Li Lu, "the Chinese Warren Buffett," Has Invested 70% of Himalaya's Capital in Just 2 Stocks

Himalaya Capital (Li Lu) runs a highly concentrated 7-stock portfolio valued at $3.7B+, with 70% of capital allocated to just two names: Alphabet (48% of fund) and PDD Holdings (22%). The piece highlights Alphabet trading at 16.4x forward earnings despite DOJ antitrust scrutiny and intensifying competition from LLM-driven search, while PDD trades below 9x forward earnings but faces China macro deceleration, U.S.-China trade-war headwinds, and competition in ecommerce.

Analysis

The meaningful signal here is not the portfolio disclosure itself; it is the market endorsement of a few large-cap platforms with durable cash generation despite headline risk. That tends to compress the “disruption discount” on GOOG/GOOGL over a 6-18 month horizon, especially if AI monetization keeps pace with capex and the DOJ overhang remains procedural rather than structural. The second-order winner is AAPL indirectly: preserving default-search economics protects a high-margin services stream and removes a reason for the street to haircut ecosystem value.

PDD is the more fragile setup. A low multiple only matters if earnings are high-quality and policy-sensitive revenue can compound; here, trade friction and China cyclicality can turn reported EPS into a lagging indicator rather than a floor. If Temu faces tighter customs enforcement or higher cross-border fulfillment costs, the pain is likely to show up first in marketing efficiency and shipping subsidies, then in margin compression; BABA and JD are the likely relative share gainers inside China if that happens.

Contrarian take: the consensus may be overpaying for “cheapness” in PDD and underpricing the durability of GOOG’s cash engine. For GOOG, the risk is not a binary legal event but a slower grind of AI-related capex that delays margin expansion; for PDD, the risk is that forward earnings estimates are still too high if growth normalizes. This is a relative-value, not a catalyst-heavy, setup: the best trade is to own quality where the market still fears structural decay and avoid cheap stocks where the business model is more policy-dependent than the P/E implies.

Key Decisions for Investors

• Long GOOG / GOOGL versus short PDD as a 3-6 month relative-value pair: thesis is quality duration vs policy-sensitive cheapness; target 10-15% spread capture, exit if PDD prints clear reacceleration in GMV or GOOG margins compress meaningfully.
• Add modestly to GOOG/GOOGL on weakness only, not strength: use any post-news dip to build a position for a 6-18 month hold; thesis breaks if AI capex drives sustained margin dilution without offsetting monetization.
• Avoid initiating fresh PDD longs until there is visible proof of stable cross-border unit economics; if already exposed, trim into rallies unless management can show marketing efficiency and fulfillment costs improving over the next 1-2 quarters.
• Relative beneficiary watchlist: BABA and JD on any evidence that Temu is losing pricing power or facing enforcement friction; consider a small BABA or JD versus PDD pair if China e-commerce share shifts become visible in monthly data.