🧭 MACRO SNAPSHOT

The macro setup is quietly constructive but the narrative is shifting. Fed funds at 3.64% (FRED), 2Y at 3.99%, HY credit spreads at just 272bps, and real GDP growth of 1.6% point to a soft-landing tape with no obvious credit stress. The 2Y above Fed funds suggests the market is pricing modest re-tightening risk rather than imminent cuts — and Kevin Warsh's WSJ op-ed arguing the Fed should "think about inflation differently" (WSJ, 5/30) is worth tracking as a leading indicator of how the next Chair pick reframes policy. Reported CPI YoY of 332.4 in the FRED pull is clearly the index level, not the rate — flagging as a data hygiene issue, not a signal.

The more relevant macro signal today is the Yahoo Finance/Reuters story that "After the AI binge, companies balk at soaring bills." This is the first real crack in the AI capex narrative — not a thesis-breaker for $NVDA, $AVGO, or $MSFT, but a reminder that enterprise inference cost discipline is starting to bite. Pair that with SemiAnalysis flagging Nvidia's Rubin CPX (a prefill-optimized inference chip) and you have the picture: hyperscaler capex is still accelerating, but enterprise customers downstream are price-sensitive. That's the dynamic to underwrite for the next 12 months.

₿ BITCOIN DAILY WRITE-UP

Price & Market Structure
BTC trades at $72,620, down 1.6% on the day, -6.3% on the week, -7.2% on the month. Dominance at 57.1% — relatively high, meaning alts are underperforming, which historically signals risk-off within crypto. We are now ~42% below the $126K ATH and ~14 months past the April 2024 halving.

Structural Thesis
The core thesis is unchanged: post-halving supply compression + permanent ETF demand sink + broadening institutional/sovereign adoption creates a multi-year structural bid that does not require a parabolic cycle to deliver compounding returns. We hold Bitcoin as a non-sovereign monetary asset with asymmetric upside, not as a trade.

What Happened This Week
Two notable items: (1) U.S. Treasury announced seizure of ~$1B of Iran-linked crypto (Bitcoin Magazine) — net neutral for BTC itself but a reminder that state-level engagement with crypto is expanding, not retreating. (2) The Bitcoin Layer published "Impasse," explicitly calling this "the late stage of the bear market" — meaningful because TBL has been a relatively measured bull voice. No fundamental on-chain or ETF flow data has materially shifted this week; the price drift is sentiment-driven, not flow-driven that we can verify.

Bull / Bear Scorecard

Bull:
- Halving supply shock still inside its 12-18 month historical expression window; we are at month 14.
- Institutional plumbing (ETFs, custody, corporate treasuries) is infrastructural and doesn't reverse — each cycle compounds the base.
- Sovereign accumulation expanding beyond El Salvador; U.S. policy posture has shifted to engagement vs. suppression.

Bear:
- 14 months post-halving and 42% below ATH is structurally unusual — the "cycle already peaked at $126K" scenario cannot be ruled out.
- BTC dominance rising while price falls suggests crypto-wide risk-off, not BTC-specific accumulation.
- TBL — a thoughtful bull — explicitly framing this as "late stage bear market" is a sentiment data point worth respecting.

Conviction Check
Action: HOLD | Conviction: 7/10. No change. The drawdown is uncomfortable but within the bounds of post-halving consolidations, and the structural pillars (ETF infrastructure, sovereign interest) remain intact. I would not add aggressively here without confirmation of either accumulation regime change or a macro tailwind (real yields breaking lower).

What to Watch
- 30-day rolling spot ETF net flows — sustained net outflows >30 days would be a thesis-breaker.
- Glassnode LTH supply — rotation from distribution back to accumulation would confirm cycle extension.
- DXY and 10Y real yields — a break lower in either would remove the dominant macro headwind.

Community Pulse
Reddit was dark across all tracked crypto subs (holiday weekend). Newsletter signal is mixed-to-cautious: TBL's "Impasse" piece explicitly frames us as late-stage bear, while their "Everything Everywhere All at Once" video frames Strait of Hormuz/AI/markets as overlapping pressure points. Tone is defensive without panic — the community is in "wait for confirmation" mode, not capitulation. The Coldcard MK5 launch is the kind of infrastructure news that suggests builders are still building regardless of price.

🔬 TODAY'S DEEP DIVES

Only one rolling review on deck today (no new ideas screened). Below is the COHR deep dive in full.

COHR — Coherent Corp. — ROLLING REVIEW
Conviction: 5/10 (downgraded from 7/10) | Status: Watchlist | Sector: Technology (Optical Components)

WHAT THEY DO
Coherent makes the optical and photonic components that connect AI clusters and power industrial lasers. Their bread and butter is high-speed optical transceivers (800G and 1.6T pluggables) sold to hyperscalers — directly to Meta, Microsoft, Amazon, Google, and through OEMs like $CSCO, Ciena, $ANET, and Nokia. They also sell industrial and EUV-grade lasers to semicap equipment makers, military lasers to defense primes, and optics into auto/medical. Roughly half the business is datacom (the AI story), and the other half is industrial/defense/telecom.

WHY IT'S INTERESTING NOW
The stock is up 354.6% over the trailing year and trades at $361.47 vs. a 52W range of $73.85–$413.00. This is no longer a discovery story — it's an execution story. AI cluster networking is driving 800G/1.6T transceiver demand at unprecedented rates, and Coherent is one of three names (alongside Lumentum/$LITE and Fabrinet) with the manufacturing scale to participate meaningfully. The question is whether the current valuation already prices the next 2-3 years of upside, or whether 1.6T ramps and co-packaged optics extend the runway.

BULL CASE
- Structural beneficiary of AI cluster networking: every GPU cluster expansion is a transceiver order, and 800G→1.6T is a clean ASP step-up with real margin lift.
- Diversification cushion: industrial laser and EUV power amplifier exposure provides counter-cyclical ballast that pure-play optical names (e.g., LITE) lack.
- Customer concentration is shifting from OEMs to direct hyperscaler relationships, which historically improves margins and stickiness.

BEAR CASE
- Valuation has fully re-rated: 1Y +354% with a 52W high already pierced this cycle. Most of the AI optical thesis is in the tape.
- Optical transceiver pricing is historically cyclical and competitive (LITE, FN, plus Chinese competitors like Innolight/Eoptolink). Co-packaged optics (CPO) could disintermediate pluggables over a 3-5 year horizon.
- Post-Finisar (2019) and Coherent Inc. (2022, $7B) acquisition debt and integration overhang remain — this is a complex, levered story to underwrite.

KEY METRICS
- 1Y return: +354.6%
- Diversification: ~50% datacom, ~50% industrial/telecom/defense
- Comp set: LITE (downgraded to 5/10 this week), Fabrinet, Innolight
- Differentiator: only Western player with both high-volume transceiver manufacturing AND industrial laser depth

BOTTOM LINE
Coherent stays on the watchlist at 5/10 — the business is real and the thesis remains structurally sound, but the price has done most of the work and I'd rather re-enter on a pullback or with confirmation of 1.6T design wins than chase here.

📋 TARGET LIST STATUS

Conviction changes this week: $COHR downgraded 7→5 on full valuation; $LITE downgraded to 5/10; $CIEN abandoned after 50-day review (the optical complex is broadly losing relative conviction as the AI networking premium gets fully priced). $SMAR added as new screen at 5/10. The cluster of 5/10 watchlist names is getting crowded — expect displacement pressure this month.

💼 YOUR PORTFOLIO

• $AAPL | HOLD | 5/10 — Trading at ATHs after a +6.5% run; BofA $380 target is constructive but conviction is capped until we see real AI integration revenue, not just installed-base monetization.
• $AVAV | HOLD | 7/10 — $43M PANTHER contract and Huntsville expansion reinforce the attritable drone/interceptor thesis. Securities fraud lawsuit headline is noise unless materiality is established.
• $AVGO | STRONG HOLD | 8/10 — Custom XPU/ASIC moat plus VMware cash machine. Near 52W high; let it run.
• $BABA | STRONG HOLD | 8/10 — Down 9.9% in 3 weeks and -34.5% from 52W high — the dislocation is the opportunity, not the warning. Cloud + AI re-rating thesis intact.
• $FSLR | TRIM | 6/10 — Up 51.96% in a month, +94% in a year. The thesis worked; take some off. Domestic solar tailwind is real but valuation has moved well ahead of it.
• $GOOGL | STRONG HOLD | 9/10 — Gemini 3.5 and Spark agent launch at I/O 2026 reinforces the structural-winner-in-AI thesis. Highest conviction in the book.
• $ISRG | BUY MORE | 8/10 — Down 5.9% to near 52W lows, -23% YoY. The single-port platform thesis hasn't changed; this is the entry point.
• $MKL | HOLD | 7/10 — Position is +103% on cost basis. TTM revenue contraction is the concern; specialty insurance cycle still our friend.
• $MP | HOLD | 7/10 — Only Western-hemisphere scaled rare earth. Selective trim opportunity on strength but core thesis intact.
• $SYM | HOLD | 6/10 — Down 21.45% on the month; warehouse automation thesis is structurally sound but execution risk is real. Don't add here, don't sell here.
• $TSLA | TRIM | 4/10 — Texas robotaxi fleet "dwarfed by Waymo's" (per recent news) is exactly the kind of data point that undermines the autonomy premium. Trimming continues to be the right move.
• $UNH | STRONG HOLD | 7/10 — Bernstein reiterated Outperform on PBM transparency. The thesis is recovery from the 234.60 low, and it's playing out.

⚠️ WATCH LIST

• $MU — Up 87.76% on the month, +29% on the week. Conviction is 6/10 but price action is screaming. Either upgrade on HBM4 design win confirmation or downgrade on memory cycle peak signal. Cramer "trillion-dollar market cap" framing is exactly the kind of late-stage narrative that warrants caution.
• $NOW — Up 40.83% on the month with no fundamental catalyst in the data. Either we're missing something (earnings beat, AI agent traction) or this is multiple expansion on enterprise AI hopes. Needs a fresh look this week.
• $ENPH — Up 107.4% on the month with no recent news. That's a 6/10-rated stock doubling in 30 days on what appears to be vibes. Either the solar inverter story is materially better than we credit, or this is a sentiment unwind waiting to happen. Flagging for fresh analysis.
• $PANW — Up 57.09% on the month with multiple street price target raises (Wedbush to $325 — new street high). Conviction at 6/10 is starting to look stale; if the platform consolidation thesis is working, this deserves an upgrade review.

🔁 RE-REVIEW QUEUE

• $CEG | Was 6/10 | Dropped 2026-05-13 — Dropped on relative conviction (not thesis breakage). Nuclear/AI power demand thesis has only strengthened since (see $VST up, $GEV pullback), so a fresh look is warranted.
• $DDOG | Was 6/10 | Dropped 2026-05-07 — Observability play dropped on valuation. With $ESTC and $NOW re-rating this month, the observability complex is back in motion — worth re-underwriting.
• $TTD | Was 6/10 | Dropped 2026-05-07 — Dropped on competitive pressure from retail media networks. Ben Evans' piece on OpenAI ad monetization is a reminder that the ad-tech competitive set is shifting — relevant input.
• $BWXT | Was 6/10 | Dropped 2026-05-07 — Small modular reactor exposure. Given nuclear renaissance momentum, deserves a fresh look.
• $NET | Was 6/10 | Dropped 2026-05-07 — Cloudflare; edge compute + AI inference at the edge thesis. The "companies balk at AI bills" macro narrative could be a tailwind for edge inference economics.
• $KTOS | Was 6/10 | Dropped 2026-05-13