Nokia, the Finnish company founded as a paper mill in 1865, which became the world’s largest mobile phone manufacturer in the late 1990s, which was crushed by the iPhone and collapsed over the following decade into a telecom equipment business most investors had forgotten about, is in 2026 one of the best-performing stocks in Europe.
Shares have surged more than 140 percent year-to-date, making Nokia the fourth-best performer in the entire Stoxx Europe 600 index.
The stock is at its highest level since 2008. And the reason is not phones.
Bloomberg’s analysis published Tuesday frames what has happened to Nokia’s valuation as a puzzle that the market is still working out.
The AI and cloud business driving most of the excitement around Nokia’s stock accounted for just 8 percent of the company’s total sales in the first quarter of 2026.
Yet the stock’s forward price-to-earnings ratio has more than doubled, from approximately 17 times at the start of the year to approximately 36 times now.
The market is valuing Nokia as an AI infrastructure company in the present tense. The actual AI infrastructure revenue is 8 percent of the business.
The question Bloomberg and every analyst who has upgraded the stock is trying to answer is whether that 8 percent becomes 20 percent, 30 percent and eventually the dominant part of Nokia’s revenue base, or whether the multiple re-rating has already priced in growth that the fundamentals will take years to justify.
The October 2025 Investment That Changed Everything
Nokia’s transformation from forgotten telecom equipment company to one of Europe’s hottest AI infrastructure stocks has a specific starting point.
In October 2025, Nvidia invested $1 billion in Nokia at $6.01 per share, taking approximately a 3 percent stake in the company.
The investment was not a passive financial bet. It was a strategic partnership announcement built around a specific technology called AI-RAN.
AI-RAN, AI Radio Access Networks, is a system that combines AI workloads with wireless network infrastructure.
The concept is that the computing resources in cellular base stations, which have historically been used only to manage wireless communications, can also be repurposed to run AI inference tasks during periods of low network demand. Nokia and Nvidia are building this system together.
T-Mobile has already signed on as the first deployment partner.
When Nvidia announced the $1 billion investment and the AI-RAN partnership, investors immediately reframed how they thought about Nokia.
The company had been known as a traditional telecom equipment supplier, important but unglamorous, trading at telecom equipment multiples rather than AI multiples.
The Nvidia investment signaled that Nokia was not just a supplier to telecom operators but a potential partner in building the AI-native infrastructure of the next decade.
The stock went from $6.50 to where it trades today.
The subsequent months added evidence that the reframing was not purely speculative.
Nokia acquired Infinera, the California-based optical networking company whose equipment is critical to the data center interconnects that AI workloads require.
It opened an AI Networking Innovation Lab in Sunnyvale, California, a physical facility dedicated to data center networking for AI training and inference, with partners including AMD, Keysight, Lenovo, Supermicro and others working under a shared open-standards framework.
It reported Q1 2026 earnings showing 49 percent growth in AI and cloud revenue year over year, with nearly $1.16 billion in new orders primarily from optical networks.
The wave of analyst upgrades that followed in May 2026 was consequential.
CFRA upgraded Nokia from Hold to Buy and more than doubled its target to $16, explicitly beginning to value the company more like an optical networking and AI infrastructure peer rather than a traditional telecom equipment supplier.
Argus moved to Buy with a $15 target citing AI-related demand. JPMorgan, Morgan Stanley, Deutsche Bank, Arete and Nordea all lifted targets or shifted to more bullish positions in the same period.
When that many analysts move simultaneously in the same direction, the market tends to respond.
What Nokia Actually Makes And Why AI Needs It
The specific products driving Nokia’s AI narrative are not phones, not mobile networks in the consumer sense and not anything that the public-facing Nokia brand of the 1990s and 2000s would be recognized by.
They are optical networking equipment, the high-speed fiber optic cables, amplifiers, switches and routing hardware that physically connect data centers to each other and connect the GPUs inside those data centers to storage and to the internet.
Optical networking is the physical infrastructure that makes AI possible at scale.
A data center containing thousands of Nvidia H100 or Blackwell GPUs cannot function as an AI training system unless those GPUs can communicate with each other at extraordinarily high speeds and low latency.
The optical interconnects that carry data between GPUs, between servers and between data centers are as essential to AI infrastructure as the chips themselves, they are just less visible and less famous.
Nokia is one of the major global suppliers of this equipment. As AI data center buildout has accelerated, Nvidia guided Q2 revenue to $91 billion, Meta and Microsoft have committed to $60 to $80 billion annually in capital expenditure, the demand for optical networking equipment has accelerated proportionally.
Every new data center that gets built, every hyperscaler expansion that adds GPU capacity, creates additional demand for the optical infrastructure Nokia sells.
The company reported optical networks revenue growing 20 percent in Q1 2026, and it has raised its guidance for the combined optical and IP networks segment to 18 to 20 percent growth for 2026, up from a prior range of 10 to 12 percent.
Nokia management believes that hyperscalers will spend more than $700 billion as cloud and AI customers scale infrastructure for AI workloads.
The market for AI and cloud networking is expected to grow at approximately 27 percent annually from 2025 to 2028.
The Valuation Puzzle Bloomberg Is Describing
The Bloomberg framing of Nokia’s valuation as a puzzle captures the specific tension that makes the stock interesting and uncertain simultaneously.
If Nokia’s AI and cloud business continues growing at 49 percent annually and represents 8 percent of current revenue, the mathematical path to it becoming 20 or 25 percent of a larger revenue base is real but takes time.
At the current growth rate and base, Nokia’s AI-adjacent revenue roughly doubles in less than two years.
If the optical networking market grows at 27 percent annually as projected, the underlying demand is there to support that trajectory.
The puzzle is whether 36 times forward earnings is the right price to pay for that trajectory today.
At the start of 2026, Nokia was trading at 17 times forward earnings, a multiple consistent with a telecom equipment company with some AI optionality attached.
At 36 times, the market is paying an AI infrastructure multiple for a company where AI infrastructure is currently 8 percent of sales.
Amanda Lyons, head of research at Energy Group Capital, put the investor position precisely in Bloomberg’s analysis. “The easy re-rating is gone. The interesting question is whether there is a second leg.”
The first re-rating, from telecom equipment to AI infrastructure story, has already happened. The 140 percent gain year-to-date reflects it.
The second leg, if it exists, requires Nokia’s AI revenue to grow into the multiple the market has already awarded.
The European AI Scarcity Premium
Part of Nokia’s extraordinary performance in 2026 reflects something specific about European equity markets that J.P. Morgan’s head of cross-asset strategy described precisely. “In Europe, scarcity amplifies the trend. There are few large, liquid AI pure-plays, so flows concentrate in a small group of perceived AI proxies.”
European investors who want AI exposure have a much smaller universe of liquid, large-cap options than American investors do.
The AI boom’s most direct beneficiaries, Nvidia, Microsoft, Meta, Google, Amazon, are all American companies. European institutions with mandates to hold European equities, or European investors who prefer not to deal with the currency and regulatory dimensions of American market exposure, have limited options.
Nokia, at its current scale and with its Nvidia partnership and optical networking growth story, is one of the few credible AI-adjacent names available in European markets.
That scarcity dynamic does not invalidate Nokia’s fundamental story. But it does help explain why the stock has responded so dramatically to what is, by objective measure, an AI business that represents 8 percent of current revenues.
In a market with limited options, that 8 percent and its growth trajectory carries a premium that the same 8 percent would not command in a universe with dozens of pure-play AI infrastructure options.
Nokia was founded as a paper mill. It became the world’s dominant mobile phone maker. It lost that position and spent a decade rebuilding in telecom equipment.
In 2025, Nvidia invested $1 billion in it to build AI radio networks. In 2026, it is up 140 percent. The puzzle is whether it keeps going from here.
