Over Half of Billionaire Chase Coleman's Portfolio Is Invested in 7 Brilliant Artificial Intelligence (AI) Stocks

Tiger Global Management had about 56% of its portfolio concentrated in seven large-cap AI-related stocks at quarter-end, led by Alphabet at 13.4%, Nvidia at 9.2%, and Amazon at 9.1%. During Q1, Chase Coleman added to Taiwan Semiconductor by 49% and Broadcom by 25%, while cutting Microsoft by 54% and slightly trimming Amazon. The article frames this as a bullish AI positioning update, but the information is backward-looking and likely to have limited immediate market impact.

Analysis

The portfolio message is less “AI exposure” and more “AI supply chain leverage.” The biggest second-order signal is the rotation from model-layer software exposure toward picks-and-shovels capacity: foundry, custom silicon, networking, and cloud infrastructure. That mix suggests Coleman is increasingly underwriting the durability of AI capex rather than just the adoption curve, which matters because the market can re-rate equipment and manufacturing bottlenecks before end-demand growth is visible in revenue.

The Microsoft reduction looks more like a relative-value decision than an outright AI skepticism call. When managers cut a dominant platform name while adding to TSM and Broadcom, they are effectively saying incremental AI dollars are migrating away from bundled cloud/office monetization toward the firms that control wafers, advanced packaging, interconnect, and ASIC design. If that view is right, the next leg of AI outperformance should be less about software beta and more about gross-margin expansion at the infrastructure layer, especially if hyperscalers keep pushing custom silicon to lower inference cost.

The contrarian risk is that the AI trade may be becoming too consensus in the exact names crowded into every hedge fund top 10 list. That creates a shallow downside catalyst: any capex pause, inventory digestion, or pricing pressure in foundry/networking could hit these leaders simultaneously over a 1-2 quarter horizon. In that scenario, the market may punish the most levered AI enablers first, even if secular demand remains intact, because expectations are already high and positioning is likely crowded.

The cleanest tell to watch is whether TSM and AVGO continue to outperform MSFT and AMZN on the back of AI capex revisions over the next earnings season. If they do, the market is confirming that compute scarcity remains the binding constraint; if not, the trade is in danger of turning into a late-cycle momentum basket. The asymmetry here favors owning the infrastructure beneficiaries but only on pullbacks, not chase levels, because the upside is tied to multi-quarter capex visibility while the downside can reprice in days if hyperscalers blink.

Key Decisions for Investors

• Long TSM / short MSFT as a 3-6 month relative-value pair: express the view that AI value capture is moving from software bundling to manufacturing leverage; stop if MSFT re-accelerates cloud demand or TSM guide turns flat.
• Add AVGO on weakness for a 6-12 month horizon: best pure-play proxy for custom AI silicon and networking attach; reward is continued multiple expansion if hyperscaler ASIC adoption broadens, risk is a near-term digestion after crowded ownership.
• Hold NVDA but hedge with a short-dated call overwrite into earnings: upside remains intact, but implied expectations are elevated and the stock is vulnerable to any sign of gross margin normalization or slower backlog conversion.
• Avoid chasing MSFT on AI enthusiasm until the next quarter: the recent de-risking suggests leadership may be transitioning away from platform software; better entry is after cloud growth inflects or valuation resets.
• For more aggressive expression, pair long TSM / long AVGO against short a basket of crowded megacap software names over 1-2 quarters to isolate the capex beneficiaries from the more sentiment-sensitive platform layer.