands it rather than when conviction fades.
Tracked daily and published in full — entries, exits, and the reasoning behind both.
Outside the mandate
Long-Horizon Research
Theses measured in years, where a deep interim decline is the price of admission rather than a signal to exit. The sleeve structurally cannot hold these. The research covers them anyway, and says plainly that it cannot hold them.
Every idea carries its horizon and its mandate status on the idea itself — not buried in a footnote.
Where a thesis needs more room than the envelope allows, the research says so, states roughly how much interim drawdown the thesis assumes, and leaves the decision where it belongs — with you.
Horizon and mandate-status labels describe the research and the constraints of the model sleeve. They are not advice, a recommendation, or a suitability assessment, and nothing on this site is tailored to any individual's circumstances, holdings, risk tolerance, or objectives. The model portfolio may not hold positions the research covers. The drawdown envelope is a design constraint of a hypothetical model rather than a guarantee — markets can gap through any intended exit, and a rule constrains risk only when it is actually executed. Past performance — hypothetical, modelled, or actual — does not guarantee future results.