Tiger Global Management LLC Doubles Down on AI Plumbing After a Brutal Quarter

The biggest buys table is effectively an AI plumbing roll call. Tiger is piling into the companies that manufacture compute, package it, and sit at the chokepoints of data and workload growth.

• TSM: A 49.4% share increase and an extra $621.7M signals TSMC as the core way Tiger wants to own AI volume and complexity, not just the GPU brand.
• AMAT and AVGO: An $260.3M add to Applied Materials and $219.5M more into Broadcom are a statement that equipment, packaging, and connectivity are where incremental economics of AI will accrue.
• NVDA: Even after huge gains versus cost, they still added $174.4M; Tiger is not declaring “peak Nvidia” yet.
• META: Another $192.4M into Meta shows conviction that the ad + AI flywheel and cost discipline have more room, despite the huge run from their average entry.
• CPNG and SPOT: Adding $157.2M to Coupang and $154.8M to Spotify leans into scaled consumer internet platforms where operating leverage from logistics, data, and engagement is still underpriced.
• MELI: A new $233.4M stake in MercadoLibre adds a Latin American e-commerce/fintech engine to that same thesis of regional scale and margin expansion.

Taken together, the “buys” tell you Tiger is willing to average up on winners in the AI stack and selectively average into pain where unit economics are improving but sentiment is still fragile.

On the other side, Tiger’s biggest trims are less about abandoning growth and more about culling where upside looks capped relative to AI and commerce plumbing.

• MSFT: Cutting Microsoft by 54.4% and freeing up about $1.10B looks like classic funding: a big, liquid, richly-valued AI beneficiary swapped for higher-beta semis and equipment.
• TTWO: A 65.7% reduction in Take-Two (~$758.3M out) suggests they see better risk/reward away from hit-driven content and toward infrastructure-like assets.
• APO: Nearly halving Apollo (down 47.0%, about $324.9M) looks like a call that alt-credit and PE carry streams are less compelling than AI-driven earnings growth.
• RDDT, APP, XYZ, NOW: Trims in Reddit, AppLovin, Block, and ServiceNow signal reduced appetite for more speculative or fully-priced software/platform names where expectations are already rich or execution risk is real.
• CHYM: Cutting Chime by 21.9% despite being underwater points to diminishing patience for unproven consumer fintech at scale.

Put simply, the sells look like a rotation out of broad software and financial exposure into narrower, higher-conviction AI and commerce pipelines, using both winners and lukewarm positions as funding sources.

The sector-allocation chart shows Tiger doing what Tiger does: leaning harder into its best-understood edge. Technology moved from 59.66% to 60.6% of the book, but under the surface that’s a swap from generalist software to semis, tools, and enabling infrastructure.

Consumer Discretionary slipped from 25.74% to 24.71%, even as they added to Coupang, Spotify, and brought in EquipmentShare. That’s net down, but the quality of exposure is up the value chain: less ad-hoc consumer apps, more scaled networks with operating leverage.

Finance took the real hit, dropping from 5.99% to 4.3%. Trims in Apollo, Chime, CoStar and a decision not to add to a long list of bruised fintech positions show a clear deprioritization of the “alt-finance + neobank” complex.

Real Estate exposure rose from 3.49% to 4.83%, but that’s mostly a quirk of classification: MercadoLibre and Zillow are digital real-asset and marketplace plays, not REITs. Health Care edged down (UnitedHealth trim), while a new Lumentum stake lifted “Telecom” to 0.42%, consistent with the broader bet on bandwidth for AI and data-heavy apps.

The net effect is a portfolio that’s more tightly wired to data, compute, and logistics rails, and less to consumer finance experimentation and generic enterprise SaaS.

Viewed as a whole, this 13F reads like Tiger doubling down on the idea that AI, semis, and scaled commerce are still in the middle innings, even after a punishing quarter. The firm is willing to wear volatility to own manufacturers (TSM, AMAT), integrators (AVGO, NVDA), and data-rich platforms (GOOGL, META, AMZN) at increasing size.

At the same time, they are narrowing the number of “stories” in the book. Capital is coming out of generalized software winners, content, and alt-finance — areas where Tiger once pushed breadth — and being redeployed into a smaller set of high-conviction rails and regional champions like Coupang and MercadoLibre.

The adds to deeply underwater names such as Zillow, alongside new positions like EquipmentShare and Intel, show an appetite for asymmetric turnarounds where Tiger believes the market underestimates normalized earnings power. But they’re being selective: some loss-making fintech and consumer internet bets are simply being left to shrink as a percent of the book.

Going forward, investors should read this as Tiger preparing for a world where AI capex, semiconductor intensity, and logistics-driven commerce matter more than low-rate financial engineering or “good enough” SaaS. The portfolio now more cleanly expresses that worldview — if the AI and e-commerce capex cycle extends, this book is built to participate aggressively; if it doesn’t, Q1’s drawdown may prove to have been an expensive reshuffle rather than a turning point.