The 1R cushion: what the banked half is really for.

The stated function

Early monetization stabilizes the week.

A branch whose outcomes depend on trail behavior produces a lumpy weekly R distribution: a few large captures, several small trail-outs, and the occasional runner that reverses before the 1.6R unlock. Banking 50% at 1R converts a meaningful share of that lumpiness into settled, deterministic contribution. The week's floor rises. In gate terms this matters directly — a branch that can post a moderately positive week during a mediocre trend regime keeps the account further from the drawdown bands where variant aggression starts getting withdrawn.

The unstated function

The cushion is what makes the runner tolerable to hold.

The manuals put this bluntly when discussing the Time-Aggressive variant: delaying the partial removes the psychological cushion that makes Trend Partial stable. That is a statement about the operator, not the market, and it belongs in the design rather than being dismissed as soft. An operator watching a full-size position round-trip from 1.8R to break-even behaves differently from one watching a half-size position do the same thing, having already banked 1R. The second operator can let the trail work. The first starts intervening — tightening stops manually, closing early, or overriding the structure — and every one of those interventions is an adherence failure that corrupts the branch's evidence as well as its returns.

The variant ladder

Three cushion sizes, three different branches in practice.

Trend Partial's variants are best understood as adjustments to cushion size and timing rather than as adjustments to aggression in the abstract. The Exposure-Conservative variant adds a second 25% partial at 1.4R, thickening the cushion for regimes where 1.4R is frequently reached but the runner conversion is weakening. The Time-Aggressive variant delays the first partial to 1.3R, thinning the cushion to preserve a larger free-roll into the 1.6R unlock. Standard sits between them. The manuals are explicit that Exposure-Aggressive is the safer of the two aggressive postures precisely because it keeps the first monetization event intact — it takes size from the banked portion, not timing from it.

When the cushion is wrong

Over-cushioning turns a trend branch into an over-defensive scalp.

The failure mode runs both ways, and the conservative direction is the one that hides. An operator who habitually selects Exposure-Conservative or Time-Conservative on Trend Partial produces steady, unremarkable weeks and a branch that has quietly stopped doing its job — the fat-tail capture that justified the branch's 33.75% weight never materializes, and the blended EV slowly converges on something a heavier Normal allocation would have produced with less complexity. The manuals name this directly: over-defensive variant selection can turn a trend branch into a scalp. The variant usage audit exists to catch it, because it is invisible in weekly P&L and obvious in variant attribution.

• Time-Conservative is correct when trend is valid but fragile, volatility is chaotic, or ATR is stalling.
• It is wrong in strong expansion regimes with high MFE potential — and that misuse looks like prudence.
• Default to Standard when the matrix is mixed or contradictory. Do not invent conviction in either direction.

The gate interaction

Cushion preference should follow account state, not personality.

Most operators develop a habitual cushion setting and carry it across every regime and every gate — which is the one thing the variant layer exists to prevent. The gate permission matrix already biases the answer: Growth leans toward Standard and the aggressive expressions, Recovery toward Standard and Exposure-Conservative, Buffer toward Exposure-Conservative with Time-Aggressive usually blocked, and Floor toward the conservative variants only. An operator whose Trend Partial variant distribution looks identical in Growth and in Buffer is not being consistent. They are ignoring the layer that was supposed to adjust them, and the variant usage audit reads that flat distribution as a finding.