AI Posts Strong Q2 Returns as Inference Economy Takes Hold

The artificial intelligence investment landscape reached a critical inflection point in Q2 2025, with the ROBO Global Artificial Intelligence Index (THNQ) delivering exceptional returns of 24.4%. Those returns significantly outpaced broader market performance. All 10 subsectors posted gains, reflecting broad-based AI adoption as the market transitions from infrastructure buildout to operational deployment, a shift we call the “inference economy.”

Broad-Based AI Rally Validates Investment Thesis

The quarter’s performance tells a compelling story of AI maturation. While the VettaFi Developed World Index posted solid gains of 11.5%, THNQ’s outperformance reflects the accelerating monetization of AI technologies across the entire value chain. Every single subsector within the index posted positive returns, indicating that AI adoption has moved well beyond experimental phases into sustained commercial deployment.

This broad-based strength validates our long-held investment thesis that sustainable AI growth requires a diversified ecosystem approach. Rather than concentrating on a handful of mega-cap winners, THNQ’s strategy of capturing value across semiconductors, cloud infrastructure, security, and emerging technologies proves prescient as each segment contributes meaningfully to overall performance.

Sector Leadership Reveals Market Evolution

Network & Security Takes the Lead

Leading the charge was network & security (19.6% portfolio weight), which surged 28.5% during the quarter. Strategic new additions proved particularly rewarding, with Astera Labs climbing 51.5% and CrowdStrike advancing 44.5%. This sector’s outperformance reflects growing enterprise recognition that traditional security architectures are inadequate for AI-native environments.

“The growing emphasis on zero-trust frameworks and edge computing security creates a compelling multi-year growth trajectory for these holdings,” noted the ROBO Global research team. As enterprises scale their AI deployments, the critical need for AI-native security architectures becomes increasingly apparent, positioning this sector for sustained outperformance.