DOSSIER · 13F-HR · Q1 2026

Castle Hook Partners LP

$23.4B in tracked AUM across 50 positions as of Q1 2026.

CH
CIK 0001687241 · last filed Mar 31, 2026
Total AUM
$23.4B
as of Q1 2026
Holdings
50
positions
Whale Score
83
83
strong
Activity
no data
changes this Q
AI Analysis · Q1 2026

Castle Hook Partners' $23.4 billion 13F filed for March 2026 tells a story of a portfolio that is quietly making one very large sector bet while cleaning up an unusually messy slate of legacy positions. The book's defining feature is — and has been — its massive core ETF holdings: SPDR S&P 500 at roughly $10.8 billion and SPDR Gold at $6.0 billion together represent 71.4% of assets, with the remaining quarter of the portfolio spread across 48 individual stock and sector bets. That ETF anchor is not the interesting thing about this quarter. The interesting thing is what the portfolio manager did with the active sleeve: Q1 saw 28 new positions initiated and 28 positions liquidated, effectively replacing almost the entire active book, and the choices signal a focused, almost archetypal rotation. The two largest additions in the active sleeve were airlines — United Airlines at $1.16 billion and Delta at $488 million, which together became the largest single-stick positions in the portfolio and constitute a remarkably directional bet on airline demand recovery and pricing power. That bet was supported by further additions in technology infrastructure — Seagate, Western Digital, Lumentum, and TTM Technologies each joined the book in the $176–$260 million range, names that point toward a semiconductor storage and advanced electronics sub-theme that runs counter to the software-universe direction many managers took out of the post-2022 drawdown. At the same time, the portfolio liquidated roughly $4.96 billion of holdings including QQQ, Citigroup, Eli Lilly, Shopify, Tesla, Carvana, Alcoa, and Cleveland-Cliffs — positions that represented a range of prior mandates from fintech to materials to legacy tech. The thematic through-line on the exits is coherence: names whose theses had run their course or where the manager's conviction had weakened were excised in favor of a narrower airline-plus-semiconductor-thesis complex. The portfolio is now 50-name but actually quite focused in the non-ETF layer: airlines, defense-aerospace (RTX and L3Harris joined), data-center infrastructure (Equinix), energy (Exxon, Linde), and industrial names. The reduction in concentration at the individual stock level is worth flagging — the prior quarter had heavier weightings in Shopify, Tesla, AppLovin, and Dexcom, all names that had been bought at the 2023-2024 bottom and were sold on a thematic rotation rather than a price target. The messaging from this portfolio as a whole is consistent: the ETF core stays intact, and the active sleeve is run by someone making cyclical and thematic calls — this quarter, those calls were deeply pro-airline recovery and pro-semiconductor storage infrastructure.

The core is enormous. SPDR S&P 500 represents $10.79 billion at 46% of the book, and SPDR Gold is $5.95 billion at 25.4%. Those two positions combined account for 71.4% of assets in a portfolio that qualifies as an ETF-heavy fund-of-funds structure with an active overlay. The ETF positions are long-term, low-turnover allocations and appear to be the portfolio's structural anchor — the prior quarter shows SPDR S&P 500 with shares reported in thousands format (6,383,200 thousand shares = 6.38 billion shares valued at $4.35 billion) and Invesco QQQ at a similar format. The apparent share-count discrepancy between quarters reflects vendor reporting format changes rather than actual trading — the value moved, but the share-count comparison is unreliable. The ETF core on its own is not what makes this filing notable.

The active sleeve is where the action is. Castle Hook went into Q1 with approximately 39 active individual stock positions and came out with 46, but with almost no overlap — 28 were liquidated and 28 were initiated, making this one of the most transaction-heavy portfolio resets in the cohort. The sold names are especially instructive as a narrative of what the portfolio was no longer willing to hold. The five largest exits by value were QQQ (an Invesco Nasdaq-100 ETF worth $3.92 billion, replaced by the SPY-focused structure), Citigroup ($607 million), Eli Lilly ($412 million), Shopify ($319 million), and Tesla ($272 million). The losses on positions like Dexcom (sold for $117.9 million below the prior-quarter carrying value), Applied Materials into weakness (sold into the price action), and Danaher (sold for $97 million below carrying value) all point to the manager cutting positions that had been initiated on earlier theses and were no longer performing. This is what you would call a theses-reassessment cycle rather than a panic liquidation.

Quarter at a glance — Q1 2026

Position-change comparison pending.

No quarter-over-quarter changes available.

Top 10 holdings

By portfolio weight as of Q1 2026.

#HoldingValueSharesWeight
01
SPY
STATE STR SPRD S&P 500 ETF T
$10.8B166K46.0%
02
GLD
SPDR GOLD TR
$6.0B138K25.4%
03
UAL
UNITED AIRLS HLDGS INC
$1.2B126K4.9%
04
NTRA
NATERA INC
$506M2.5M2.2%
05
DAL
DELTA AIR LINES INC
$488M73K2.1%
06
FIX
COMFORT SYS USA INC
$288M209K1.2%
07
EQIX
EQUINIX INC
$259M264K1.1%
08
RTX
RTX CORPORATION
$248M1.3M1.1%
09
WMT
WALMART INC
$243M2.0M1.0%
10
WDC
WESTERN DIGITAL CORP
$238M879K1.0%

Filing history

2026Q1Mar 31
$23.4B 2.6%
50 positionsView →
2025Q4Dec 31
$24.1B 73.9%
59 positionsView →
2025Q3Sep 30
$13.8B 21.4%
36 positionsView →
2025Q2Jun 30
$17.6B 82.3%
71 positionsLocked
2025Q1Mar 31
$9.7B 19.3%
37 positionsLocked
2024Q4Dec 31
$8.1B 51.7%
45 positionsLocked
2024Q3Sep 30
$5.3B 25.6%
39 positionsLocked
2024Q2Jun 30
$4.2B 19.2%
49 positionsLocked
2024Q1Mar 31
$3.6B 55.3%
56 positionsLocked
2023Q4Dec 31
$8.0B 181.6%
47 positionsLocked
2023Q3Sep 30
$2.8B 16.0%
41 positionsLocked
2023Q2Jun 30
$3.4B 23.7%
34 positionsLocked
Showing 12 of 36