Nokia posts higher Q2 profit as AI infrastructure demand surges

Nokia reported Q2 net sales of €4.82B (+8% YoY) and comparable operating profit of €434M (+18% YoY), with margins improving to 9.0% from 8.3%. AI and cloud revenue more than doubled (+105% YoY), driven by Network Infrastructure growth (+12% constant currency) and a 20% jump in Optical Networks. Restructuring charges weighed results (operating loss of €50M), but AI order intake was €2.8B and expected to convert ~half into revenue over 12 months; 2026 guidance was slightly raised to €2.1B–€2.6B from €2.0B–€2.5B.

Analysis

The investable signal is not the top-line print; it is that AI/network spend is now showing up in order books with a plausible 2-4 quarter revenue lag. For NOK, that creates a cleaner bridge from hyperscaler capex to earnings than in prior cycles, but only if conversion holds and mix stays tilted toward optical/IP rather than low-margin legacy gear. The market should still discount restructuring noise until gross margin and operating margin prove the AI mix is accretive ex one-offs.

Second-order, the bigger beneficiaries of AI traffic growth are likely the most exposed interconnect names such as CIEN, ANET, and select optical suppliers before a diversified telecom vendor fully rerates. Nokia’s upside is more about being a lagged catch-up beneficiary if it keeps taking share in network infrastructure; the risk is that hyperscalers concentrate spend in in-house compute and custom networking, leaving vendors with headline orders but weak pricing power. For GOOGL, heavier infra spend is strategically positive for moat-building, but it can still cap multiple expansion if FCF and cloud margin expansion do not keep pace.

Contrarian view: the consensus may be underestimating how much of this is a timing shift rather than a durable earnings step-up. If Nokia’s conversion rate slips below roughly 40% or 2026 guidance stops moving higher, the stock should be treated as a false breakout. Time horizon matters: the next 1-3 months are about sentiment and order-follow-through; 6-18 months are about whether AI infrastructure remains a high-ROIC spend category or devolves into a capital-intensive race.

Key Decisions for Investors

• Long NOK on 3-6 month horizon, ideally on any pullback post-earnings; use a call spread if you want defined risk. Base case is 10-15% upside if AI/cloud orders convert as guided and margin noise fades; invalidate on weaker-than-expected backlog conversion or any 2026 guidance stall.
• Pair trade: long NOK / short GOOGL into the next hyperscaler capex update, small size. Thesis is supplier revenue recognition versus spender FCF drag; risk/reward improves if Google keeps raising infra spend without clear cloud-margin expansion over the next 1-2 quarters.
• Do not chase GOOGL strength solely on AI capex headlines; prefer to add only after the next quarter if monetization metrics improve. If capex rises again and cloud profitability does not, the market can re-rate the stock down on multiple compression rather than earnings.
• Watch CIEN and ANET as confirmation names; if they fail to respond to the same AI-capex theme, treat NOK’s move as company-specific rather than a durable sector inflection. Use that divergence as a falsifier for the broader AI-network thesis.