Lone Pine Just Did the Aschenbrenner Trade at $2B, Coatue Cut Microsoft by $1.6B, the AVGO Convergence Fires

The Lead

Q1 2026 13F filings landed Friday — eight of the nine watchlist funds we track filed at the deadline. The single most-revealing single-fund positioning move is Lone Pine Capital’s wholesale rotation out of AI mega-caps and into AI-infrastructure secondary beneficiaries, executed at $2B+ of moved capital and in textbook Chamath/Aschenbrenner alignment.

The trade, on the data:

• EXITED at $2.4B combined: Microsoft ($597M), Amazon ($557M), Broadcom ($599M), Wingstop ($380M), Affirm ($284M)
• INITIATED at $2.7B combined: Teradyne ($555M), Corning ($506M), MasTec ($493M), Ciena ($314M), Argan ($214M), TeraWulf ($287M), Hut 8 ($285M), McKesson ($339M), Performance Food ($341M), US Foods ($350M)

If you strip out the food-distribution names (McKesson / Performance Food / US Foods — likely defensive sector positioning unrelated to AI), the remaining new positions are an almost-perfect map of the “money flowing to asset-heavy data-centre and power infrastructure” thesis from Issue 03 and re-confirmed in Issue 04 by the chip-tier prints.

This is a top-decile long-only growth fund doing the Aschenbrenner trade at institutional scale. The single largest signal in a 13F season often is — and this is it.

Falsifiable form: Teradyne / Corning / MasTec / Ciena should outperform MSFT / AMZN over the next 6-12 months for Lone Pine’s rotation to look prescient. By the Q3 2026 13F (filed November 14), we’ll know whether the rotation was early or mistimed. If by then they’ve reversed any of these positions, the conviction wasn’t there. If they’ve added, the thesis is hardening.

Threads

Lone Pine just did the Aschenbrenner trade at $2B

Steve Mandel’s Lone Pine Capital filed its Q1 2026 13F-HR on 15 May. Source: SEC EDGAR, accession 0000919574-26-003439. Total reported AUM: $12.54B (down marginally from $13.61B prior quarter — likely partly mark-to-market). Position count: 36 in Q1 vs 32 in Q4.

The new positions, ranked by initial position size:

Ticker	Initial $M	Sector	Aschenbrenner-thesis fit
TER (Teradyne)	$555M	Semi test equipment	Direct — test gear sees demand from any custom silicon program
GLW (Corning)	$506M	Optical fibre + glass	Direct — fibre buildout for hyperscaler interconnects
MTZ (MasTec)	$493M	Construction (infra)	Direct — power & telecom construction for DC sites
CIEN (Ciena)	$314M	Optical networking	Direct — transport layer for hyperscaler DCs
AGX (Argan)	$214M	Power-plant construction	Direct — natural-gas + renewable power for AI loads
WULF (TeraWulf)	$287M	Crypto-pivot to AI compute	Direct — same trade Aschenbrenner / Chamath called
HUT (Hut 8)	$285M	Crypto-pivot to AI compute	Direct — same

Seven new positions, all in either compute-adjacent infrastructure or direct power/construction. None of these names overlap with Mag 7 or with any traditional growth-fund holdings. Lone Pine is signalling, with $2B+ of new capital, that the highest-IRR way to play the AI capex cycle is one degree removed from NVDA.

The funded-by side: $2.4B of exits in MSFT / AMZN / AVGO / WING / AFRM. Three of those are AI-narrative names (MSFT, AMZN, AVGO — though AVGO is the off-watchlist convergence beneficiary other funds are buying, see below). Lone Pine’s bet: the secondary-beneficiary basket re-rates higher than the primary-beneficiary basket over the next 12-24 months.

Lab-calibrated: This sits in future_bet × bullish setup at the issuer level for each name. +2.5% mean direction-adjusted alpha at 20d (n=365). Note the relevant lab calibration is on the narratives of these signals — the actual 13F position itself is a stock-selection signal, not a filing-derived signal, so the historical direction-adjusted alpha framework doesn’t perfectly apply. Treat as a thematically-strong reading not yet validated by our lab’s exact methodology.

Coatue cut Microsoft by $1.58B, exited Oracle / Snow / Adobe / AMD / ARM

Coatue Management’s Q1 2026 13F (filed 15 May, accession 0000919574-26-003501) is the most aggressive single-quarter repositioning in our 4,669-row historical data set. $29B AUM as of Q1, down from $40B prior. Position count: 62 vs 52.

Exited entirely:

• ORCL ($865M)
• SNOW ($544M)
• ADBE ($306M)
• AMD ($265M)
• ARM ($251M)
• MRVL ($177M)

Materially trimmed:

• MSFT: $2.76B → $1.39B = -$1.58B (-57%) ← single largest position-cut in the data set
• TSLA: $1.64B → $59M = -$716M (-96%) ← effectively a full exit
• DASH: -$906M
• CRVN: -$612M

Initiated:

• EQIX $1.07B (NEW)
• ASML $655M (NEW)
• V $217M, QCOM $179M (NEW)

The rotation is loud: OUT of AI software / services / Mag-7 narrative names, INTO infrastructure (EQIX) + semi-cap (ASML). Combined with the AMD warrants disclosure from Issue 04 and the MSFT-OpenAI exclusivity erosion from Issue 03, this is institutional confirmation that the “buy the obvious AI mega-cap names” trade peaked.

Worth noting that Coatue is the only fund of the eight that took down MSFT this aggressively. Altimeter trimmed MSFT modestly ($618M → $438M = -29%); the other six funds either held flat or were never material holders. So this is a Coatue-specific call — not a converging exit signal. But the conviction is highest in the data set.

The AVGO convergence fires — Issue 03 thesis confirmed

Issue 03 (1 May): “AVGO is the biggest beneficiary nobody is talking about… AVGO meets the criterion (off-watchlist ticker that is a primary beneficiary of multiple watchlist-company actions). Worth proactively flagging.” Issue 04 (10 May): config-level CIK fix shipped to resume AVGO ingestion that had been silently failing since 2018.

Q1 2026 13F data confirms:

Fund	Q4 2025 AVGO	Q1 2026 AVGO	Δ
Altimeter	0	$21M (NEW)	new
Coatue	$1.04B	$850M (held + flow)	trim
D1Capital	$359K shares	$981K shares ($180M added)	+2.7x position
Duquesne	small	held	held
WhaleRock	held	held	held

Five of the eight deadline-day funds hold AVGO in Q1 2026. D1Capital is the high-conviction adder (2.7x the position, +$180M). The convergence-detection threshold (≥3 funds buying an off-watchlist name) fires cleanly. AVGO should be added to the formal watchlist with priority status before Issue 06.

The disagreement: Lone Pine exited AVGO entirely ($599M position → 0). They were one of the rotation funds described in the lead thread above — selling AVGO to fund the Teradyne / Corning / MasTec / Ciena infrastructure pivot. Net read: AVGO has both believers and skeptics among top funds; the believers outnumber the skeptics 5:1.

Off-watchlist convergence: Carvana (6 funds), AppLovin (4 funds)

The convergence detector flags two genuinely off-watchlist names that meet the multi-fund threshold:

Carvana (CVNA) — 6 of 8 funds: Coatue, D1Capital, Dragoneer, Durable, LonePine, WhaleRock. Combined holding $1.96B in Q1 2026. Not an AI thesis — this is a recovery / re-rating play on the online auto-retail market. Worth tracking because (a) the convergence ratio is unusually high and (b) at least three of these funds (Coatue, Dragoneer, Durable) are not typically auto-sector investors.

Caveat: Coatue trimmed CVNA by $612M in Q1 (from 1.96M shares to 681K shares), so it’s a “still held but materially reduced” position rather than a fresh conviction add. The 6-fund convergence is more of a “everyone bought at some point and most are still holding” pattern than a “Q1 stampede.”

AppLovin (APP) — 4 of 8 funds: Coatue, D1Capital, LonePine, WhaleRock. Combined $1.64B. LonePine added +$57M (+87% to share count) in Q1 — the most material add. AppLovin is AI-adjacent: their AXON ad engine uses ML for ad-targeting. The thesis: pure-play software with embedded ML monetisation, not the obvious “buy NVDA” trade. Worth proactively adding to watchlist along with AVGO.

Other off-watchlist multi-fund names worth noting (not yet promotion-tier):

• NU Holdings (4 funds, $1.4B) — Brazilian fintech
• Coupang (CPNG) (4 funds, $0.99B) — Korean e-commerce
• Mercadolibre (MELI) — interesting CONVERGING EXIT signal: 3 funds (Altimeter, Dragoneer, WhaleRock) all exited MELI in Q1. Rare to see three top funds simultaneously dump a position.

Atreides + Aschenbrenner didn’t file — that’s its own signal

Two prominent funds we track did NOT file Q1 2026 13F-HR by the 15 May deadline:

• Atreides Management (Gavin Baker) — most recent 13F-HR was 17 February (Q4 2025); since then only a Schedule 13G/A on 15 May
• Aschenbrenner’s Situational Awareness LP — most recent 13F-HR was 11 February

Both have historically been deadline-day filers. Plausible explanations:

1. Form NT 13F filed for extension — gives an extra 5 business days; we’d see the actual 13F by Friday 22 May. Most likely explanation.
2. AUM dropped below $100M reporting threshold — required only above that level. Unlikely for Atreides (multi-billion fund); possible-but-unlikely for Aschenbrenner’s relatively young vehicle.
3. Material change in advisor structure — re-registration or fund-restructuring delay.

We’ll know by next Friday. The deferred filings, when they land, may carry signal value — managers sometimes delay 13F when they’re materially building positions and don’t want to broadcast.

Action: re-poll EDGAR for both funds Monday 18 May and again Wed 20 May. If 13F-HR lands, integrate findings into Issue 06 watchlist updates.

Watchlist Updates

Ticker	Direction	Driver	Lab-calibrated reading
AVGO	bullish	5-fund convergence; D1Capital added +$180M	Convergence-promotion target — add to formal watchlist before Issue 06
TER, GLW, MTZ, CIEN, AGX, WULF, HUT	bullish	LonePine new positions confirming Aschenbrenner thesis	Add all to watchlist as “AI-infra basket”
APP	bullish	4-fund off-watchlist convergence	Promotion-tier — add to watchlist
CVNA	mixed	6-fund hold but Coatue trimming	Off-watchlist; monitor but don’t promote yet
MSFT	neutral-to-bearish	Coatue -$1.58B (the single largest single-position cut)	Reinforces Issue 03 / 04 reading on OpenAI exclusivity erosion
TSLA	bearish on margin	Coatue -$716M (-96%, effective exit)	Lab caution — bearish future_bet -1.3% mean alpha. Don’t oversell.
ORCL, SNOW, ADBE, AMD, ARM, MRVL	mixed	Coatue exited all of these	Single-fund signal not yet a converging exit — but conviction is significant
EQIX	bullish	Coatue NEW $1.07B; reinforces Issue 03 colocation thesis	Strong — bullish future_bet calibration applies
ASML	bullish	Coatue NEW $655M	Semi-cap concentration thesis

The Week Ahead

• Monday 18 May & Wednesday 20 May: Re-poll EDGAR for Atreides + Aschenbrenner deferred filings
• Tuesday 19 May: NVDA reports Q1 FY27 after-close. The Q1 13F data has set the stage — Coatue cutting MSFT and exiting AMD/ARM/MRVL is a bearish-mega-cap read. NVDA’s print will be the bull/bear referee. Capex commentary + China commentary the two things to watch.
• Wednesday 20 May: TGT Q1 (consumer read-through — less relevant for AI thesis).
• Thursday 21 May: PANW (cybersecurity tied to AI usage — relevant secondary read), DECK.
• Friday 22 May: Atreides + Aschenbrenner DEADLINE under the 5-day extension if they filed Form NT.
• Mon-Sun 18-24 May: Continued NVDA pre-earnings positioning + initial reactions to today’s 13F data.

What would change next week’s thesis: a deferred Atreides 13F showing concentrated positions in any of the LonePine names (TER / GLW / MTZ / CIEN / AGX / WULF / HUT) would create a 2-fund convergence on the AI-infrastructure basket and accelerate the promotion case for those names. Conversely, if Atreides shows a de-risked book leaning back to mega-cap tech, that’s evidence against the rotation thesis being broad.

Methodology and disclosures

Filings Intel Digest is a bona fide financial publication. Nothing here is personalised investment advice. The editor may hold positions in companies discussed; current positions and policy are at /about. All claims are sourced to publicly filed documents linked inline. Backtested or historical figures are direction-adjusted and calibrated against the signal lab; calibration is partial — treat any forward statement as a hypothesis, not a forecast.

Frozen as of close-of-trading 15 May 2026 (Friday). Source 13F filings retrieved direct from EDGAR Sunday 17 May. Holdings parsed from informationTable XML; positions aggregated per CUSIP per fund; Q1 2026 (filed 15 May) compared against Q4 2025 (filed 17 February) baseline.

Methodology note: 13F data is backward-looking by ~45 days (positions as of 31 March 2026, filed 15 May). Funds may have materially changed positions between the as-of date and today; the data should be read as a snapshot of mid-quarter conviction, not current holdings. Coatue specifically filed a 13F-HR/A amendment same day — we used the primary 13F-HR.

Lone Pine’s exit-MSFT signal is one fund’s view. The lab-calibration on multi-fund convergence signals is incomplete — we have only 5 quarters of multi-fund 13F data and have not yet computed direction-adjusted alpha for convergence-promoted tickers. Treat the AVGO / APP promotion case as thesis-grounded, not yet edge-grounded.