The AI Value Chain and AI Technology ETFs
Checking ETF Holdings Is Essential
From Equity ETFs to Covered Call Structures
Investor buying has been heavily concentrated in newly listed AI-related ETFs. Among ETFs listed in Korea, 23 products include “AI” in their names. Excluding three actively managed ETFs that use AI in portfolio management—FOCUS AI Korea Active, WOORI AI ESG Active, and TIGER AI Korea Growth Active—the total rises to 24 AI-focused ETFs when recently listed ACE AI Value Chain series products and Nvidia bond-balanced ETFs are included.
Nine AI ETFs were listed in 2023, but by June 2024, ten new products had already been launched, surpassing the total number listed in the previous full year. Notably, both portfolio composition and product structure are evolving rapidly—from AI semiconductors to the broader AI value chain, and from pure equity exposure to covered call strategies.
From a portfolio perspective, AI ETFs can broadly be categorized into AI semiconductors, AI value chain, and AI technology (AI tech). The single most representative stock across AI ETFs is unquestionably Nvidia (NVDA). Parallel computing is essential for artificial intelligence, and Nvidia’s GPUs lie at the heart of that process. ETFs that explicitly include Nvidia in their names—ACE Nvidia Value Chain Active and ACE Nvidia Bond Balanced—allocate approximately 18% and 29%, respectively, to Nvidia shares.
AI semiconductor ETFs focused on Korean equities can be divided into two groups. One centers on HBM (High Bandwidth Memory) manufacturers, such as SK Hynix, while the other targets the HBM value chain, including equipment and materials suppliers. ACE AI Semiconductor Focus and DAISHIN343 AI Semiconductor & Infrastructure Active emphasize portfolios built around Samsung Electronics and SK Hynix. ETFs focused on HBM equipment and components include SOL AI Semiconductor Materials & Equipment, TIGER AI Semiconductor Core Process, and KODEX AI Semiconductor Core Equipment. All three are heavily weighted toward companies such as Hanmi Semiconductor and Isu Petasys.
The AI value chain itself can be divided into foundation models—best known through ChatGPT—hardware such as semiconductors required for computation, and software at the final service stage. SOL U.S. AI Software, listed in May, is a representative value chain ETF. The ACE Nvidia Value Chain Active, ACE Microsoft Value Chain Active, and ACE Google Value Chain Active series also offer differentiated exposure across various segments of the AI value chain.
More recently, on-device AI, where GPT-like models are embedded directly into smartphones and other devices, has emerged as another key link in the value chain—alongside AI power infrastructure, developed in response to severe electricity supply constraints. TIGER Global On-Device AI focuses on this theme, with concentrated exposure to Apple and ARM, a leader in mobile memory architecture. ETFs targeting AI power infrastructure—SOL U.S. AI Power Infrastructure, KODEX AI Power Core Equipment, and KODEX U.S. AI Power Core Infrastructure—are scheduled for listing in July.
Samsung Asset Management has also constructed an AI Tech Top 10 portfolio by combining the U.S. big-tech “Magnificent Seven”—a term borrowed from the Western film The Magnificent Seven—with the AI 5 defined by Light Street Capital analyst Glen Kacher: Nvidia, AMD, Broadcom, TSMC, and Microsoft. ETFs built on this portfolio include KODEX U.S. AI Tech TOP10 and KODEX U.S. AI Tech TOP10 +15% Premium.
KODEX U.S. AI Tech TOP10 +15% Premium employs a covered call structure, selling weekly call options on the Nasdaq-100 Index. While the AI tech portfolio does not perfectly match the Nasdaq-100—the option’s underlying index—approximately 30% of net assets are allocated to call option sales to target an annualized 15% option premium. As a result, the ETF retains about 70% participation in the upside of the equity portfolio and aims to distribute 1.25% per month using the collected premiums.
Volatility in Nvidia’s share price has been increasing, heightening concerns among investors exposed to AI themes. Even those with a long-term investment horizon may feel uneasy. The most fundamental response to volatility is diversification—across dividend stocks, bonds, or other asset classes. Beyond that, diversification within the AI theme itself, as well as across different ETF structures, represents a more advanced approach to portfolio design.
Rather than relying on a single stock or a single ETF structure, understanding the AI value chain and carefully selecting diversified exposure may prove essential for navigating the next phase of AI-driven markets.