Tuesday plan: no trades with CPI one session out and the inflation-supply row back on
Monday's flat index hid a hard rotation — energy and gold bid, tech and small caps sold, long bonds lower — and Wednesday's CPI print is now one session away. The plan for Tuesday is zero actions: the regime moves to R4 event-suspended on a named tier-1 catalyst, the pre-committed MSFT trim stays armed rather than executed, and the book holds its five growth positions unchanged into the print.
Grade: the prior plan holds up — it refused to front-run Friday's broadening into an R1 upgrade and pre-named MSFT as the first source of funds if the inflation-supply row came back on. Monday withdrew the broadening and switched that row on, the book returned -0.03% and matched SPY at -0.03% with no trades required, and the pre-commitment is now live rather than hypothetical.
The Call
Zero trades. Wednesday's CPI print is one session away, which suspends discretionary decisions under the regime framework, and the honest read of Monday is that the inflation-supply rates row — the single most negative line for a book that is 77% long-duration growth — came back on. The pre-committed MSFT trim is therefore armed and dated, not executed: if the print runs hot, capital moves from the book's most multiple-sensitive position into AVGO's contracted-backlog stream, never into cash, defensives or hedges.
Since Last Session
Monday looked like nothing happened and quite a lot did. SPY finished -0.03%, but the composition was a hard rotation: energy was the day's engine with XLE +4.66% and XOM +4.41% as US oil moved back above $82 on growing doubt that Washington and Tehran will reach a Hormuz deal, and it was funded out of the leadership — QQQ -0.30%, XLK -0.88%, IWM -0.52%. Long bonds fell with it (TLT -0.85%), gold was bid (GLD +1.02%), and the defensive sleeves absorbed the flow (XLV +1.67%, XLF +0.36%). Underneath the oil headline sat a structural detail: the Strategic Petroleum Reserve fell below 300 million barrels, its lowest since 1983, which is less buffer against exactly the shock currently live.
Inside the book, the split mirrored the tape. AMZN +1.32%, MSFT +1.21% and JPM +0.63% carried; AVGO -1.25% and NVDA -2.86% cost, with NVDA falling further than its sector. The net was -0.03%, matching SPY at -0.03%, and the risk snapshot shows the era at 0.00% drawdown from its peak. A flat day that neither gave anything up nor learned anything cheap.
Plan for Tuesday
The intention for Tuesday is zero actions, and that is a decision rather than an absence of one. With a tier-1 macro print one session out, pre-event price action is positioning noise; the framework's stated best action count in that window is zero, and nothing in tonight's packet broke a holding's own thesis, which is the only exception that would override the suspension.
What the open needs to confirm, in order. First, whether Monday's energy move extends or was a single-session Hormuz repricing — a second day of XLE-style strength with long bonds lower would mean the inflation-supply driver is settling in rather than reacting, and that is what converts the armed MSFT trim into an executed one after Wednesday. Second, whether the long end stabilizes; the rates impulse is already negative over the intermediate window and a further leg lower in TLT does the most damage to the most extended positions. Third, how NVIDIA's $500 billion financing arrangement gets read — as an extension of the GPU demand runway, or as evidence that AI demand is increasingly credit-funded and therefore newly sensitive to the very print landing on Wednesday. That question is a reason not to add to NVDA; it is not, on tonight's evidence, a reason to cut it.
If NVDA falls again Tuesday with no fresh reason in the packet, the unexplained-decline rule arms a half-trim. Monday's fall had a reason — a rates-driven de-rating of the whole complex — so it does not qualify.
Regime Check
- Volatility state: VIX 15.46 — the normal band, standard rules. Direction is the useful part: it rose back above the 15 line Friday had briefly cleared, so the complacent/trend-friendly reading is withdrawn.
- Rates impulse + driver: negative, and the driver is now identified rather than assumed. TLT -0.85% with momentum20 -1.47% and momentum60 -3.09%, driven by inflation and supply — oil above $82 on Hormuz deal doubt, XLE +4.66%, and the SPR at its lowest since 1983. That is the framework's explicitly NEGATIVE row for long-duration growth, which is 77% of this book.
- Leadership breadth: rotation without breadth. Leaders sold (QQQ -0.30%, XLK -0.88%), hard assets and defensives bid (XLE +4.66%, GLD +1.02%, XLV +1.67%), small caps absent (IWM -0.52%), index flat (SPY -0.03%). The framework reads that as distribution until proven otherwise. Friday's broadening never got its second confirming session.
- Credit & dollar: unverified in tonight's packet. It can only lower risk appetite, never support raising it.
- Event proximity: CPI Wednesday, one session away — inside the two-session window.
Regime: R4 — event-suspended, a change from R2 that is permitted the same night because it rests on a named tier-1 catalyst rather than on price action. The two strongest supporting inputs are event proximity and the inflation-supply rates impulse. The contradicting input is genuine and is why nothing is being sold tonight: the tape is orderly rather than stressed, VIX at 15.46 is not a warning level, SPY was essentially unchanged, and prediction markets cited in tonight's packet point to a tamer CPI print — an outcome that would argue for holding maximum long-duration growth into Wednesday rather than trimming ahead of it. That is the whole case for arming the MSFT trim instead of executing it.
Positioning
| Holding | Weight | Thesis state | Action + tree branch | Evidence |
|---|---|---|---|---|
| MSFT | 23% | INTACT | Hold — INTACT branch (adds only on pullback-to-structure with regime agreement; neither is present) | +1.21% against XLK -0.88%, no company-specific item in the packet; extension is the risk, and the pre-committed trim is armed for post-CPI |
| JPM | 23% | INTACT | Hold — INTACT branch; explicitly not a funding source | +0.63% with XLF +0.36%; the only stream in the book that improves if the inflation-supply argument wins |
| AMZN | 20% | INTACT | Hold — INTACT branch; no add after a strong session | +1.32%, no Amazon-specific news; catalyst check unverified, which caps conviction at 7 |
| AVGO | 16% | INTACT | Hold — INTACT branch; the intended destination if MSFT is trimmed post-print | -1.25% with an identifiable market reason and an untriggered falsifier; contracted ASIC backlog is the least discretionary AI spend in the book |
| NVDA | 18% | INTACT | Hold — INTACT branch; new funding-quality flag logged, no add | -2.86% explained by the rates-driven de-rating, so the unexplained-decline half-trim is not triggered; $500bn financing extends the runway and adds a credit dependency |
No actions — no state changes. Sector exposure stands at 57% Information Technology, 23% Financials, 20% Consumer Discretionary; the largest single theme is 23%. Three actions have been taken in the last five sessions against a cap of two per session, and LLY remains blocked from re-entry.
Institutional Signals
The filings themselves are unchanged (all dated 2026-05-15, with Scion's stale at 2025-11-03), so there is no new 13F read. What is new is the route institutional capital is taking into this trade. NVIDIA lining up $500 billion in financing with Wall Street asset managers, and Huang framing chips as an 'investable asset', means a meaningful share of institutional AI exposure is now arriving as structured credit rather than as the equity sleeve that shows up in a 13F. Bridgewater's NVDA at 3.7% and Burry's at 13.5% understate how much institutional money is actually underwriting the compute build-out. For this book that cuts in two directions at once: the demand runway has a funding channel it did not visibly have before, and the AI complex has acquired a sensitivity to financing conditions — which is why a CPI print now matters to NVDA through a second channel, not just through the discount rate.
What Could Break It
The single event that would hurt the largest share of the book is Wednesday's CPI print coming in hot — or oil pressing higher into it on a Hormuz escalation, with the SPR at a 43-year low leaving less to lean on. That combination re-prices the front end on inflation grounds and compresses the multiple on MSFT, AMZN, NVDA and AVGO, which are 77% of the book between them. MSFT takes it hardest because it trades furthest above its 20-day average of the group. JPM at 23% is the only position that improves in that world, and it is not enough on its own to offset the rest — which is the risk being accepted deliberately, not hedged away.
Second, NVIDIA's financing structure becoming the story rather than the footnote: if $500 billion of chip-backed financing gets read as circular demand, the AI complex de-rates on funding quality instead of on capex, and NVDA and AVGO are 34% of the book between them.
Third, the position-specific falsifiers, unchanged and armed: a hyperscaler capex guide-down or China export escalation for NVDA, a hyperscaler ASIC program slipping or being in-sourced away from Broadcom for AVGO, AWS growth or capex efficiency disappointing for AMZN, cloud/AI deceleration or margin compression for MSFT, and materially wider HY spreads or a growth-scare curve move for JPM.