BNP Likes Chinese Tech Stocks as AI Rotation Play Unfolds

BNP Paribas AM portfolio manager Sophie Huynh argues tech stocks offer value in China, citing a potential lag to the US AI frontier models of about six months. She frames the outlook around regional divergence (US vs China vs Europe) and suggests China could be competitive without being at the US frontier immediately. The comments are primarily positioning/strategy rather than a company or policy catalyst.

Analysis

The investable takeaway is not that China has caught up, but that the market may be underpricing how fast sentiment can re-rate if investors stop treating the gap as permanent. That is a relative-multiple story first and an earnings story later: China internet/cloud names such as BABA, BIDU, and TCEHY can rerate on any evidence that AI demand is moving from training to deployment, while the US frontier leaders keep their fundamental edge but lose some of the narrative premium.

Second-order, a credible China catch-up thesis is mildly negative for the scarcity premium embedded in US AI software and platform names because open-source and local deployment tend to commoditize inference faster than bulls assume. It also leaves Europe in an awkward middle ground: no frontier-model leadership, weaker domestic AI monetization, and limited direct exposure to the China re-rating trade, which argues for continued underperformance in European tech proxies and China-sensitive cyclicals unless earnings inflect.

The key risk is timing. A six-month gap can persist for years if export controls, compute access, and weak domestic demand keep China from converting model progress into monetization. I would frame this as a 1-3 month positioning catalyst, not a 6-18 month certainty. Falsifiers: KWEB/BABA failing to hold breakout levels, or upcoming China cloud/earnings prints showing no AI revenue lift and no capex rebound.

Key Decisions for Investors

• Tactically long KWEB vs short QQQ on pullbacks; target 8-12% relative outperformance over 1-3 months if China AI re-rating broadens, with a stop if KWEB lags QQQ by ~5% from entry.
• Buy BABA or BIDU call spreads 1-3 months out to express China AI optionality with defined downside; use weakness after any headline-driven spike rather than chasing strength.
• Trim incremental exposure to crowded US AI winners in XLK/SMH into strength; the risk/reward has shifted from narrative expansion to multiple compression if China catch-up becomes more credible.
• Watch BNPQY only as a sentiment/AUM-flow beneficiary from renewed EM-China interest; no direct position unless we see hard evidence of inflow acceleration.