🧭 MACRO SNAPSHOT

The single most important variable this morning is the regime change at the Fed. Kevin Warsh was confirmed Friday by a 54-45 margin (the narrowest since 1977) and inherits a stagflationary cocktail: CPI running hot, PPI at 6%, and a bond market that has flipped from pricing cuts to a 45% probability of a hike. The 10Y at 4.47% and 2Y at 4.00% gives us a +47bp curve — meaningfully steeper than a month ago and consistent with the market repricing the long-end higher rather than the short-end lower. Fed funds at 3.64% (FRED) is the policy anchor, but the market is now skeptical that further cuts are coming. Warsh has floated swapping Core PCE for Trimmed Mean PCE as the inflation reference — that would create political cover to ease, but the bond market won't be fooled and term premium will widen if credibility erodes.

Layered on top: Brent back above $110 with the Strait of Hormuz blockade now expected to last months (WSJ noting coal making a comeback, Fortune flagging "non-linear" oil risk in June). Persistent $110 oil is directly inflationary and reinforces the hike-odds story. The setup matters most for high-multiple growth: $NVDA reports Wednesday at a 48 P/E into a tape where rate-sensitive growth is suddenly vulnerable. VIX at 17.26 and HY credit spreads at 276bp say the broader market is not yet pricing the stagflation risk that bonds and oil are flagging. That gap is the trade to watch.

₿ BITCOIN DAILY WRITE-UP

Price & Market Structure. BTC sits at $77,204, down 1.49% on the day, -4.62% on the week, but still +1.33% over 30 days. We are now ~37% below the $126K ATH printed earlier in this post-halving cycle. Dominance at 58.3% remains elevated — capital is concentrated in BTC, not flowing aggressively down the risk curve into alts.

Structural Thesis. Thirteen months past the April 2024 halving, daily new issuance is ~450 BTC against a $1.5T market cap — the lowest structural supply pressure in Bitcoin's history. The investment case is simple: scarce, monetizable digital bearer asset with a permanent and growing institutional bid (spot ETFs, sovereign wealth participation) into an inflationary fiscal regime. Hold for the supply/demand asymmetry, not for the chart.

What Happened This Week. Two structurally relevant items: (1) Abu Dhabi's Mubadala increased its IBIT stake 16% to $566M in Q1 2026 (Bitcoin Magazine) — sovereign wealth is adding, not trimming, even at elevated prices. (2) The Microsoft BitLocker "YellowKey" exploit story is the security narrative of the week — not bearish for BTC fundamentals, but a reminder that self-custody hygiene matters. Nothing in the news flow this week alters the structural thesis; what it doesn't tell us is the direction of current ETF flows, which remains the most important unconfirmed variable.

Bull / Bear Scorecard.
- 🟢 Halving supply shock still compressing sell-side; ~450 BTC/day new issuance against rising sovereign and ETF demand.
- 🟢 Post-halving month 13-18 has historically delivered the largest moves of the cycle; the early ATH is unusual but not disqualifying (2021 had a dual-peak structure).
- 🟢 Mubadala adding 16% to IBIT signals institutional accumulation at $77K is happening.
- 🔴 The 37% drawdown from ATH at month 13 deviates from prior cycles where the peak came LATE in this window — non-trivial probability the cycle top is in.
- 🔴 ETF flow data is the swing variable and we don't have confirmation it's still net positive. Sustained outflows would break the cycle's marginal-buyer thesis.
- 🔴 Macro alignment at 5/10 reflects rising real yields and a 45% hike probability — both directly negative for non-yielding assets in the near term.

Conviction Check. Action: STRONG HOLD | Conviction: 8/10. No change. Price weakness is consistent with the bear case but does not by itself invalidate the structural thesis. We hold because the supply asymmetry and institutional adoption layer are intact.

What to Watch. (1) ETF net flows — five consecutive weeks of outflow would force a conviction review. (2) BTC dominance — a break below 55% suggests risk-on rotation; a sustained move above 60% suggests defensive consolidation. (3) Warsh's first FOMC and any signal on the PCE measurement change — a credibility loss at the Fed is structurally bullish for BTC over 6-12 months.

Community Pulse. Tone is defensive but not panicked. The top Bitcoin Reddit thread — "AI Agents Have Already Chosen Their Money: Bitcoin" (746 upvotes) — is classic long-horizon thesis content, not capitulation. The BitLocker backdoor thread (274 upvotes) is generating unanimous "hardware wallet, not software" advice — operationally relevant for William's custody. The Bitcoin Layer newsletter is flagging an "Unconfirmed Red Dot" liquidity signal and writing about a "Flight to US Equities" — meaning sophisticated macro readers see USD assets as the flight-to-safety, not BTC, in the very near term. The community isn't bullish on the next 30 days; it's bullish on the next 30 months.

🔬 TODAY'S DEEP DIVES

UPST — Upstart Holdings — NEW IDEA
Conviction: 4/10 | Status: MONITORING (Screened, Not Added) | Sector: Financials / Fintech

WHAT THEY DO. Upstart runs an AI-driven consumer lending platform. They don't lend their own money primarily — they use machine learning models to underwrite borrowers (unsecured personal loans, auto refi, HELOCs, small-dollar) and route those loans to bank/credit union partners and institutional capital markets buyers. They earn fees on origination and servicing. The pitch: their ML beats FICO at predicting default, which means lower losses and higher approval rates for the same risk.

WHY IT'S INTERESTING NOW. The optical setup is striking — TTM P/E of 71.97 vs. forward P/E of 8.78 (yfinance) implies the market is pricing a massive earnings inflection. Revenue is back to +44.6% TTM growth after the 2022-2023 collapse, suggesting the platform is reaccelerating as rates normalize. In theory, a Fed easing cycle is rocket fuel for an AI lender: lower funding costs, higher origination volume, better unit economics.

BULL CASE:
- Forward P/E of 8.78 vs. TTM 71.97 signals an expected earnings inflection (yfinance). If $3.36 forward EPS materializes, the stock is at single-digit earnings on a recovery cycle.
- Revenue growth +44.6% TTM confirms the platform is back in growth mode (yfinance).
- Secular tailwind from rate normalization should expand the credit box and improve loan economics.

BEAR CASE:
- Active securities class action litigation — Pomerantz, Schall, and Gross all filed in May 2026 (PRNewswire/GlobeNewswire 2026-05-15). Multiple simultaneous fraud filings typically allege misrepresentations about loan performance or model accuracy. This is a hard stop.
- The forward P/E assumes a Fed easing cycle that the bond market is now actively un-pricing (45% hike odds per Daily Pulse).
- Business model has high beta to credit cycle and funding markets — both potentially deteriorating.

KEY METRICS. Revenue +44.6% TTM; TTM P/E 71.97; Forward P/E 8.78; differentiator is the ML underwriting model (which is also exactly what the litigation likely targets).

BOTTOM LINE. Not adding to the target list at 4/10 — active fraud litigation plus a macro tape moving against the forward-EPS assumption is a no-go until the legal picture clarifies.

SMAR — Smartsheet — NEW IDEA
Conviction: 2/10 | Status: WATCHLIST (Screened, Not Added) | Sector: Technology / SaaS

WHAT THEY DO. Smartsheet is enterprise collaborative work management software — think a spreadsheet-grid interface for project management, workflow automation, and team coordination. Sold primarily to mid-market and enterprise via seat-based SaaS subscriptions.

WHY IT'S INTERESTING NOW. Honestly, it isn't — at least not in a way I can verify today. The data payload was empty: market cap $0.0B, no price, no filings, no news. The likely explanation is that the previously-rumored take-private transaction has closed, meaning there is no public security to recommend. I will not fabricate around a data hole.

BULL CASE:
- Historical strength: high dollar-based net retention (~115%+ in past 10-Ks) suggested durable upsell.
- FCF inflection occurred in FY24 — if continued, validated operating leverage.
- Secular CWM tailwind from hybrid work.

BEAR CASE:
- Take-private status: if the deal closed, there is no public security to own. This is the dominant fact.
- Microsoft Loop/Planner/Copilot bundling compresses standalone CWM TAM.
- AI-native workflow tools (Notion AI, ClickUp) disrupt the category.

KEY METRICS. None reliable — data pipeline failure.

BOTTOM LINE. Not added at 2/10. The first action item is to confirm whether SMAR is still publicly traded; if it is, re-screen with clean data. Otherwise, archive.

AVGO — Broadcom — ROLLING REVIEW
Conviction: 7/10 (maintained) | Status: MONITORING | Sector: Semiconductors

WHAT THEY DO. Broadcom is two businesses under one roof. (1) Semiconductors: custom AI accelerators (XPUs) for hyperscalers like Google's TPU, networking ASICs/switches (Tomahawk, Jericho) for AI fabrics, plus broadband, wireless, and storage chips. (2) Infrastructure Software: the VMware franchise, security (Symantec), and mainframe software (CA), generally sold under multi-year enterprise contracts. Roughly half hardware, half high-margin recurring software — a unique combination.

WHY IT'S INTERESTING NOW. AVGO is the cleanest beneficiary of two simultaneous AI buildout trends: custom silicon (as hyperscalers reduce Nvidia dependency) and AI networking (Ethernet displacing InfiniBand). Management has guided to $60-90B serviceable AI revenue by FY27 across the top 3 hyperscaler customers (AVGO Q4 FY24 call). AI revenue reportedly grew +106% YoY. But at $440 the stock is at the 52-week high, market cap is $2.01T, and Trefis is now openly asking whether the stock is "expensive at $440" (2026-05-15). The whole chip complex sold off on rising yields last week (Financial Post, 2026-05-15).

BULL CASE:
- Custom AI silicon TAM expansion: $60-90B serviceable opportunity by FY27 across top 3 hyperscalers. 50-60% capture implies $35-45B AI revenue vs. ~$15B run-rate today.
- Ethernet AI networking inflection — every InfiniBand-to-Ethernet decision incrementally benefits AVGO's Tomahawk/Jericho franchise.
- Software half (VMware, Symantec, CA) provides high-margin recurring base that smooths semi cyclicality.

BEAR CASE:
- Valuation prices in perfection. $2.01T market cap at P/S 29.5x and EV/EBITDA 55.5x leaves zero margin for execution slippage. Trefis (2026-05-15) explicitly questions valuation here.
- Customer concentration: AI revenue is dominated by 3 hyperscalers (reportedly GOOG, META, plus a third). Any in-sourcing pivot or share shift is a step-function risk.
- Rising 10Y yields (4.47%) and 45% hike probability are direct headwinds for high-multiple semis.

KEY METRICS. Market cap ~$2.01T; AI revenue growing +106% YoY; P/S 29.5x; EV/EBITDA 55.5x; differentiator vs. peers is the hybrid hardware+recurring software model.

BOTTOM LINE. Conviction held at 7/10, monitoring — fundamentals are excellent but valuation is at the level where I need a pullback or an earnings beat with raised guidance to upgrade to RECOMMEND.

📋 TARGET LIST STATUS

No conviction changes today. Two new screens (UPST 4/10, SMAR 2/10) failed to clear the threshold to displace existing names. Recent changes: VRT and ADBE dropped this week at 6/10 and 5/10 respectively after being displaced; VEEV and CPRT auto-screened in. Target list at 49 — one open slot for a high-conviction idea.

💼 YOUR PORTFOLIO