The true accelerator: why some trades need room to become outliers.

The founding argument

Partial profit-taking is a tax on the outcomes that matter most.

Every partial is a decision to convert an uncertain large outcome into a certain small one. Applied to the median trade, that is sound risk management. Applied to the rare trade that was going to run 5R, it is expensive — and the problem is that the two are indistinguishable at 1R. Trend No-Partial's answer is to accept a worse median outcome in exchange for an intact right tail on a small, governed subset of trades. Not because the median doesn't matter, but because the other three branches are already optimizing it and something in the system has to be positioned to capture the trades that pay for the flat stretches.

Arithmetic versus geometric

The tail is not a bonus on top of the mean — for this branch it is the mean.

A branch whose outcome distribution is dominated by a handful of large winners has a mean that is almost entirely determined by those winners. Remove the top few trades from a TNP sample and what remains is roughly break-even after costs. This is not a defect to be engineered away; it is the branch's shape, and it has a direct consequence for how the branch is judged. Small-sample reads on TNP are close to meaningless, because the sample either contains its outliers or it doesn't, and which of those happened has more to do with when you drew the boundary than with whether the branch is working.

The evidence stack

Seven lenses, and the branch clears all of them or routes elsewhere.

Before a Trend No-Partial trade is permitted, the selection matrix runs seven lenses: branch EV support, gate permission, SDE alignment, clean ATR expansion, supportive market structure, execution-quality history, and a stable variance regime. No single lens is sufficient and any one of them can veto. A strong trend setup with a weak gate state is still constrained; a healthy account state with weak structure is still not permission to force aggression. The design intent is stated plainly in the manual — aggressive variants must be harder to qualify than conservative ones, because if they become a feeling-based override they will increase variance without improving expectancy.

• The 1.8R hit flag exists to test delayed trail activation — to prove, not assume, that the highest-variance management earns its keep.
• Execution red flags for this branch: low MFE, high MAE, poor survival to 1R and 1.6R, or emotional tightening that prevents fat-tail capture.
• Buffer gate blocks casual aggression; Floor and below block aggressive variants entirely.

The failure mode

Fat-tail logic becomes emotional acceleration the moment it's used without evidence.

That sentence is doctrine, and it describes a specific and recognizable sequence. An operator takes a loss, or several. The fat-tail branch is the one that can recover the deficit in a single trade. The reasoning that follows feels like conviction and is indistinguishable, from the inside, from the reasoning that accompanies a genuine A+ setup. What separates them is not the strength of the feeling but the state of the evidence stack — and the stack is checked before the trade rather than defended after it. This is the entire reason TNP's permissions are mechanical and the branch is quota-governed rather than left to judgment.

The boundary

What this branch is not permitted to become.

Trend No-Partial is a branch, not a posture available to the rest of the system. It does not license removing partials from Normal on a strong-looking day, extending Trend Partial's runner past its structure because momentum feels persistent, or treating Overflow as a place to express fat-tail conviction under a quieter label. Each of those is the same trade — full exposure without the evidence stack that authorizes full exposure — wearing a branch name that was designed for something else. Branch integrity is what keeps the attribution honest, and the moment fat-tail logic leaks across labels, every branch's expectancy estimate becomes a blend of things that were never supposed to be blended.