According to MarketWatch, Daniel Loeb’s Third Point LLC made significant new investments in SK Hynix and Ebara during the third quarter, viewing them as strategic plays in the artificial intelligence infrastructure buildout. The hedge fund’s investor letter specifically highlighted SK Hynix as a South Korean memory manufacturer and Ebara as a Japanese semiconductor-production equipment maker, noting both companies trade at “undemanding absolute valuations and meaningful discounts to their U.S. peers.” Loeb positioned these investments as a way to participate in the AI trade through companies that benefit from the technology’s expansion while avoiding the premium valuations of American alternatives. This move comes as Third Point continues to hold winning positions in Nvidia and TSMC, suggesting a broadening of their semiconductor and AI exposure across the global supply chain.
Sponsored content — provided for informational and promotional purposes.
The Valuation Arbitrage That’s Driving Smart Money
What Loeb is executing here is a classic global valuation arbitrage strategy that sophisticated investors have used for decades, but with a modern AI twist. While U.S. semiconductor stocks like Nvidia trade at forward P/E ratios exceeding 30x, South Korean and Japanese semiconductor companies often trade at single-digit or low-teens multiples despite similar exposure to the same structural growth trends. This discrepancy isn’t just about geographic risk—it reflects different market structures, investor bases, and historical trading patterns. Asian markets have traditionally valued manufacturing and hardware companies lower than their U.S. counterparts, creating opportunities for global investors who recognize that AI infrastructure requires both the glamorous designers and the essential manufacturers.
Why Memory Markets Are Both Opportunity and Risk
SK Hynix’s position as a leading memory manufacturer exposes Loeb’s fund to both the tremendous upside and historical volatility of the memory cycle. While SK Hynix has become a critical supplier of high-bandwidth memory (HBM) for AI accelerators, memory markets are notoriously cyclical with periods of oversupply and price wars. The current AI-driven demand for HBM creates a favorable pricing environment, but memory manufacturers have historically struggled to maintain pricing power during industry downturns. What makes this cycle different is that AI-specific memory represents a premium product with higher margins and more stable demand profiles than commodity DRAM, potentially smoothing out some of the historical volatility.
The Unsung Heroes of Semiconductor Manufacturing
Ebara represents a more nuanced play on the semiconductor equipment side that many investors overlook. While companies like Ebara don’t have the brand recognition of Applied Materials or ASML, they provide critical components like chemical mechanical polishing (CMP) systems and vacuum technology that are essential for advanced chip manufacturing. The investment thesis here is that regardless of which chip designer wins in AI—whether it’s Nvidia, AMD, or new entrants—the equipment suppliers benefit from the increased capital expenditure across the entire industry. This creates a diversified exposure to the AI buildout without betting on specific end-winners in the highly competitive AI accelerator market.
The Geopolitical Risks Wall Street Is Underestimating
What the optimistic valuation analysis often misses are the substantial geopolitical risks embedded in Asian semiconductor investments. South Korean companies operate in the shadow of North Korean threats and complex China relationships, while Japanese manufacturers face regional tensions and currency volatility. The U.S. CHIPS Act and export controls have created a fragmented global semiconductor landscape where companies must navigate competing national interests. While these risks contribute to the valuation discounts Loeb is exploiting, they represent real operational challenges that could impact supply chains, market access, and ultimately shareholder returns in ways that don’t affect U.S.-based semiconductor companies to the same degree.
How the Competitive Landscape Is Evolving
The traditional competitive dynamics in semiconductors are being rewritten by AI demands. Memory manufacturers like SK Hynix are no longer just commodity suppliers—they’re becoming strategic partners in AI system design, working directly with companies like Nvidia on custom memory solutions. Similarly, equipment makers like Ebara are benefiting from the increased technical requirements of advanced nodes needed for AI chips. This represents a fundamental shift from the past where memory and equipment companies had less pricing power and technological influence. However, this elevated position also brings increased R&D requirements and capital intensity that could pressure margins if the AI growth trajectory slows.
The Realistic Investment Timeline
While Loeb’s move appears strategically sound, investors should understand that these are not quick trades. Building semiconductor manufacturing capacity and developing advanced equipment takes years, meaning the full benefits of AI-driven demand may not materialize in quarterly earnings for some time. The market’s patience with these investments will be tested during periods when AI enthusiasm wanes or broader economic conditions deteriorate. Historical patterns suggest that while semiconductor cycles can create tremendous wealth during upswings, they require stomach for volatility that many investors underestimate when chasing the AI narrative.
