The AI selloff is spreading beyond a few tech leaders. Markets are rotating from highly valued AI pioneers to firms with nearer-term cashflows and control over pricing.
Pressure has been most visible in software, wealth management and brokers, insurance, logistics and transport, and real estate services. These areas face risks from AI tools that can lower fees, compress margins, and replace manual processes.
Us Downstream Exposure
The US market has more exposure to downstream AI such as apps, software, and services. This has led to closer scrutiny of monetisation, capital spending payback, and the risk that AI shifts value to customers.
Asia has more exposure to upstream AI infrastructure such as memory, foundries, and assembly and packaging. Demand for physical build-out can support these areas even when service models face disruption risk.
Within Asia, Korea and Taiwan are most linked to AI hardware cycles, supply tightness, component pricing, and utilisation rates. Japan is more tied to AI adoption in industry and enterprise, including automation, robotics, sensors, and process upgrades.
Diversification comes with caveats. Asia indices can be top-heavy, with Taiwan and Korea influenced by a small number of large chip-related names.
Risks And Drawdowns
Asia can still fall in global risk-off moves, broad tech drawdowns, or semiconductor down-cycles. Asia-listed software and IT services also fell alongside US peers during the selloff.
The current selloff is not a signal that the AI boom is over; it is a rotation. We are seeing a clear shift away from US-based software and service companies toward the Asian upstream hardware and component makers. This is a move from businesses facing questions about AI disruption to those providing the essential infrastructure for it.
This performance divergence has become quite clear in the first seven weeks of 2026, with the Nasdaq 100 falling approximately 8% while Taiwan’s TAIEX index has managed a 3% gain. This follows the period in 2025 when US downstream AI stocks were priced for perfection, leaving them vulnerable to this scrutiny. The market is now rewarding the tangible cash flows of the AI enablers over the more speculative promises of AI applications.
For traders, this suggests setting up pairs trades that capture this rotation. We should consider buying call options on ETFs focused on Korean and Taiwanese semiconductors while simultaneously buying puts on US software-as-a-service indices. This structure is designed to profit from the widening gap between the upstream “picks and shovels” and the downstream service models facing margin pressure.
The increased “dispersion” within the US market also presents an opportunity in volatility. Implied volatility on specific US software and financial services stocks is likely to rise more than that of the broader S&P 500. The CBOE Volatility Index (VIX) has already climbed above 22, and we can use this environment to trade relative volatility between sectors.
Drilling down into Asia, Korea and Taiwan offer the purest exposure to the AI infrastructure buildout. Recent industry reports confirm that prices for high-bandwidth memory (HBM) chips continued to climb in January 2026, directly benefiting the major Korean manufacturers. This pricing power confirms their resilience even as US software names are being questioned on their ability to monetize AI.
Japan presents a different, arguably more stable, opportunity for AI exposure. The theme there is less about building the core components and more about industrial adoption and automation to offset labour shortages. A good strategy here could be selling puts on Japanese industrial automation or robotics ETFs, collecting premium on the view that their businesses face less immediate disruption risk.
However, we must manage the risks inherent in the Asian trade, particularly the concentration risk. With Taiwan Semiconductor Manufacturing Company now accounting for over 30% of Taiwan’s main stock index, a single company’s performance can dominate returns. To mitigate this, we could hedge a long position in Asian markets by buying puts on a global semiconductor ETF, protecting against a broader cyclical downturn in the chip sector itself.
