Executive Summary

The week’s data resolves into a clear signal: the economy is running hotter than expected on inflation while simultaneously fragmenting beneath the surface. Core PPI’s 0.8% January surge, well above consensus, pushes PCE Friday from a 25% hot-print probability to 35-40%, and raises the prospect that the Fed doesn’t cut at all in 2026. This is the single most consequential macro data point of the week because it directly triggers the DHI exit condition flagged in prior briefs. Meanwhile, NVDA’s post-earnings drop despite “phenomenal” results, CoreWeave’s 8% revenue guidance miss, and a software rotation emerging from retail buying confirm that AI infrastructure demand is real but the profitability of its downstream participants is fragile. The divergence between NVDA (chip seller, profitable) and CoreWeave (chip buyer/reseller, loss-widening) is the clearest illustration: NVDA validated demand at the top of the stack; CoreWeave proved the economics deteriorate as you move down.

Since Brief #25, five developments require portfolio-level attention. First, PPI’s 0.8% hot print approaches the DHI exit trigger — mortgage rates already ticked back to 6.09% from 5.99%, and if PCE Friday confirms persistent inflation, rate cut expectations compress further. Second, CoreWeave’s revenue miss and widening losses at 8% stock decline mean the Blue Owl / private credit AI infrastructure stress identified in Brief #20 is unresolved by NVDA’s beat. Third, BofA’s problem loan warning makes seven named institutional sources flagging credit stress in two weeks, upgrading credit cascade probability from 20-25% to 22-27%. Fourth, the M&A cycle has reached unprecedented breadth: Paramount-WBD ($31/share), EA ($55B rumored), Equinix-atNorth ($4B), Brink’s-NCR Atleos ($6.6B), DAE-Macquarie ($7B), Victory-Janus Henderson, and KORE ($726M) all announced within days. The JPM IB fee thesis has never been better supported. Fifth, SEBI’s authorization for India’s $384B equity fund universe to allocate up to 35% to gold/silver adds an 11th structural driver to the NEM thesis that operates independently from all prior gold catalysts.

The key evolution from prior briefs: I’m downgrading DHI from the portfolio after PPI provided the first half of the exit trigger (hot inflation print). Even before PCE Friday, the accumulated counter-evidence now exceeds the supporting evidence. Simultaneously, the Zscaler weak guidance raises the first material challenge to the CRWD thesis — while Zscaler operates in a different cybersecurity subsector (cloud/SASE vs. CRWD’s endpoint), two data points against cybersecurity (Anthropic AI tool selloff + Zscaler miss) require monitoring. I’m maintaining CRWD but flagging it as no longer at maximum evidence strength. The PYPL acquisition thesis from prior briefs must be downgraded to low probability given Stripe explicitly denied takeover talks — a direct contradiction of the two-source confirmed thesis.

Key Events & Analysis

Hot PPI: DHI Exit Trigger Approaching

Core PPI at 0.8% in January, well above consensus and accelerating for the second consecutive month, is hard data that overrides the softer rate path expectations that had supported housing. The mechanism is direct: PPI feeds into PCE calculation → PCE Friday likely comes in hot → Fed rate cut expectations compress → mortgage rates reverse their brief dip below 6% → housing demand weakens.

Mortgage rates already moved from 5.99% to 6.09% by February 27. The DHI thesis, which I have been reducing across Briefs #23-25 ($1,100 → $1,000 → $900), now faces its most unfavorable evidence configuration: five negative data points (January home sales -8%, PPI hot, bond crowding headwind, builder confidence weak, shutdown confidence drag) vs. three positive (HD comps, builder M&A, Morgan Stanley TACO). The prior brief’s explicit exit trigger was “PCE Friday hot + Fed hike rhetoric = reduce DHI $300, add ACGL.” PPI provides the first half of this trigger. Rather than waiting for PCE to confirm what PPI strongly implies, I’m executing the exit now and rotating to ACGL, which at 0.38 beta and 17.1% operating ROE is the superior risk-adjusted holding in a rising-rate environment.

IMF’s statement that the Fed is “nearing the end of its easing cycle” and Goolsbee’s resistance to using productivity as justification for cuts both reinforce the higher-for-longer thesis. The Warsh transition in May adds additional uncertainty — his confirmation environment just got more complex given hawkish data leaves less room for the rate cuts Trump demands.

CoreWeave: NVDA Beat Did Not Resolve AI Credit Stress

CoreWeave’s 8% stock drop on disappointing revenue guidance, widening net losses, and rising interest expenses is the most important disconfirming data for the narrative that NVDA’s beat resolves AI infrastructure credit concerns. The specific finding: even with validated demand (NVDA’s $200B revenue confirms customers are buying compute), CoreWeave cannot generate profits because the cost of capital for GPU infrastructure exceeds the revenue from reselling compute. This is the semiconductor-vs-downstream profitability split operating in real time.

For the portfolio: Blue Owl Capital’s stress identified in Brief #20 is NOT resolved. CoreWeave’s revenue miss means its order book has lower collateral value than NVDA’s beat implied. The $4B funding gap may widen rather than close. Credit cascade probability ticks up from 20-25% to 22-27%, driven by CoreWeave’s own financials (not NVDA’s demand signal) and BofA’s seventh institutional credit warning.

Equinix’s $4B atNorth acquisition provides a useful contrast: established data center REITs with investment-grade credit and lower leverage are winning the AI infrastructure buildout, while startup GPU cloud operators like CoreWeave struggle with capital costs. This favors EQIX and DLR over CoreWeave’s venture-backed model.

M&A Cycle at Maximum Breadth: Seven Deals Across Six Sectors

The deal flow this week validates the JPM IB fee thesis at maximum strength: Paramount-WBD (media), EA $55B (gaming), Equinix-atNorth $4B (data centers), Brink’s-NCR Atleos $6.6B (fintech infrastructure), DAE-Macquarie $7B (aircraft leasing), Victory-Janus Henderson (asset management), KORE $726M (IoT). Seven announced deals across six distinct sectors in five days. This breadth has not existed at any point in the 25-brief series.

JPM at 14.4x P/E with this pipeline is underpriced relative to the fee revenue these deals generate. SPGI and MCO benefit from debt issuance to finance acquisitions. The media consolidation specifically (WBD 50% surge, Netflix walking away) reshapes competitive dynamics — Netflix preserving its balance sheet while competitors consolidate declining assets may prove strategically superior.

Software Rotation: Short Squeeze Risk Crystallizing

NVDA’s post-earnings drop triggered a rotation into software stocks, exactly the Risk #1 (35% probability) flagged across Briefs #23-25. Jensen Huang’s specific comment that “AI won’t cannibalize software companies” is a narrative-shifting statement from the most influential voice in AI. Combined with retail buying software dips and CRM’s $50B buyback providing mechanical support, the Long NOW / Short CRM pair trade faces temporary reversal pressure.

However, I distinguish between narrative shift and fundamental shift. CRM’s revenue still missed guidance. WDAY’s margins still compressed. The five data points confirming the two-front squeeze haven’t been invalidated by one NVDA CEO comment. The trade may temporarily reverse (timing risk) but the underlying thesis remains intact. Importantly, NOW benefits from the same rotation if quality SaaS re-rates — the long side of the pair trade captures upside from the rotation while the short side faces squeeze risk. This argues for maintaining the pair but acknowledging elevated near-term volatility.

India’s SEBI Gold Authorization: 11th Structural Driver

India’s SEBI authorizing $384B in equity mutual funds to allocate up to 35% to gold and silver is a structural demand shift that did not exist before this week. Even conservative 5-10% reallocation from India’s equity fund universe represents $19-38B in new gold demand. This operates independently from all ten prior gold drivers (de-dollarization, sovereign hoarding, safe-haven, debt sustainability, anti-AI rotation, etc.). China’s record household savings shifting to gold + India’s institutional authorization = the two largest gold-consuming nations both increasing structural gold exposure simultaneously.

NEM is now the single position in the portfolio supported by the most independent evidence sources: eleven structural gold drivers, six institutional equity warnings, sovereign hoarding per Goldman, anti-AI rotation per Barron’s/Motley Fool, and SEBI’s institutional demand unlocking.

Iran and Pakistan: Global Instability Expanding

Iran talks ended inconclusively — the neither-deal-nor-escalation outcome maintains the current risk premium without triggering position adjustments. The LMT position stays at current sizing. Pakistan’s airstrikes on Kabul add a third active conflict theater (alongside Ukraine and Iran/Middle East), structurally reinforcing defense procurement urgency across NATO allies. India’s 1,000-point Sensex crash reflects this regional instability.

Block’s 24% Surge: AI Workforce Displacement Signal

Block cutting 50% of its workforce and being rewarded with a 24% stock surge sends an unambiguous signal to every CEO: the market rewards aggressive AI-driven headcount reduction. Combined with eBay’s 800 cuts and Duolingo’s AI threat disclosure, three companies disclosed AI-driven workforce reduction on the same day. The 30,000 tech layoffs in early 2026 are structural, not cyclical. This accelerates the labor market’s resolution from “contradictory” to “eventually higher unemployment” as AI displacement broadens from tech to financial services to professional services.

For IT outsourcing shorts (ACN, CTSH): if Block can operate with 50% fewer employees using AI tools, the demand for external consulting to deploy those tools initially rises but then faces an existential question — will companies need consultants if AI replaces the roles consultants were managing?

Portfolio Implications

What Changed Since Brief #25

  1. DHI EXIT: PPI 0.8% hot print triggers accumulated exit condition. Five negative data points now exceed three positive. Rotating to ACGL ($900 allocation).

  2. CoreWeave revenue miss contradicts NVDA-resolves-everything narrative. Blue Owl stress unresolved. Credit cascade probability up to 22-27%. APO thesis strengthened.

  3. Seven institutional credit warnings (BofA added). Credit warning density unprecedented. APO + ICE/CBOE remain best positioned.

  4. M&A at maximum breadth — seven deals in six sectors in five days. JPM IB thesis at strongest evidence point in series.

  5. SEBI gold authorization = 11th structural NEM driver. $384B Indian equity funds authorized for 35% gold/silver allocation.

  6. PYPL acquisition thesis downgraded. Stripe explicitly denied takeover talks — contradicts two-source confirmed thesis from prior briefs.

  7. Zscaler miss raises first challenge to broad cybersecurity bullishness. CRWD differentiated but sector headwinds emerging. Maintaining position but no longer maximum evidence strength.

  8. Software rotation underway (Risk #1 crystallizing at 35%). NOW/CRM pair trade faces temporary pressure. Thesis intact but timing risk elevated.

Tracking Prior Calls

  • DHI exit trigger (Briefs #22-25): PARTIALLY TRIGGERED. PPI hot print is the first half of “PCE Friday hot + Fed hike rhetoric.” I’m executing early given accumulated negative evidence. Call: Correct to have been reducing from $1,100 across briefs; exiting now.
  • NVDA beat validation (Briefs #21-25): Beat validated demand but post-earnings drop + CoreWeave miss mean the trade is “sell the news.” NVDA position needs GTC March catalyst. Call: Partially correct — demand validated but stock didn’t rally.
  • Software squeeze (5-6 data points): Thesis intact but software rotation creates timing risk for pair trade. Jensen’s “AI won’t cannibalize software” comment is a narrative headwind. Call: Fundamentally correct but facing near-term reversal.
  • PYPL acquisition (2-source): DISCONFIRMED. Stripe denied talks. Call: Wrong — two sources were insufficient; should have been treated as hypothesis.
  • Gold thesis (now 11 drivers): SEBI authorization adds 11th. Every brief has added supporting evidence. No disconfirming data. Call: Strongest thesis in series.
  • JPM M&A fees: Seven deals in six sectors validates at maximum strength. Call: Correct.
  • Credit cascade (22-27%): Seven institutional warnings. CoreWeave revenue miss adds confirming data. No resolution. Call: Probability increasing as predicted.

Risk Scenarios

Risk 1: Software Short Squeeze (35%, CRYSTALLIZING). NVDA post-earnings software rotation + Jensen “AI won’t cannibalize” comment + CRM $50B buyback + retail buying = fuel for 10-15% SaaS snap-back. NOW long side benefits but CRM short side faces squeeze. Timing risk, not thesis risk. Monitor for two consecutive weeks of software outperformance.

Risk 2: Credit Cascade (22-27%, UP from 20-25%). Seven institutional warnings. CoreWeave revenue miss. BofA problem loan warning. Private credit retail redemption risk. Bank-loan ETF structural vulnerability. Second fund gate = 30%+.

Risk 3: Tariff Legal Chaos (35%, UNCHANGED). Supreme Court ruling aftermath. FedEx refund lawsuits. Steve Madden guidance withdrawal. 150-day Congressional window.

Risk 4: Valuation Correction (25-30%, UNCHANGED). 9/10 sell indicators. S&P 500 concentration risk (NYT). Russell Microcap 57% unprofitable. Consumer staples crowding.

Risk 5: PCE Friday Confirms Hot Inflation (35-40%, UP from 25%). PPI 0.8% feeds into PCE calculation. If PCE confirms, rate cut expectations compress to one or zero in 2026. Broad equity repricing risk.

Risk 6: Iran Military Action (15-20%, UNCHANGED). Inconclusive Geneva talks maintain status quo risk. Pakistan-Kabul escalation adds regional instability.

Risk 7: Anti-AI Rotation Intensifies (25%, MATERIALIZING). NVDA dropped post-earnings while materials outperform. Global investor rotation out of US equities (MarketWatch/Bloomberg). Benefits NEM, pressures NVDA multiple.

Risk 8: CRWD Cybersecurity Sector Headwind (NEW, 15%). Zscaler weak guidance + Anthropic AI tool selloff = two data points against cybersecurity. CRWD differentiated (endpoint vs. cloud security) but sector backdrop deteriorating.

$10,000 Model Portfolio

Ticker Company Allocation ($) Shares Thesis
NEM Newmont Corporation $2,400 44 Eleven structural gold drivers including new SEBI $384B fund authorization; Pakistan-Kabul conflict adds geopolitical tailwind; anti-AI rotation flows; sovereign hoarding broadening; maximum evidence strength in portfolio
APO Apollo Global Management $1,700 12 Seven-source credit warning density unprecedented; CoreWeave revenue miss proves AI infrastructure credit stress unresolved; disciplined underwriting premium expanding; 16x vs BX 54x; private credit retail risk benefits disciplined managers
NVDA NVIDIA $1,300 7 Demand validated at $200B revenue but post-earnings drop means GTC March required for re-rating; CoreWeave miss shows downstream profitability fragile; reducing $100 to fund ACGL rotation
LMT Lockheed Martin $1,200 2 Iran talks inconclusive maintains risk premium; Pakistan-Kabul airstrikes add third conflict theater; Canada 5% GDP structural; no position adjustment triggers fired
ACGL Arch Capital Group $900 14 Replaces DHI after PPI exit trigger; 0.38 beta with 17.1% operating ROE; rate-insensitive insurance compounder; primary portfolio hedge against rising rates and credit stress
CRWD CrowdStrike $800 4 Zscaler miss introduces first sector headwind; reducing $100 to reflect no longer maximum evidence; still non-discretionary and differentiated from cloud security peers
NOW ServiceNow $900 1 Software rotation benefits long side of NOW/CRM pair; five data points intact despite Jensen’s narrative comment; workflow integration layer for enterprise AI; CEO $3M insider buy
JPM JPMorgan Chase $800 3 Seven deals across six sectors in five days — M&A fee thesis at maximum evidence; 14.4x P/E; Dimon’s public anxiety historically precedes outperformance

Changes from Brief #25: Three adjustments. (1) DHI exited, ACGL added at $900: The core change. PPI 0.8% hot print triggers the exit condition I’ve been flagging since Brief #23. Accumulated evidence ratio flipped: five negative data points (January sales -8%, PPI hot, bond crowding, builder confidence, shutdown) vs. three positive (HD comps, M&A validation, TACO). ACGL at 0.38 beta, 17.1% operating ROE, and 8.7x trailing P/E is the best risk-adjusted return in the S&P 500 and thrives in rising-rate environments where insurance investment income increases. (2) NEM increased from $2,300 to $2,400 (+$100): SEBI’s authorization of $384B in Indian equity funds for up to 35% gold/silver allocation is a genuinely new structural demand source. Eleven independent drivers is the highest evidence count for any single thesis in the 26-brief series. Funded from NVDA reduction. (3) NVDA reduced from $1,400 to $1,300 (-$100): Post-earnings stock drop despite beat, CoreWeave revenue miss showing downstream fragility, and software rotation suggesting NVDA is not a sentiment vehicle but an idiosyncratic fundamental story. GTC March remains required catalyst. (4) CRWD reduced from $900 to $800 (-$100): Zscaler weak guidance + Anthropic AI tool cybersecurity selloff = two data points challenging the sector. CRWD is differentiated (endpoint vs. cloud) but can no longer claim maximum sector tailwind. Funded the NEM increase.

Portfolio theme allocation: Physical assets and geopolitical protection (NEM + LMT = 36%, up from 35%) — SEBI gold authorization and Pakistan-Kabul escalation strengthen both positions. Financial/credit expertise and insurance (APO + ACGL + JPM = 34%, up from 33%) — ACGL replaces DHI as the highest risk-adjusted return per unit of evidence. DHI’s exit and ACGL’s entry transforms this sleeve from “credit expertise + housing value” to “credit expertise + insurance compounding.” AI infrastructure (NVDA + NOW = 22%, down from 23%) — demand validated but downstream profitability fragile; reducing to reflect CoreWeave reality. Cybersecurity (CRWD = 8%, down from 9%) — first headwinds emerging; maintaining at reduced size.

Exit triggers: NVDA: GTC March disappoints on Blackwell roadmap = reduce $400, add NEM + ACGL. LMT: Iran deal materialization = reduce $600, add $300 FCX + $300 NEM. NEM: reassess only on sustained gold below $4,500 (increasingly unlikely with 11 drivers). APO: second fund gate = increase $200. CRWD: if two more cybersecurity companies report weak guidance or AI displacement acceleration = reduce $400, add ACGL. NOW: if software rotation causes CRM to outperform NOW by >10% over two weeks = reassess pair trade thesis. NEW: PCE Friday hot = increases confidence in ACGL entry timing; no further action needed as DHI already exited.