Goldman Lifts MLCC View, Says Capacitor Stocks Could Extend AI Upcycle to 2030
Goldman Sachs Research says “capacitor stocks” could benefit from an unusually long AI data-center build cycle, led by multi-layer ceramic capacitors (MLCCs). The bank revised its 2026 MLCC pricing forecast from flat to a 0–5% increase and now expects the AI-driven upcycle to run to around 2030 (vs. a prior 2028 endpoint).
Goldman met Murata Manufacturing, a major MLCC supplier, and reaffirmed a Buy rating on Murata with a target price of 5,400 yen. It also rated Nantong Jianghai Capacitor Co. (China) as Buy, citing direct exposure to data-center power infrastructure demand.
Fundamental demand backdrop: Goldman projects global power demand could rise 220% by 2030, up from 175%, as AI hyperscalers accelerate capital expenditures. The article notes hundreds of billions of dollars in commitments from Microsoft, Google, Amazon, and Meta, with new data centers requiring large electrical systems that use capacitors.
Investor angle: data-center electrical component stocks have already outperformed hyperscalers by ~85 percentage points since early 2025. Goldman argues upside may still remain because the upcycle is expected to extend several more years.
Key risks: MLCC price gains are modest, so if AI spending slows, the upcycle thesis could weaken quickly. Geopolitical tensions and supply-chain uncertainty add pressure for Chinese component makers via tariffs or export controls.
Overall, Goldman’s call reframes “capacitor stocks” as a longer-duration theme tied to AI infrastructure spending through 2030.
Bullish
This is bullish for risk sentiment around AI infrastructure supply chains, though it’s indirect for crypto markets. Goldman’s thesis implies a multi-year demand tailwind for MLCC “capacitor stocks” tied to hyperscaler capex—so traders typically react positively to durable AI trade narratives. The article notes that component/electrical stocks already outperformed hyperscalers by ~85 percentage points since early 2025; that raises the risk of near-term “good-news is priced in” pullbacks, but extending the upcycle to ~2030 can reduce the speed at which the narrative fades.
Short-term: headlines from large sell-side banks often trigger momentum trades in AI-adjacent equities and may slightly lift broader “tech/AI” sentiment, which can spill into crypto via correlation (risk-on periods). However, because the forecast is specific (0–5% pricing) and dependent on sustained capex, any sign of AI spending slowdown would likely cause faster sentiment reversals.
Long-term: if hyperscalers indeed keep investing through 2030, it supports a sustained industrial/semiconductor theme. In past market cycles, long-duration capex narratives (e.g., AI/compute buildouts) tended to attract consistent capital flows, supporting broader risk assets. For crypto traders, this usually matters less as a direct fundamental input and more as a macro/liquidity sentiment signal—bullish if it reinforces risk-on, neutral-to-bearish if geopolitical or supply-chain shocks dominate.