The Rotation Playbook: Exit Paths by Sleeve

Builds-on: regime-check-august-7-2026 (B leads, 42; modes decomposed), how-inflation-dies-the-empty-reservoir Related: ai-crash-portfolio-defense, portfolio-rebalance-april-2026, failure-cascade-index, two-economy-gauge, the-hidden-bank-and-channel-1-yen-carry-unwind, china-oil-buffer-thesis-audit (Test A = the XLE adjudicator) Informs: the household book — his side + polly-fidelity-403b-allocation

Why this doc exists, and its governing constraint

B leads the board (42) and the thesis has gone mainstream — which raises exit-crowding risk: when everyone shares the same dials, the moves gap before slow hands act. The response is not to exit sooner on vibes; it's to pre-commit rotations precisely, front-load only the cheap and reversible ones, and prefer private dials over consensus ones. The governing constraint is the modal B-mode itself: the caught break (B-1998, 14 of B's 42) punishes premature full exits with a terminal melt-up. A 40-year-old with a 24-year horizon and steady contributions is structurally long the re-entry; the playbook's job is to make the round trip cheap, not to win the top-tick.

Three principles:

1. Rotations, not liquidations. Every move below swaps one exposure for another with a defined re-entry. Nothing goes to "out."
2. Act into strength where possible; triggers are for the rest. Selling insurance when it's expensive (BE >2.7, oil spikes) beats selling assets when everyone's dial fires at once. Consensus-dial triggers (payrolls, CPI prints) get half-sized because they're half-priced on arrival.
3. Account mechanics favor action: nearly everything sits in tax-advantaged wrappers — rotations are free. The scarce resource is decision quality, not tax lots.

Sleeve-by-sleeve

1. TIPS sleeve (SCHP + FIPDX ~10.5% household; STIP)

• Standing trigger (armed, amended 8/7): a second consecutive negative payroll print OR continuing claims >1.95M OR U-3 ≥4.4% — executed only if 10y breakeven ≥2.2% that week → rotate ⅓ to ½ of the sleeve into intermediate nominal Treasuries. Partial by design: A+D (42 combined) still owns the rest.
• Sell-into-strength side (the one nobody watches): BE ≥2.7% → trim the sleeve; the insurance has become expensive relative to every scenario path.
• New money: STIP over SCHP until the last hike is confirmed in; then extend to lock 2.4%+ real yields (per the 2026-08-05 analysis).
• Do not: dump on disinflation headlines (July 14's lesson) or on a single soft CPI print. Realized CPI vs 2.24% BE is the only scoreboard.

2. XLE (~4% his side)

• Dated exit test (the adjudicator): China Test A resolves structural — imports pinned ≤8.5M bpd through September while refinery runs recover → trim half. The floor thesis is impaired on evidence, not vibes.
• Terminal-value confirmation for the rest: long-dated strip (Dec-2027+) breaking and holding below ~$55 → exit the remainder. Until then the back-of-curve is the equity's true driver — front-month/Cushing fireworks are explicitly not signals (see the kinked-curve analysis).
• Sell-into-strength side: a war-premium spike carrying WTI >$105 without the long strip following → trim half into the rally (the China band caps it; capture the insurance payout).
• Do not: exit during the oscillation band on ceasefire headlines (fade-the-announcement remains the year's most durable rule), or add on physical-squeeze stories.

3. IAU (~4.6%)

• Hold. This is the D-hedge, and every rescue path terminates in D — the 1998-branch melt-up requires the Fed's balance sheet; the backstop ladder is already climbing toward it. Gold is the last sleeve to exit in this regime.
• Only trigger: a genuine C-resolution (inflation anchored ~2%, real yields >3% sustained, war resolved) → revisit. Nothing on the board points there.

4. The target-date core (~66%, TDF 2050 both sides)

The real question the mainstreaming raises — and the answer is deliberately modest:

• Now (cheap, reversible): redirect future contributions — not balances — to a more conservative mix (a 2040-vintage fund or an index/stable-value split) while B leads. Zero tax, zero timing drama, fully reversible, and it builds the dry powder that the re-entry rule below wants.
• Trigger-based single de-risk step (balances): if two of the following confirm within a quarter — continuing claims >1.95M; initial claims >240K; a failed/pulled hyperscaler bond deal (the digestion→spiral tell); the October S-1 pricing below range or breaking on debut — move balances one notch (2050→2040, ≈10-15 points less equity). Hard cap: one notch. More than that is a market call the 1998-branch exists to punish.
• The pre-authorized re-entry (more important than the exit): on a ≥30% S&P drawdown from the high, glide back to 2050 with both contributions and the notch. At 40 with 24 years of horizon, buying the break is the actual wealth event; the de-risk step exists to fund it, not to dodge volatility.
• Do not: exit equities wholesale, short anything, buy puts on vibes, or let the doom-content cadence set the tempo. The book's edge is solvency and rules, not timing.

5. Cash / new taxable money

• Maintain the base-spend buffer (the finances/ work defines it). New taxable savings to T-bills/STIP until the autumn cluster resolves — the option value of liquid dry powder into September-October exceeds any carry pickup.

The crowded-exit doctrine (why these triggers and not the famous ones)

Consensus dials — payrolls, CPI, FOMC outcomes — now move markets on release because everyone watches them; by execution time the rotation is half-priced. The playbook therefore leans on leading, less-crowded dials where we have them: new-issue concessions and subscription ratios (B-credit), SubV filings and the small-bank triple-stack (failure cascade), top-cohort spending (two-economy), the yen intervention decay clock (B-carry), secondaries discounts (B-marks), the long-dated oil strip (XLE). When a crowded dial and a quiet dial disagree, trust the quiet one; when only the crowded dial fires, act at half size.

Summary trigger table

What this playbook refuses to do

No shorts, no options, no wholesale exits, no acting on content-cycle cadence, no more than one equity notch in either direction per quarter. The mainstreaming of the thesis is information about crowding, not about timing — the correct response to everyone knowing what's next is smaller, earlier, cheaper, more reversible moves governed by quieter dials. The 93%-slow-bleed risk the video warns about (the Snapchat case) is real for concentrated single names; it is not the risk profile of a contribution-fed index book with a 2050 horizon — that book's enemy is being scared out at the bottom, and the re-entry rule is this doc's real payload.

Open questions

• Whether the one-notch cap survives contact with an uncaught B — if the fast modes (credit/marks/carry, ~18 of 100) fire without a catch, the cap will feel wrong in the moment. It stays anyway; that's what the cap is for.
• Whether Polly's side should mirror the contribution redirect — same logic applies to the 403(b), pending a check of its fund menu's conservative options.
• Where the D-confirmation threshold sits for rotating back into inflation assets aggressively (the reassertion trade) — likely the first Fed balance-sheet expansion announcement inside a selloff. To be specified before October.