Daily Market Brief - Jul 23, 2026

Editor's Notes:

• This morning, the market is experiencing a significant divergence, with technology budgets rapidly shifting from traditional software-as-a-service (SaaS) models to physical AI hardware. This dynamic, dubbed the "SaaS Shakeout," is driven by generative AI's impact on headcount-linked subscriptions, while companies like Super Micro (SMCI) and Dell (DELL) emerge as primary beneficiaries in the "Physical AI" sector, transforming semiconductors into a reservation-based asset class.

Overall Themes, Market Sentiment & Debates

The Great AI "Hardware vs. Software" Divergence

The market has entered a ruthless phase of divergence where technology budgets are being zero-summed. We are witnessing a "SaaS Shakeout" across the US and HK, as investors aggressively rotate out of seat-based software models, such as ServiceNow, Salesforce, and HubSpot. This rotation is fueled by fears that generative AI cannibalizes traditional headcount-linked subscriptions.

Conversely, the "Physical AI" backbone is capturing every available dollar. Super Micro (SMCI) and Dell (DELL) remain the primary beneficiaries of a structural shift in procurement: semiconductors are no longer cyclical components but are now considered a "reservation-based" asset class. In this new paradigm, customers prepay for supply years in advance to secure access.

The "Circular Financing" Debate

A sophisticated debate is emerging among desks regarding the quality and sustainability of AI growth. With an estimated $1.65 trillion in off-balance-sheet commitments for infrastructure across Big Tech companies (Alphabet, Microsoft, Amazon, Meta, Oracle), the longevity of the "AI Treadmill" is being questioned.

Concerns specifically center on "circular financing," where chipmakers like Nvidia and AMD invest billions into customers such as Nebius and Anthropic. These customers then utilize that capital to purchase GPUs from the very chipmakers that invested in them. While this strategy secures demand in the short term, it raises the specter of a significant Capex bubble if enterprise ROI, exemplified by negative CoPilot feedback, fails to materialize over time.

Regional Market Nuances

Japan: The ¥3 Trillion Passive Tailwind

While the Nikkei 225 faces a technical "double correction" according to Elliott Wave analysis, a massive structural catalyst is looming for autumn 2026. A projected TOPIX float-ratio adjustment for Kioxia (285A.T), set to increase from 15% to 50%, is expected to trigger ¥3 trillion in passive buying. Niche leaders in modular data centers, like Sankyo Frontier, and materials companies, such as Mitsui Kinzoku (5706.T), are outperforming as the "Japan is back" narrative shifts to specialized infrastructure.

China A-Share & Hong Kong: Technical Crossovers vs. National Team Fatigue

In China, the "National Team" interventions, designed to support the market, are reportedly losing efficacy, with rallies typically fading within 24 hours. However, technical analysts are flagging a potential positive signal: $CHAU (CSI 300 Bull 2X) is reclaiming its 200-day moving average, which could indicate a sustained reversal. Hong Kong is observing a defensive rotation into high-dividend State-Owned Enterprises (SOEs) and Traditional Chinese Medicine (TCM) stocks, as investors hedge against ongoing domestic property market weakness.

US: The Nuclear Renaissance

The Trump administration’s $200 million "AI Nuclear Initiative," which selected Oklo and X-Energy, confirms a critical insight: the bottleneck for AI infrastructure is no longer solely about chips, but increasingly about the electric grid. This development is driving high-conviction bids into utility and nuclear energy names, as the escalating energy demands from new data centers place significant strain on US electrification capabilities.

Notable Big Stock Moves, Earnings and Development

Ranked by absolute price move (%)

Interesting Comments, Facts and Ideas

The Procurement Paradigm Shift in Semiconductors

The semiconductor sector is undergoing a fundamental shift from a traditional cyclical "ordering" model to a "reservation-based" procurement model. This transformation is observed across the entire AI system stack, encompassing CPUs, High Bandwidth Memory (HBM), networking components, and packaging technologies.

Under this new model, customers are signing long-term agreements and prepaying for components to secure priority access and supply. For hedge funds, this implies that traditional "book-to-bill" ratios may be less predictive of future performance than emerging "prepayment-to-backlog" metrics.

Retail vs. Reality: The Reddit (RDDT) Pricing Tension

While Reddit shares (RDDT) experienced an 8.3% drop today due to competitive pressure from Meta’s "Forums" launch, a high-conviction bullish pitch is circulating among data-focused funds. The core thesis suggests that Google’s current $60 million/year licensing deal for Reddit data, which is crucial for training AI models like Gemini, is significantly undervalued—a "steal," according to some.

Analysts anticipate that this licensing deal will re-price towards $1 billion/year at its renewal. This potential revaluation could fundamentally pivot Reddit from primarily an advertising-driven platform to a high-margin data utility, transforming its revenue model and investor perception.

Japanese Passive Flow Arbitrage Opportunity

The anticipated TOPIX float-ratio adjustment for Kioxia (285A.T), projected to increase from 15% to 50% in autumn 2026, is becoming a crowded institutional trade. This technical event is expected to force approximately ¥3 trillion in passive buying, as index funds adjust their holdings.

Funds are already strategically positioning themselves in "follow-on" plays, such as Tokyo Electron and Advantest, which are highly correlated to the expansion of memory infrastructure and are expected to benefit from this influx of capital.

Nuclear Power: The AI "New Oil"

The Trump administration's $200 million "AI Nuclear Initiative" underscores a massive and growing strain on the electric grid caused by increasing AI infrastructure demands. Oklo and X-Energy have been effectively anointed as the preferred providers for major data centers operated by companies like Microsoft and Nvidia.

This strategic move suggests that the "Physical AI" trade is extending beyond the data center wall, now encompassing the very source of power generation itself, highlighting the critical role of reliable energy in supporting AI's expansion.

Risks to AI's "Circular Financing" Model

While tech giants like Microsoft and Google are reporting strong Capital Expenditure (CapEx) for cloud infrastructure, negative user feedback on generative AI tools, such as Microsoft CoPilot, is steadily growing. If substantial enterprise efficiency gains do not materialize by the first half of 2027—a period when AMD’s MI450 GPUs are slated for widespread deployment targeting 2GW of power—the "circular financing" model previously discussed could face a sudden liquidity pull.

This scenario poses a significant risk to the sustainability of current AI growth projections, as the economic return on large-scale AI investments becomes increasingly scrutinized.

Happy Alpha Hunt! - Distilla

Disclaimer: This content is generated using AI, synthesizing public data (filings, reports, news) and social media (Reddit, X). It may contain errors, inaccuracies, or hallucinations. Nothing herein constitutes financial advice. This newsletter is for informational purposes only; please consult a qualified professional and conduct your own due diligence before making any investment decisions.