Substack作者重申看多诺基亚AI光网络与IP网络长期重估逻辑

Nokia has had rough price action over the last couple weeks.

But I believe the long term story is intact and the market id trying to figure out how to value the company.

There are two recent analyst calls that frame the debate pretty well.

Handelsbanken downgraded Nokia from Buy to Hold while raising its target to €12 from €10.20. That was a valuation call after the stock’s run. The thesis did not break. Their model just says the stock has mostly caught up.

Then Jefferies came out on the other side, maintaining Buy and raising its target to €13.80 from €10.70.

So, has Nokia caught up to the model?

Or is the model still too small?

I lean strongly toward the second one.

The reason is demand. Crazy demand. I think the market is still underestimating how much optical and IP infrastructure the AI buildout will require.

A normal model might say AI drives a step-up in optical demand, Nokia has a strong 2026, growth normalizes, and a €12 target looks fair.

I personally think that is way too conservative.

AI data centers are getting larger, more distributed, more power-constrained, more fiber-heavy, and more supply-constrained than a normal telecom or cloud cycle. That significantly changes the math for companies sitting in the network layer.

In 2025, Nokia’s Optical Networks business did roughly €3.2 billion of revenue. IP Networks did roughly €2.6 billion. Together, Optical plus IP was roughly €5.8 billion.

Then Q1 came through and was really the first proof point of the story.

AI and Cloud sales grew 49%. Nokia booked €1 billion of AI and Cloud orders in one quarter. Optical grew 20%. Management raised the 2026 growth assumption for Optical plus IP from 10-12% to 18-20%.

That bridges Optical plus IP to roughly €6.8-7.0 billion this year.

My old upside case had that business reaching €8-9 billion by 2028.

I think that is too low.

Let’s unpack why and what I am now underwriting.

*Educational and informational only. This report reflects my personal research, opinions, and interpretation of publicly available information. It is general market commentary and should never be treated as individualized financial advice, tax advice, legal advice, or a recommendation to buy, sell, hold, or short any security.*

## Why I Think The Demand Curve Is Bigger

The AI buildout is spreading because power and space are becoming real constraints.

This is the part I think the market is still underweighting. AI compute cannot all sit in one perfect building in one perfect location. The industry is being pushed into larger, more distributed infrastructure footprints.

JLL expects nearly 100 GW of new data center capacity between 2026 and 2030.

Meta is building its largest data center in Louisiana, with more than two gigawatts of compute capacity.

Stargate has already laid out a multi-site U.S. footprint approaching seven gigawatts of planned capacity.

Microsoft has described its Wisconsin and Atlanta AI data centers as connected sites working together as an AI superfactory across state lines.

When compute spreads across campuses, states, and power regions, the network between those sites becomes part of the AI machine.

That is where Nokia lives.

This is also why scale-across is such an important concept and why I keep hammering it home. The industry is building more and more connected AI factories. That requires optical transport, fiber density, line systems, routing, switching, and automation.

Then there is the supply-chain angle and strength in positioning that Nokia has. Infinera added optical scale. San Jose supports InP manufacturing. Allentown adds advanced test and packaging for photonic chips.

That is why I think a conservative model may miss the setup as it may treat 2026 as a peak year.

I think that 2026 is just the first visible year of a larger demand curve.

## Two Numbers To Separate

There are two lenses here before we unpack further, as they could get a bit confusing if you are looking at the numbers.

Optical plus IP is **what Nokia sells.**

AI and Cloud is **who Nokia sells to.**

Optical plus IP is the product engine. It includes the optical and IP networking products that move, route, and manage AI traffic.

AI and Cloud is the customer category. It tells us how much revenue and order activity is coming from AI and cloud buyers.

In Q1, Nokia did €4.5 billion of total sales. AI and Cloud was 8% of sales, or roughly €360 million of revenue. In that same quarter, Nokia booked €1 billion of AI and Cloud orders.

So again, revenue tells us what Nokia shipped and orders tell us what customers are committing to.

That is why the order number is the key signal.

If Nokia is already booking €1 billion of AI and Cloud orders in one quarter, the 2027 and 2028 revenue base may be larger than current models assume.

## The Revenue Bridge

Optical plus IP was roughly €5.8 billion in 2025.

The 2026 guide points toward roughly €6.8-7.0 billion.

My main question is where that goes by 2028.

My old view was €8-9 billion.

Now I think €10-12 billion is a more realistic upside range, with a bull case above that.

Optical is already working. AI and Cloud orders are well ahead of current revenue. Capacity is being added. IP is still early and can become the second leg.

IP is the biggest swing factor. It grew only 3% in Q1, while Optical grew 20%. If IP starts converting design wins into real data center networking revenue, Nokia is layering on multiple types of re-rates.

Also important to know that it’s possible that Nokia’s guidance for 2026 is conservative and that their guide will continue to move up. I believe that it will move up each quarter, but of course, take that with a grain of salt.

## The Margin Bridge

Nokia’s gross margins are stronger than a legacy label suggests. Group gross margin was 45.5% in Q1. Network Infrastructure gross margin was 43.4%. Mobile Infrastructure gross margin was 48.5%.

The gap is operating margin. Network Infrastructure operating margin was only 6.7% in Q1, well below cleaner comps like Ciena. But that is also where the opportunity sits. Nokia is still investing behind Infinera, San Jose, supply chain, product development, customer engagement, and IP wins.

Nokia’s target is for Network Infrastructure operating margin to reach 13-17% by 2028.

This is directional, because Network Infrastructure also includes Fixed Networks, but the math shows why the earnings bridge can get interesting.

At €8 billion of Optical plus IP revenue, a 13-17% margin framework implies roughly €1.0-1.4 billion of operating profit.

At €10 billion, that becomes roughly €1.3-1.7 billion.

At €12 billion, that becomes roughly €1.6-2.0 billion.

So the way I see it is that revenue gets the re-rate started and operating leverage keeps it going.

## The Scenario Framework

My conservative case is Optical plus IP reaching roughly €8-9 billion by 2028. That assumes the current ramp continues but normalizes.

My base case is roughly €10-11.5 billion. That assumes optical stays strong, system sales attach, capacity improves shipment conversion, and IP starts contributing more visibly.

My bull case is >€12 billion. That assumes multiple AI networking cycles overlap and IP becomes a real second leg. In this situation, everything needs to go right.

## What Moves The Model / Risks

The model moves higher if AI and Cloud orders stay strong after the €1 billion Q1 print, Optical holds above mid-teens growth, IP accelerates from 3%, San Jose ramps cleanly, the next optical roadmap gets customer validation, and margins move toward 13-17%.

It moves lower if orders get lumpy, Optical cools after the 800G wave, IP takes too long to convert, San Jose slips, supply blocks shipments, telecom drag offsets the AI-facing growth, or the market treats Optical like a cyclical spike.

So I will adjust accordingly.

But this is how I see it today given what we know and the state of the industry.

I think that this will remain a volatile stock, especially as traders pile in. But it is one of my preferred holdings with optical / networking exposure with a multi-quarter timeframe.

Also to important to remember that even if Nokia hits these numbers, we don’t know how the market will reward them with a multiple. It would make sense to me that as they actually become more of a pure-play Optical / Networking stock, that they will deserve a higher multiple. But there is always the risk that by the time that happens, the market has decided to reward lower multiples to these optics exposed companies. That is always a risk for any of these companies.

But I believe in the growth vision for Nokia, the position of the strength they are in, and the re-rate that should continue to come.

*Educational and informational only. This report reflects my personal research, opinions, and interpretation of publicly available information. It is general market commentary and should never be treated as individualized financial advice, tax advice, legal advice, or a recommendation to buy, sell, hold, or short any security.*

*I may own positions in companies discussed, including Nokia, and my views may change at any time. I may buy, sell, add to, reduce, or exit any position mentioned with no obligation to provide an update.*

*Forward-looking statements, estimates, scenario analysis, price targets, market-size assumptions, and company projections are inherently uncertain. Actual results may differ materially from the assumptions discussed. Readers should verify all material facts independently and consult a qualified financial professional before making any investment decision.*

*Do your own work.*
From story to proof