The AI selloff is now spreading beyond a handful of tech leaders. Markets are moving away from expensive AI pioneers and toward companies with nearer-term cash flow and stronger pricing power. The pressure is clearest in software, wealth management and brokers, insurance, logistics and transport, and real estate services. In these sectors, AI tools can cut fees, squeeze margins, and replace manual work.

Us Downstream Exposure

The US market has more exposure to downstream AI, such as apps, software, and services. This is driving tougher questions about monetisation, payback on AI spending, and whether AI shifts value away from companies and toward customers. Asia has more exposure to upstream AI infrastructure, such as memory, foundries, and assembly and packaging. Demand for physical build-outs can support these areas, even when service-based models face disruption. Within Asia, Korea and Taiwan are most tied to AI hardware cycles, supply tightness, component pricing, and factory utilisation. Japan is more tied to AI adoption in industry and enterprise, including automation, robotics, sensors, and process upgrades. Diversification also has limits. Many Asian indices are top-heavy, and Taiwan and Korea can be driven by a small number of large chip-related stocks.

Risks And Drawdowns

Asia can still decline during global risk-off moves, broad tech selloffs, or semiconductor down-cycles. Asia-listed software and IT services also fell with US peers during this selloff. This selloff does not mean the AI boom is over. It looks more like a rotation. We are seeing a shift away from US software and service companies and toward Asian upstream hardware and component makers. In other words, investors are moving from businesses that may be disrupted by AI to businesses that supply the infrastructure AI needs. This split has been clear in the first seven weeks of 2026. The Nasdaq 100 is down about 8%, while Taiwan’s TAIEX is up about 3%. This follows 2025, when US downstream AI stocks were priced for perfection and became vulnerable once scrutiny increased. The market is now favouring tangible cash flows from AI enablers over more speculative promises from AI applications. For traders, this rotation could be expressed through pairs trades. One approach is to buy call options on ETFs focused on Korean and Taiwanese semiconductors while also buying puts on US software-as-a-service indices. The goal is to benefit if the gap keeps widening between upstream “picks and shovels” and downstream service models under margin pressure. Rising “dispersion” in the US market also creates a volatility opportunity. Implied volatility in specific US software and financial services stocks may rise more than volatility in the broader S&P 500. The CBOE Volatility Index (VIX) has already moved above 22, which can support relative-volatility trades between sectors. Within Asia, Korea and Taiwan offer the most direct exposure to the AI infrastructure build-out. Recent industry reports show high-bandwidth memory (HBM) chip prices kept rising in January 2026, which directly supports major Korean manufacturers. This pricing power suggests resilience, even as US software companies face pressure to prove they can monetise AI. Japan offers a different, and often steadier, way to gain AI exposure. The focus is less on core components and more on industrial adoption and automation to address labour shortages. One possible strategy is selling puts on Japanese industrial automation or robotics ETFs to earn premium, based on the view that these businesses face less near-term disruption. Even so, the Asian trade has meaningful risks, especially concentration risk. Taiwan Semiconductor Manufacturing Company now makes up more than 30% of Taiwan’s main stock index, so one company can drive overall returns. To reduce this risk, you could hedge a long Asia position by buying puts on a global semiconductor ETF, which helps protect against a wider chip-sector downturn. Create your live VT Markets account and start trading now.

here to set up a live account on VT Markets now