US listing could put Japan’s Kioxia in global AI spotlight, Voya’s Thomas says

Kioxia, up 456% year-to-date and the Nikkei 225's top performer, is preparing to list U.S. American depositary shares to broaden its investor base and potentially improve trading liquidity. Voya Investments' $14 billion AI-focused fund manager said a U.S. listing could make Kioxia more investable for large global funds, similar to SK Hynix's Nasdaq listing. Voya's AI fund has generated roughly 600% cumulative returns over 10 years and favors liquid AI-infrastructure holdings such as Nvidia, SK Hynix and Micron.

Analysis

A U.S. ADS would primarily create an investability premium rather than alter Kioxia’s NAND earnings power. The relevant second-order effect is incremental global-fund demand for a scarce, liquid memory/AI-storage proxy, which could narrow the valuation discount versus MU and SK Hynix; it also gives passive and benchmark-aware capital a cleaner route into Japan’s semiconductor supply chain. That is more consequential for Kioxia’s future financing flexibility and acquisition currency over 6-18 months than for near-term industry pricing.

For MU, the prospective listing is modestly negative at the margin because it broadens the set of investable memory exposures precisely when AI-driven memory scarcity has supported premium multiples. However, Kioxia is more NAND-heavy while MU’s upside remains disproportionately tied to HBM/DRAM execution; a new listed peer should not impair MU unless it signals aggressive capacity spending or pricing behavior. The key missing diligence item is the ADS structure, float size, lockups, use of proceeds, and whether it is a genuine primary capital raise—without those details, an inflow thesis is speculative.

Consensus may overstate the analogy to other cross-listings: liquidity can expand ownership, but a sharp pre-listing rerating often pulls forward the benefit and leaves post-listing holders exposed to lockup supply and cyclical memory-price volatility. Over the next 1-3 months, any filing, investor-day guidance, or named U.S. cornerstone allocation is a catalyst; over 6-12 months, NAND contract-price trends and capex discipline matter far more. A weakening AI-server order cycle, falling NAND spot prices, or a large primary issuance would falsify the re-rating thesis.

Key Decisions for Investors

• Maintain MU as the preferred liquid U.S. memory exposure rather than chase a prospective Kioxia ADS before formal filing terms are available; reassess only if Kioxia discloses a sizable float and valuation materially below MU/SK Hynix on normalized mid-cycle earnings.
• Monitor a relative-value trade: long MU / short a broad NAND-sensitive semiconductor basket only if NAND pricing rolls over while HBM demand and MU DRAM guidance remain intact. Use the next MU earnings report as the decision point; exit if management cuts HBM qualification, DRAM-margin, or capex expectations.
• Treat any Kioxia listing announcement as a short-duration event-driven watch item, not an automatic long. Consider participation only after lockup terms, primary versus secondary share mix, and U.S. trading liquidity are confirmed; avoid if implied valuation already exceeds established memory peers without a demonstrable HBM or enterprise-SSD margin advantage.
• NDAQ has limited but positive optionality from another high-profile cross-border listing, but the revenue impact is unlikely to be material absent a broader pipeline of Japanese ADS issuers. No standalone position is warranted on this development alone.