A new sister model to the intraday note. Here chfwd is the 6-trading-day forward return: deploy at t and capital is locked through the t+6 close — no rebalancing at t+1…t+5. To keep capital working under that lock we run a laddered ladder of six tranches (deploy ~1/6 each day, six rolling 6-day books). Over Feb 2022 → May 2026 the laddered rule earned +84.0% cumulative (15.5% CAGR, Sharpe 0.61) across 531 tranche entries at a 52.4% per-tranche win rate — and the universe + regime + probability gate lifts cumulative return by ~45 points over ungated laddering.
Two intraday findings flip over a 6-day hold: the gap rule inverts (big gaps mean-revert, so we now fade |gap|>2%), and the regime map develops two danger zones (deep-bear −3 to −2.5 and capitulation ≤ −4) around a sweet-spot bounce band (−4 to −3). Today the regime prints −4.13 — inside the capitulation danger zone — so the recommended sleeve is cash.
Recommended laddered strategy
15.5% CAGR
Sharpe 0.61 · maxDD −43.5% +84.0% cum · win 52.4% deployed 58% of sleeve-cycles
+45 pts vs no-gate ladder
◆ The headline: a laddered six-tranche book keeps capital working under the t+6 lock and compounds +84%
The t+6 lock freezes capital committed at t through the t+6 close — no rebalancing at t+1…t+5. The laddered design answers that by running six overlapping 6-day books: ~1/6 of capital is deployed each day into a fresh tranche, so the portfolio stays (almost) fully invested while every entry honours its own six-day hold. A single bad entry only ever risks ~1/6 of the book, and the universe + regime + probability gate lifts cumulative return from +39.5% (ungated) to +84.0%.
Net read: the t+6 constraint is not a reason to time harder — it is a reason to diversify across entry days. The studies below re-derive the gap and regime rules for this horizon and show where the current −4.1 tape leaves us (cash, by rule).
↻ Current sleeve = cash. The latest signal day (t=1045, 20 May 2026) ranks BRPT #1 (prob 0.672, in universe), but the regime score is −4.13 — inside the cap ≤ −4 band, where the deployable top-5 in-universe set has historically returned −3.47% over t+6 (win 41%, n=22). The laddered rule therefore parks this sleeve in cash and waits for the regime to lift back into the favourable −4 to −3 bounce band (+1.22% deployable) or to neutral.
How "today" works under the ladder. Each trading day fills exactly one of six sleeves with ~1/6 of the book. Five other sleeves remain in their existing 6-day holds. So the only decision today is the t=1045 sleeve: deploy into the day's qualifying in-universe picks, or hold cash. The most recent days that did qualify were 18 May (DSSA, TPIA at regime −3.94, inside the bounce band).
Position sizing — how much per name
Total book
100%
Rolling sleeves (6-day books)
6
Each day's sleeve
16.7% of book
Split across the day's qualifiers
1/vola × gap-mod
Today's sleeve (t=1045)
0% — cash
Each name's book weight = 16.7% × its share of that day's sleeve. If nothing passes the gate, the sleeve holds cash and the other five sleeves keep their existing 6-day holds.
Most recent active sleeve — per-name allocation
Latest ranked picks — last three signal days (top 5 by rank)
Universe filter — recent ana picks (t+6 model)
Universe list contains 282 tradeable tickers (hard pre-filter, step 0). The ana model ranks across 469 names; only are tradeable. Out-of-universe picks are shown greyed — visible, but never deployable.
Date
Rank
Ticker
Name
Prob1
Regime
In uni?
Alloc (% book)
Laddered call
Universe attrition is heavier on the ana model. Across the full history only ~48% of in-window top-5 picks are tradeable, falling to ~39% in 2026 — many of the model's strongest names (CYBR, BREN, SHIP, TINS, CUAN…) sit outside the 282-ticker list and are excluded regardless of rank. This is the same constraint as the intraday note, applied identically here.
★ The t+6 lock and why we ladder. Allocating at t commits capital through the t+6 close; no rebalancing is possible at t+1…t+5, and the next free allocation is t+7. A single account can therefore be in the market at most one day in six unless it ladders. The laddered design keeps six overlapping 6-day books so the portfolio is (almost) always fully invested while still honouring the per-tranche lock.
Six-sleeve ladder — illustrative state at t=1045
Each sleeve holds ~16.7% of the book and matures every sixth trading day. Today only the t=1045 sleeve is actionable; it is parked in cash because the regime is in the capitulation danger zone.
Laddered strategy — full-history backtest (Feb 2022 → May 2026)
Strategy
Cum return
CAGR
Sharpe (ann)
Max DD
Win %
Capital working
Equity curve — laddered (recommended) vs no-gate laddered
Even though the gate qualifies a pick on only ~58% of days, laddering six sleeves keeps the book near-continuously invested — each cash day idles just one sleeve (~1/6) while the other five stay in their existing 6-day holds.
Laddered tranche return by year (recommended gate)
Year
Mean / tranche
Entries
Win
2023 and 2025 carried the curve; 2026 YTD is a −6.1%/tranche drawdown — the current capitulation tape. The gate cannot avoid every bad patch, but it sizes each one to ~1/6 of book.
↻ The gap rule INVERTS over a 6-day hold. In the intraday note, large gaps (|gap|>2%) outperformed — momentum. Over t+6 the sign reverses: big gaps mean-revert. |gap|>2% returns −1.11% vs +0.50% for |gap|≤0.5% (Welch t=−3.43, p=0.0006). Big gap-ups are the single worst bucket at −1.56% (win 36%). So for the swing horizon we fade gappers, not chase them.
|gap| > 2% (t+6)
−1.11%
vs +0.50% for small gaps
|gap| ≤ 0.5%
+0.50%
best |gap| bucket
Big gap-up
−1.56%
worst cell · win 36%
Welch t / p
−3.43
p = 0.0006 (significant)
Gap coverage
37%
rows with pre-open trade
t+6 forward return by signed gap bucket
Signed gap bucket
Mean chfwd (t+6)
Win rate
N
Read
Sizing implication. Apply a gap modifier at the t+6 entry that is the mirror of intraday: ×1.0 for |gap|≤2%, downsize to ~×0.4 for |gap|>2% (most aggressively for big gap-ups). At the portfolio level the effect is modest — big gappers are rare among in-universe top-5 — but it trims the worst-tail entries without hurting Sharpe.
Regime maps to t+6 returns with two danger zones, not one. By market_condition the relationship is cleanly monotonic (bull > neutral > bear). On the finer regime_score_short scale, the deployable set reveals a non-monotonic shape: a deep-bear trap at −3 to −2.5 and a second danger zone at full capitulation ≤ −4, sandwiching a bounce sweet-spot at −4 to −3 (+1.22% deployable, win 68%). The recommended gate excludes both danger zones.
t+6 return by market_condition (all realized rows)
Condition
Mean chfwd (t+6)
Win rate
N
t+6 return by regime_score_short tier — broad vs deployable
Tier
All rows
N
Top-5 in-universe
N
Gate
Where we are now: regime −4.13 → capitulation danger zone. The latest tape sits below −4, the deployable band that has historically lost −3.47% over t+6. By rule the current sleeve stays in cash. The door reopens if the score lifts into −4 to −3 (the bounce band) or recovers toward neutral (> −0.5, the largest and most reliably positive band at +0.94%).
Today's recommended allocation (t=1045 sleeve, 20 May 2026)
✗ Cash for the current sleeve — regime −4.13 is inside the capitulation danger zone (≤ −4)
✗ BRPT ranks #1 (prob 0.672, in universe) but the deployable ≤ −4 analog is −3.47% over t+6
✓ Five other sleeves continue their existing 6-day holds (laddered book is not flat overall)
↻ Re-engage this sleeve when regime lifts into −4 to −3 (bounce band) or toward neutral
t+6 laddered sizing pipeline
0. Filter to universe list (282 tickers, hard pre-filter)
1. Rank gate: keep top-5 by ana rank
2. Probability gate: keep prob1 ≥ 0.50
3. Regime gate: exclude both danger zones — deep-bear (−3 < score ≤ −2.5) and capitulation (score ≤ −4); allow the −4 to −3 bounce band and everything > −2.5
4. Weight survivors by inverse volatility (1/vola1m)
5. Gap modifier (mirror of intraday): ×0.4 for |gap| > 2%, else ×1.0
6. Deploy ~1/6 of book into this day's tranche; hold to t+6 close, no intermediate rebalance
7. Repeat daily across six rolling sleeves; if no name qualifies, the sleeve holds cash for that cycle
What the t+6 study confirmed
✓ The laddered six-tranche design keeps capital working under the lock (+84.0% cum, Sharpe 0.61)
✓ The regime gate adds ~45 pts of cumulative return over no-gate laddering (+84.0% vs +39.5%)
✓ Inverse-vol weighting lifts cumulative return vs equal weight
✓ The universe filter is as binding here as on the intraday model
Caveats & what differs from intraday
✗ Gap rule is inverted — fade big gaps over t+6, do not chase them
✗ Regime has two danger zones (−3 to −2.5 and ≤ −4), not a single extreme-bear floor
✗ Max drawdown is large (−43.5%) — this is an aggressive small/mid-cap swing book
✗ The ≤ −4 carve-out rests on a small sample (n=22 deployable) — monitor as more data arrives
✗ Per-tranche return is the t+6 close-to-close chfwd; live fills, slippage and the gap modifier will shift realized numbers