Viking Global Investors
Ole Andreas Halvorsen (co-founder, CEO); Ning Jin (CIO)
Large, diversified Tiger-cub long/short fund running a more balanced, lower-AI-concentration book than its peers — financials, healthcare, and industrials alongside semis.
- Not open — Not an open-ended vehicle — analysis only, not an available allocation.
- Partial 13F — The disclosed book is longs-only and ~45 days lagged — a slice, not the strategy.
Top holdings
Top position; payments, not AI
Largest direct AI/semi exposure
Financials
Media
Industrial tech
Recent moves
Held a broad 77-name book led by Visa and financials, with TSMC its main direct AI exposure (~4.2%); the diversified, short-inclusive posture protected capital in past drawdowns but underperformed during the AI-led rally.
Loading disclosed positions…
Our take
Viking is the contrarian on this list — a deliberately diversified, less AI-concentrated book under a newer CIO; its 2025 lag is the cost of that caution, and the positioning signals skepticism that the AI trade's risk/reward still favors crowding in.
Two things get conflated when people read Viking's breadth as an anti-AI view.
First, breadth is partly structural: a multi-billion book with a real short sleeve and a private arm cannot express a thesis the way a 15-name fund can — 77 positions is what capacity looks like, not necessarily what conviction looks like. Second, the disclosed book is the least informative slice of Viking specifically: a 13F shows long US-listed positions only, ~45 days stale, and omits the short book that is the whole point. A fund whose edge is relative-value stock-picking is exactly the fund whose 13F tells you least. The durable asset is the research organisation — and the evidence for its quality is perverse: it is good enough that its output keeps self-funding into competitors (D1, Alua, Anomaly, Avantyr). That is both the strongest testament to the training system and the clearest risk to continuity. The 2017 return of $8B to reset size remains a capacity signal most managers never send.
Thesis
Viking is the Tiger cub that industrialised the analyst-to-CIO pipeline, then kept losing the CIO. Every CIO in firm history has departed — Ott (2010), Purcell (2015), Sundheim (2017), Jacobs (2019), Jin (2024) — most to found rivals. Halvorsen, 27 years from the 1999 founding, is the only constant. The 2025 lag versus AI-heavy peers is framed as a positioning call; the sharper question is whether a diversified, short-inclusive 77-position book is conviction or the natural output of a firm whose highest-conviction pickers keep leaving to run concentrated books.
Fundamental long/short global equity of Tiger lineage: deep-research sector teams underwriting multi-year theses, with a genuine short book alongside longs. Broader than concentrated peers — 77 disclosed 13F positions at Q1 2026 (13(f) securities, ADRs included) across financials, healthcare, industrials, semis. Sleeves: the flagship long/short, a long-only (Viking Long Fund), and private vehicles (>$14B, 70-plus companies).
Assessment
- Willingness to return capital ($8B in 2017) — a manager acting against its own fee base is a rare, credible capacity signal.
- A real short book, not a long-only fund wearing a hedge-fund fee. Breadth plus shorts is what produced past drawdown protection.
- Talent development is demonstrably world-class: multiple CIO-level alumni have raised institutional capital on a Viking track record.
- Diversification across financials, healthcare and industrials means the record is not a single-factor bet on one sector cycle.
- Key-person exposure concentrated in Halvorsen after 27 years; five CIOs have departed and none stayed to succeed him. Succession is unresolved, not solved.
- The 'deliberate anti-AI positioning' framing is a narrative fitted to an outcome. We found no disclosed statement of a capacity-constrained AI thesis — only a diversified book that lagged.
- Firm AUM reportedly grew past $55B (Sept 2025) after the 2017 reset to a smaller size. Capacity discipline looks like a one-time act, not a standing policy.
- Private-book marks (>$14B, 70-plus companies) are manager-estimated and were reported as lagging the public sleeves — a valuation-judgment layer inside a fund sold on liquid stock-picking.
- Attribution between skill, gross leverage and beta is unknowable from outside: no public gross/net exposure, no factor decomposition, no audited series.
- 2026-07Jul 24, 2026: Viking's July client letter calls the AI underweight a 'missed opportunity,' with Halvorsen citing valuations offering 'little margin of safety.' The stance is now disclosed, not inferred.
Record
Fund-level returns are private and reach the public only via press. The best-sourced datapoints are Q3 2024: Viking Global Equities +9.3% and Viking Long Fund +18.2% through September, while Viking Global Opportunities was -1.6% YTD and Opportunities Drawdown +8.3% — the private sleeves lagging the public ones, single-sourced to a paywalled trade report. The ~+5.8% full-year 2025 figure (versus the S&P's ~16%) traces to secondary press reporting, not a manager disclosure; treat it as reported, not established. Note the shape that gap implies: if the long-only sleeve materially outran the long/short sleeve, the hedge was the drag, which is the expected cost of the structure in a rising market — not evidence of stock-selection failure. Judging the strategy on one up-market year inverts what it is built for. The honest verdict: the record cannot be attributed from public data alone.
- 2026-07Jun 30, 2026: the lag now spans 18 months — FY2025 ~+5.8% followed by +2.6% in H1 2026, against Coatue +24.5% and Lone Pine +43%. One up-market year no longer covers it.
Risks & fit
- Halvorsen key-person and unresolved succession after five CIO departures — the single largest structural risk.
- Continued alumni attrition draining the research bench that is the actual asset.
- AUM growth toward and past prior capacity limits, diluting the per-name edge.
- Private-book (>$14B) mark-to-market credibility and illiquidity inside a liquid-strategy wrapper.
- Crowding: Tiger-cub alumni funds hold overlapping names, so a de-risking event hits many of them at once.
Two observations would overturn the 'durable research franchise, temporary lag' read. First, if the long/short sleeve underperforms the long-only sleeve again across a down or flat year — the regime the short book exists for — the hedge is destroying value structurally, not paying a cyclical cost. Second, if another CIO-level departure follows Jin's with no internal successor visibly holding the seat, key-person risk stops being a footnote and becomes the thesis. Conversely, a documented capital return at today's >$50B scale would confirm capacity discipline is standing policy, not a one-off.
A study in how a research organisation compounds — and leaks — human capital, and in why 13F reading breaks down for long/short managers. Viking's funds are private vehicles for institutional and qualified investors, not open to the general public; nothing here implies the reader can or should access them. The transferable lesson is analytical: when the edge is relative value, the disclosed long book is the wrong evidence.
Fee terms are not publicly disclosed. Tiger-cub long/short funds historically carry fees well above index costs, but we found no primary source for Viking's current schedule and will not estimate one. Any fee-drag figure here would be invented.