AI money, tax battles, and data-center backlash: who’s winning the next market shift?

Hedge funds entered the second quarter “all-in” on the AI trade, but after summer volatility triggered a massive portfolio cleanup, they are now diversifying rather than doubling down. Market coverage indicates the shift is not a retreat from AI so much as a rotation: investors are adding exposure to healthcare, energy, and financials while memory and other AI-adjacent themes struggle to regain momentum. In parallel, “smart money” appears to be moving on from memory stocks after their strong first-half run, with concerns about the broader AI trade’s health weighing on sentiment despite solid fundamentals. Separately, Bloomberg analysis and a podcast segment argue that parts of the AI industry misread public backlash, particularly around data centers, suggesting a growing gap between hyperscaler expansion plans and community acceptance. Geopolitically, the cluster points to a new phase of the AI economy where capital allocation, regulatory design, and local social license all interact. Hong Kong’s attempt to attract global money managers by introducing legislation to erase levies on carried interest has drawn a direct competitive response from Singapore, highlighting how financial hubs are using tax policy to win talent and assets. This “tax competition” matters because it can re-route where management fees and deal flow concentrate, influencing regional capital markets and the political bargaining power of financial regulators. Meanwhile, the Minnesota story shows that even when the technology is global, deployment constraints are local: Google’s “charm offensive” in two towns signals that hyperscalers face reputational and permitting risks that can slow capacity additions and reshape investment timing. The winners are likely those who can align AI infrastructure build-outs with community and political constraints, while the losers may be memory and data-center supply chains that depend on uninterrupted scaling. Market implications cut across semiconductors, healthcare, energy, and financials. Memory stocks are described as struggling for momentum, implying downside pressure on the AI hardware complex even as fundamentals remain supportive, a pattern consistent with investors de-risking after volatility. The hedge-fund rotation toward healthcare, energy, and financials suggests relative outperformance potential in those sectors versus pure-play AI hardware, with possible near-term flows into defensive growth and cash-flow generators. On the policy side, carried-interest tax changes in Hong Kong—and the competitive counter-move from Singapore—can affect private capital formation, which typically influences leveraged finance, asset management revenues, and risk appetite. For investors tracking AI exposure, the key instruments to watch are memory-related equities and broader AI supply-chain baskets, alongside sector ETFs that capture the rotation away from the most crowded trades. Next, investors and policymakers should watch whether the AI trade’s “state of play” improves enough to pull capital back into memory and other high-beta components, or whether diversification becomes structural. For the data-center front, the trigger is whether community opposition in Minnesota translates into delays, stricter zoning, or higher compliance costs that force hyperscalers to revise capex schedules. On the financial-hub side, the key signal is the legislative timeline and implementation details for Hong Kong’s carried-interest levy repeal, and whether Singapore escalates with additional incentives or regulatory adjustments. Finally, the market’s immediate stress test will be whether “smart money” continues to rotate out of memory momentum or finds a new entry point after volatility subsides. Escalation would look like renewed volatility plus evidence of deployment delays; de-escalation would be smoother permitting outcomes and clearer policy certainty for capital formation.

Geopolitical Implications

• 01

  AI competitiveness is increasingly shaped by domestic regulatory and social constraints, not just global technology cycles.

• 02

  Financial-hub tax competition (carried interest) can shift where management talent and deal flow concentrate, altering regional influence in capital markets.

• 03

  Hyperscaler expansion faces a “social license” bottleneck in the US, which can delay infrastructure build-outs and affect supply-chain timing.

• 04

  Market volatility is translating into cross-sector reallocation, potentially reducing the political leverage of pure-play AI hardware narratives.

Key Signals

• —Legislative progress and effective dates for Hong Kong’s carried-interest levy repeal and Singapore’s counter-incentives
• —Memory-sector earnings guidance and inventory commentary indicating whether demand fears are easing
• —Permitting outcomes and timelines for data centers in Minnesota (zoning, environmental review, utility interconnection)
• —ETF/flow data showing whether capital continues rotating into healthcare/energy/financials or re-concentrates into AI hardware