Baupost Group LLC Ma 13F Portfolio · Seth Klarman

Baupost is adding where it sees either a dominant cash machine or a mispriced cyclical. The biggest dollar add was Genuine Parts, where share count rose +89.0% to a 6.1% position, while Amazon remains the largest holding at 16.5%. These are not momentum buys; they are size-ups in businesses with repair demand, distribution scale and e-commerce/cloud cash flow.

• Amazon: added +20.0%, now 16.5%; the bet is AWS, advertising and retail margin expansion, not just consumer spending.
• Genuine Parts: nearly doubled shares +89.0%, now 6.1%; auto aftermarket demand is non-discretionary and cash-generative.
• CME Group: new 2.5% position, $136.7M; a capital-light exchange and clearing toll road that benefits from volatility and rate uncertainty.
• Norwegian Cruise Line: shares up +112.9%, now 3.0%; average cost $19.91 and the position is -1.8% versus cost, a classic contrarian cyclical recovery bet.
• Alphabet: added +16.2%, now 9.0%; the position is +96.6% versus average cost, yet Baupost still size-ups the AI/search/cloud winner.
• Axalta: new 0.8% position, $43.6M; coatings and refinish demand add a materials and pricing-power sleeve.
• GDS Holdings: added +23.8%, now 2.1%; China data-center exposure is up +166.3% versus cost, a deep-value infrastructure bet.
• Ferguson: added +0.6%, now 6.4%; a small top-up to an industrial distributor with housing and plumbing leverage.

The pattern: fund the new money with industrial and cable sales, and concentrate it in platform cash flows, auto repair, cruise recovery and exchange volatility infrastructure.

The sell side is more revealing than the buy side. Baupost cut Wesco by -53.9%, a $267.4M reduction, after the position had gained +113.2% versus average cost. That is harvesting a double, not panic.

• Wesco: -53.9% shares, $267.4M sold; industrial distribution winner monetized after a +113.2% gain.
• Union Pacific: -23.1%, $96.7M sold; rail exposure reduced despite a +27.7% gain versus cost.
• Restaurant Brands: -16.4%, $96.2M sold; still 9.0% of book, so this is trimming concentration rather than exiting.
• Eagle Materials: -40.5%, $81.5M sold; building-materials cyclical cut while the position is -10.5% versus cost.
• Liberty Global: -36.2%, $53.4M sold; cable/telecom remains a small 1.7% position and is -17.3% versus cost.
• Aon: -10.7%, $27.2M sold; insurance broker trimmed modestly even as finance weight rises.
• Teleflex: -11.7%, $23.6M sold; medical devices position reduced.
• Americold: -12.2%, $14.9M sold; cold-storage REIT trimmed.
• DNOW: -23.1%, $10.9M sold; oilfield distribution reduced.

The common thread: cyclical, rate-sensitive and lower-conviction names are being sold to fund the adds. The largest cuts are also among the fund's biggest historical winners or largest positions, which is how a concentrated value manager raises cash without breaking the core.

The reported sector weights say Consumer Discretionary rose to 42.1% from 40.9%, Technology climbed to 11.1% from 9.4%, and Finance increased to 6.8% from 4.7%. Industrials fell to 8.2% from 11.5%, Telecom dropped to 1.7% from 2.7%, and Health Care slipped to 17.4% from 17.8%. Those headline moves overstate the consumer-discretionary story.

The screener files Wesco under Consumer Discretionary, but it is industrial distribution; Ferguson is industrial distribution; Axalta is a coatings and materials company; DNOW is oilfield equipment. Visa is filed under Real Estate, but it is a payments network. After those corrections, the real shift is toward technology platforms, financial exchanges and select consumer cash flows, funded by cuts in industrial cyclicals and cable.

• Technology: weight up to 11.1% from 9.4%, driven by Alphabet and GDS; AI/cloud and data-center infrastructure.
• Finance: up to 6.8% from 4.7%; new CME adds an exchange and volatility toll road, while Visa remains a core payment network.
• Consumer Discretionary: headline 42.1% is inflated by misclassified industrial names; true consumer exposure is Amazon, Restaurant Brands, Genuine Parts and Norwegian Cruise.
• Industrials: reported weight down to 8.2% from 11.5%; the real reduction is clearer when Wesco and Ferguson are reclassified.
• Health Care: roughly flat at 17.4% from 17.8%; Elevance and Molina unchanged, Teleflex trimmed.
• Telecom: down to 1.7% from 2.7%; Liberty Global cut -36.2%.

Net: Baupost is not making a broad consumer bet. It is buying platform cash flows, payments and exchange infrastructure, and non-discretionary auto repair, while reducing rail, building materials, cable and other cyclical or levered exposure.

The forward read is that Baupost expects a market where company-specific cash-flow repair beats broad cyclical beta. The new CME position is a hedge-like asset: if volatility stays elevated or rate uncertainty persists, exchange and clearing volumes should benefit regardless of direction. Amazon and Alphabet are the platform compounders; Norwegian Cruise, Genuine Parts and Axalta are the mispriced cyclicals.

The biggest risk is concentration. Top 10 at 74.9% means one bad thesis can move the whole book. Norwegian Cruise is already -1.8% versus average cost, GDS is a China data-center bet with regulatory and currency risk, and Alphabet's +96.6% gain versus cost could tempt profit-taking if AI spending disappoints.

Performance gives the fund room: latest quarter +11.4%, 3-year annualized +13.9%, though 5-year annualized is +3.6%. The 2026-Q2 filing suggests Klarman's team is using that strong recent run to rebalance out of industrial winners and into what it sees as cheaper or more durable cash-flow streams. Watch whether next quarter continues the same funding pattern: more sales in rails, industrial distribution and cable, and more adds to platforms, exchanges and contrarian consumer cyclicals.