The semiconductor sector has become shorthand for the AI boom, and Invesco PHLX Semiconductor ETF (NASDAQ:SOXQ) sits at the center of that trade. Trading at $64.24 per share as of February 9, 2026, the fund has surged 64% over the past year as investors bet on AI infrastructure spending. That momentum carried into 2026 with a 15% year-to-date gain, though it also exposes holders to cyclical chip demand swings.
SOXQ has scaled rapidly since its 2021 launch, reaching $1 billion in assets as investors sought concentrated semiconductor exposure during the AI buildout. The fund tracks the full chip value chain from design to manufacturing equipment, offering broad sector exposure at a competitive cost structure that undercuts many rivals.
AI Capex Cycles Will Drive the Next Leg
The macro factor that matters most for SOXQ is the pace of AI infrastructure spending by hyperscalers like Microsoft, Amazon, and Google. These companies are collectively spending hundreds of billions on data centers, and semiconductors are the core input. When capex budgets expand, demand for GPUs, high bandwidth memory, and advanced packaging surges. When budgets tighten or utilization disappoints, chip orders flatten quickly.
SOXQ concentrates heavily in AI accelerator leaders. NVIDIA (NASDAQ:NVDA | NVDA Price Prediction) alone represents 10.75% of the portfolio, while Advanced Micro Devices (NASDAQ:AMD) and Broadcom (NASDAQ:AVGO) add another 14.82% combined. This trio serves as a direct proxy for hyperscaler AI spending, meaning the fund’s performance rises and falls with data center capex cycles.
Watch quarterly earnings calls from cloud providers for commentary on capital intensity and AI workload growth. If spending plans moderate or utilization metrics disappoint, semiconductor equipment and memory suppliers will feel it first. Any acceleration in AI model training or inference deployment ripples through the entire supply chain. Monitor the Semiconductor Industry Association’s monthly sales reports and equipment billings data from SEMI for early demand signals.
Concentration Risk in the Top Ten
Concentration defines SOXQ’s risk profile, with the top 10 holdings commanding 59% of assets. NVIDIA alone represents over 10% of the portfolio, meaning the fund’s performance is directly tied to a handful of AI accelerator leaders. This concentrated approach amplifies both upside during AI spending booms and downside when chip demand cycles turn.
The fund provides exposure across the value chain, including equipment makers like Lam Research (NASDAQ:LRCX) and Applied Materials (NASDAQ:AMAT), but those names are also cyclical and sensitive to fab spending cycles.
Check Invesco’s monthly fact sheet for meaningful shifts in top holdings or sector weights. Rebalancing can alter exposure to specific sub-segments like analog, memory, or design.
The biggest factor for the next 12 months is whether AI capex spending sustains its current trajectory, while the key micro risk is concentrated exposure to a small group of high-performing but cyclical chip leaders.