As analysts widely speculate that Japan's NAND flash memory leader Kioxia is poised to knock on Wall Street's door, potentially following global semiconductor giants like TSMC and SK Hynix into US public listings, the storage chip market that recently returned to a bullish trajectory is experiencing dual catalysts from both demand fundamentals and capital flows. The core investment significance of the anticipated US listing lies in significantly expanding the trading market accessible to institutional investors, thereby adding a new funding source for a company already propelled by AI demand. Kioxia's shares have surged 456% year-to-date on the Japanese stock market, leading the Nikkei 225 index, yet Sebastian Thomas, an investment manager at Voya Asset Management which runs a $14 billion AI-focused fund, still does not hold the stock, primarily due to liquidity concerns. Thomas noted that for large funds, beyond corporate fundamentals, actual position-building capacity, trading impact costs, exit convenience, and liquidity strength all influence allocation decisions. Emulating SK Hynix's recent ADR listing on the US stock market could improve these conditions and enhance Kioxia's investability within global institutional AI portfolios.

Storage chips break through the midsummer consolidation as Goldman Sachs detects a new storage bull cycle

Wall Street financial giant Goldman Sachs has re-endorsed the bullish trading environment surrounding the storage chip sector, with core rationale rooted in the expansion of frontier high-performance AI computing power demand brought by OpenAI's Astra launch, combined with the RSI (Recursive Self-Improvement) training paradigm beginning to dominate AI training contexts, which has heightened demand expectations for storage chips. This coincides with traditionally low positioning from Wall Street asset managers and hedge funds, especially as stocks like Micron and SanDisk begin breaking above summer downtrend technical lines, semiconductor implied volatility retreats, and the risk exposure previously trimmed by hedge funds leaves room for re-accumulation. On the fundamental side of storage, UBS senior analyst Timothy Arcuri projects that average selling prices for storage chips in the third quarter will rise more than 20% quarter-over-quarter on top of a record-high second-quarter base, with DRAM and NAND supply shortages persisting through 2027. Combining these two types of bullish investment views forms a constructive thesis of "earnings expectations supported with room for low-position capital to rebuild exposure"; whether sustained re-rating continues ultimately depends on actual selling prices, shipment volumes, and profit realization. The market action prior to Thursday's selloff triggered by strong US PPI data had already reflected this storage theme recovery. The Korean KOSPI index, often dubbed the "AI computing power barometer," rebounded approximately 21% from its late-July low by mid-August, reaching the commonly defined technical bull market threshold; on September 7th it gained another 4.6%, with Samsung Electronics and SK Hynix rising 5.7% and 8.1% respectively. In the US market, as of the September 9th close, the Philadelphia Semiconductor Index rose for five consecutive trading sessions with cumulative gains near 6%, while the Roundhill Storage ETF gained nearly 12% over the same period, demonstrating the storage sector leading this rebound.

The significant industry signal from Astra is that more complex work now holds commercial value for delegation to AI execution. On September 10th, media reported that OpenAI launched a ChatGPT product tailored for the financial services industry, combining GPT-6 Astra with professional data sources to support research, financial modeling, and client material production. Extrapolating from this, the growth variables of AI demand will further expand to encompass concurrent agent counts, task execution durations, tool invocation frequencies, and context scale: when the cost of completing a task declines and success rates improve, enterprises have greater incentive to deploy more workflows. This opens exceptionally broad new space for cloud-based AI inference computing power and AI-related high-performance storage demand, serving as the latest evidence for the market to reassess the sustainability of AI infrastructure growth.

From Tokyo's top performer to Wall Street's spotlight? A US listing could make Japan's Kioxia a global AI investment focus

From an architectural perspective, Kioxia's direct beneficiaries are primarily NAND flash memory and enterprise-grade SSDs: AI agents accessing enterprise knowledge bases, retrieving vector data, loading models, saving task states, and maintaining audit records all require persistent storage; long-context and multi-round tasks also expand demand for reusable key-value cache capacity, pushing some caching into SSDs through hierarchical architectures. HBM handles high-bandwidth compute access, server DRAM handles low-latency working sets, and NAND provides more economical large-capacity storage, with the three complementing one another. Kioxia has positioned SSDs with high random read/write performance as an important product direction for enhancing AI inference efficiency, and related research indicates that optimizing SSD cache scheduling can reduce redundant computation and GPU waiting. Therefore, Kioxia's long-term growth thesis should center on enterprise SSD capacity, performance value, and customer share gains; the US listing improves capital market participation conditions, but sustained earnings growth still requires product competitiveness, supply-demand dynamics, and order momentum to jointly support it. The chief fund manager of a $14 billion AI-focused fund stated that Kioxia's planned US listing could reposition this Japanese chipmaker as a global investment focus. Kioxia has surged 456% year-to-date, ranking first among constituents of the benchmark Nikkei 225 index. Sources familiar with the matter revealed that the company indicated at an internal meeting in May that it was preparing for a US stock ADR listing to broaden its investor base. Voya Investment Management's global AI fund holds no position in Kioxia, nor does it hold any of the technology stocks driving the Nikkei's advance. The fund includes $5 billion sourced from Japan. Voya portfolio manager Sebastian Thomas stated that the issue lies in liquidity, and trading in the US market could change that calculus, much like the AI financing and investment enthusiasm that followed Korean storage giant SK Hynix's July listing on the Nasdaq.

"There are many companies worth watching in Japan, especially those in the AI computing infrastructure supply chain," Thomas said in an interview on Thursday. "The challenge is finding companies with sufficient liquidity that allow us to invest." The AI fund, affiliated with Sumitomo Mitsui DS Asset Management, has delivered cumulative returns of approximately 600% on a pre-tax distribution reinvestment basis since its inception a decade ago this month. The fund concentrates its investments in the most core AI computing infrastructure companies in the compute supply chain, including AI chip superpower NVIDIA (NVDA.US), AMD, and AI ASIC leader Broadcom, with NVIDIA as its largest holding; it also invests in cutting-edge software application developers related to AI applications, as well as manufacturing companies like Eli Lilly (LLY.US) that stand to benefit from adopting frontier AI technologies. Thomas said Voya has previously invested in Japanese companies, and while it currently holds no Kioxia position, it maintains exposure to other storage chipmakers, including SK Hynix and Micron Technology. He added that a US listing makes such investments easier to execute. "We generally prefer technology companies with better liquidity, larger position capacity, and greater market capitalization," he said.