Analysis-With tariffs stalled, Trump’s China policy drifts

Trump’s China strategy remains inconsistent despite tariffs that briefly reduced the U.S. goods trade deficit with China by 32% to $202 billion in 2025. The article highlights repeated policy reversals on AI chip exports, military blacklists, export controls, and port fees, alongside a February Supreme Court ruling that invalidated many tariffs. Beijing has not materially changed course, while U.S. credibility and coordination appear weakened heading into Trump’s planned May 14-15 China visit.

Analysis

The market implication is not “more tariffs” but a higher probability of policy whiplash becoming the baseline state. That is structurally bullish for incumbents with diversified supply chains and pricing power, and bearish for companies whose capex decisions depend on stable export-control regimes. In practice, the biggest hidden winner is not the obvious domestic manufacturing basket; it is firms with optionality to reroute production and re-source inputs without having to make binary China/U.S. commitments.

For semis, the key second-order effect is that inconsistent licensing raises the value of gray-area revenue while simultaneously increasing the discount rate applied to China exposure. That hurts names with concentrated advanced-node demand and helps the ecosystem operators that monetize tooling, packaging, and services across geographies. The signal is also mixed for AI capex: even if sales are permitted episodically, the risk premium on China-bound revenue should widen, which tends to compress multiple expansion in the most China-levered platforms before any actual unit impact shows up.

The other important read-through is that “managed trade” reduces the probability of a clean reshoring supercycle. If policy can be reversed within weeks, corporates will hedge with inventory and dual-sourcing rather than commit to domestic capacity, which means the inflationary benefit to U.S. industrials may be smaller and slower than consensus expects. Over the next 3-6 months, the more tradable setup is around headline volatility and licensing decisions; over 12-24 months, the risk is institutional erosion of U.S. credibility, which is harder to model but likely negative for multinational valuations tied to China growth assumptions.

Key Decisions for Investors

• Maintain / add to short NVDA into any rally tied to China-access headlines; use 1-3 month put spreads to express the view. Risk/reward favors the short because incremental upside from approvals is episodic, while downside from another policy reversal is immediate and multiple-compressing.
• Pair trade: long AVGO / short NVDA over 3-6 months. AVGO has more diversified end-demand and less single-point China policy risk; NVDA carries higher headline beta and a wider distribution of outcomes if export enforcement tightens again.
• Buy downside protection on semiconductor ETF exposure via SOXX 3-6 month puts rather than outright shorting a single name. The catalyst is not earnings deterioration but policy inconsistency, so index hedges are cleaner and less vulnerable to idiosyncratic announcements.
• Fade pure-play U.S. reshoring beneficiaries in the next 6-12 months unless they have visible backlog conversion. The market is likely overpricing immediate capex repatriation; use any post-headline bounce to short industrial suppliers with China-sensitive end markets.
• Avoid taking a directional long in AAPL on China-policy noise alone. The better trade is volatility selling around headline spikes only if you can hedge with component or handset-supply chain exposure, because the article implies uncertainty rather than a durable demand shock.