Chris Wood boosts SK Hynix and Samsung bets amid South Korea chip pullback

Summary

Jefferies’ global equity strategist Christopher Wood has increased his exposure to South Korean memory chipmakers SK Hynix and Samsung Electronics, trimming positions in Indian, U.S. and Chinese stocks to fund the shift. Wood argues the recent correction in semiconductor stocks offers an attractive long-term buying opportunity as artificial intelligence adoption drives sustained demand for memory chips and computing infrastructure.

MUMBAI, June 30, 2026 — Christopher Wood, Global Head of Equity Strategy at Jefferies, has increased allocations to South Korean chipmakers SK Hynix and Samsung Electronics across his flagship GREED & Fear portfolios, taking advantage of the recent selloff in semiconductor stocks.

According to Wood’s latest GREED & Fear newsletter, the portfolio changes were funded by reducing or exiting positions in several Indian, U.S. and Chinese companies, reflecting his growing conviction that AI infrastructure will continue to outperform software-related investments over the long term.

The move comes after South Korean semiconductor shares pulled back following a strong rally earlier this year. Investors booked profits amid concerns over elevated valuations and the capital expenditure required to expand manufacturing capacity, although AI-driven demand for memory chips remains robust.

Wood believes the correction presents an attractive entry point rather than signalling the end of the AI hardware cycle. He argues that falling AI inference costs and wider adoption of generative AI applications will increase demand for high-bandwidth memory (HBM), DRAM and storage technologies.

Jevons Paradox underpins the investment thesis

Wood bases his outlook on Jevons Paradox, the economic principle that greater efficiency and lower costs can lead to higher overall consumption.

He argues that as AI models become cheaper to operate, businesses will deploy AI more broadly, increasing demand for computing power, memory chips and data storage rather than reducing it. In his view, memory manufacturers represent the “picks and shovels” of the AI boom and are better positioned than many software companies to benefit from rising infrastructure spending.

Wood also pointed to the growing use of long-term supply agreements by memory manufacturers as evidence that the industry has become structurally stronger than in previous semiconductor cycles.

Portfolio changes

Wood made several changes across his model portfolios:

• Global Long-Only Portfolio: Added a 4% allocation each to SK Hynix and Kioxia, exited Alphabet and Alibaba, and increased Samsung Electronics.
• International Long-Only (Ex-U.S.) Portfolio: Added SK Hynix and raised Samsung Electronics while exiting Alibaba.
• Asia Ex-Japan Portfolio: Sold PolicyBazaar to initiate a new position in SK Hynix.
• India Long-Only Portfolio: Exited Ambuja Cements and reduced holdings in GMR Airports, JSW Energy and Adani Energy Solutions.

The changes reflect a broader shift away from internet and infrastructure stocks toward companies expected to benefit directly from AI hardware investment.

Why this matters

• AI infrastructure remains the focus: Wood expects memory chipmakers to remain among the biggest beneficiaries of expanding AI workloads.
• Buying after the correction: He views the recent decline in South Korean semiconductor stocks as a valuation opportunity rather than a deterioration in fundamentals.
• Hardware over software: The strategy favours companies supplying critical AI infrastructure instead of software platforms facing increasing competition and pricing pressure.
• Global portfolio rotation: The changes highlight a continued shift in institutional portfolios toward semiconductor manufacturers tied to AI investment.

FAQs

Q1: Why has Chris Wood increased exposure to SK Hynix and Samsung?

He believes AI adoption will continue driving long-term demand for memory chips, making the recent share-price correction an attractive buying opportunity.

Q2: What is Jevons Paradox?

It is the economic theory that improved efficiency and lower costs often increase total consumption rather than reduce it. Wood applies this concept to AI, arguing cheaper AI services will increase demand for computing infrastructure.

Q3: Which stocks were sold to fund the purchases?

Wood exited or reduced positions including Alphabet, Alibaba, PolicyBazaar, Ambuja Cements, GMR Airports, JSW Energy and Adani Energy Solutions across various portfolios.

Q4: Why are memory chips important for AI?

High-bandwidth memory (HBM), DRAM and storage chips are essential for training and running advanced AI models, making them a critical part of AI infrastructure.

Q5: Does Wood believe the semiconductor cycle has changed?

Yes. He argues that stronger AI demand and more long-term customer agreements have made the memory industry less cyclical than in previous decades.