Nokia ($9.15) Jumps 8.4% as €2.8B AI Order Intake Offsets a Negative €732M Cash Quarter — Path to $11

Nokia shares are trading at $9.145, up roughly $0.71 or 8.4% from Wednesday's close of $8.41, with an intraday range of $9.04 to $9.34. That follows a 6.05% decline on Wednesday that took the ADR to $8.39 on a day when the broader tape fell hard — the Dow shed 1,153 points and the Nasdaq-100 entered correction. Today's move tracks the semiconductor and AI-infrastructure rebound, with the semiconductor ETF up more than 7% and Microsoft's 14% surge on 43% Azure growth reviving the entire capex-linked complex.

The bounce recovers a fraction of what has been lost. Nokia peaked at $17.45 inside the last twelve months and traded near $13.49 as recently as late June. It fell roughly 25% across thirty days to $10.12 by mid-July, then dropped a further 7.1% on the July 23 earnings release, and kept sliding through the following week. From the 52-week high, the stock is down approximately 48%. From the $4.00 low at the other end of that range, it is still up 129%. Both facts describe the same twelve months, and the gap between them is why this name has become one of the most contested large-cap positions in European technology.

Market capitalisation sits near $48 billion at Wednesday's close, or roughly €57 billion, on 5.74 billion shares outstanding. Trailing earnings per share of $0.15 puts the trailing multiple around 57 at $8.39 and roughly 61 at today's $9.15. Average daily volume runs near 94 million shares — heavy enough that the post-earnings selling was genuine distribution rather than thin-tape drift. The forward dividend of $0.163 yields about 1.8%, with the most recent ex-date on July 28.

The setup is unusual and worth stating plainly. This is a company that beat on adjusted earnings, grew net sales 9%, expanded gross margin, doubled its AI and cloud revenue, booked €2.8 billion of AI-related orders, and guided above the midpoint of its full-year profit range — and the stock fell 7% on the day and kept falling. That disconnect is the entire subject of this forecast. Either the market is pricing something the headline numbers do not capture, or a crowded AI-adjacent position is being unwound for reasons that have nothing to do with Nokia's operations. The evidence points toward both being true simultaneously.

The Q2 Print: €4.82 Billion, Nine Percent Growth, and a GAAP Line at Zero

Nokia reported second-quarter 2026 results before the open on July 23. Net sales came in at €4.82 billion, up 9% year over year on a constant-currency basis and 8.3% as reported — narrowly short of one consensus figure by about €10 million while beating others depending on the currency basis used. Comparable earnings per share landed at €0.07, a cent ahead of expectations, which translated to $0.08 on the ADR against a $0.07 estimate, a beat of roughly 17%. The comparable year-ago figure was $0.04, so adjusted earnings doubled.

The margin lines were the constructive part. Comparable gross margin expanded 70 basis points to 46%, reflecting optical network growth and the integration of Infinera. Comparable operating profit reached €434 million with the operating margin also up 70 basis points to 9.0%. Management flagged that the quarter benefited from some software revenue recognition landing in the second quarter rather than the third — a timing pull-forward that flatters the comparison and needs to be backed out of any run-rate assumption.

The reported line is where it breaks. GAAP earnings per share printed at €0.000 against a €0.051 expectation, and the reported operating margin swung to negative 1.0% against the positive 9.0% comparable figure. That ten-percentage-point gap is entirely restructuring, and it is the single number that explains the share price reaction. A comparable margin tells you what the business earns in a steady state. A reported margin tells you what it earned this quarter. When the two diverge by a thousand basis points, the market discounts the former and prices the latter until the charges stop.

Segment performance was cleanly split. Network Infrastructure delivered 12% net sales growth led by Optical Networks and IP Networks, driven by hyperscaler and AI-related demand. That is the segment carrying the entire story. Mobile Infrastructure — the traditional radio access business that was Nokia's identity for two decades — is the drag, operating in a 5G market that is past its capex peak with 6G still years out. The 2026 reorganisation into two segments, Network Infrastructure and Mobile Infrastructure, was designed precisely to make this divergence visible to investors. It has succeeded, and the visibility is not flattering to half the company.

AI and Cloud at €446 Million: The 105% Number That Carries the Story

The disclosure that matters most is AI and Cloud net sales of €446 million in the quarter, up 105% year over year. That is the line item Nokia has built its entire repositioning around, and it more than doubled. For context, the equivalent figure grew 49% year over year in the first quarter with roughly €1 billion of new orders from the segment alone — so the growth rate itself accelerated by more than fifty percentage points across a single quarter.

The absolute scale is the caveat that keeps the stock honest. At €446 million against total quarterly net sales of €4.82 billion, AI and Cloud represents roughly 9% of revenue. Doubling a segment that constitutes under a tenth of the business contributes about four and a half points of consolidated growth — meaningful, and consistent with the 9% headline, but not transformative on its own. The mathematics of the bull case require this line to compound at triple-digit rates for several more years before it dominates the mix, and compounding at that rate off a larger base is materially harder.

What makes the disclosure credible rather than promotional is the composition. This is not a software licensing line or a services attach rate. It is optical networking and IP routing sold into data centre interconnect — physical infrastructure that hyperscalers must buy to move training and inference traffic between facilities, priced against a competitive set that includes some of the largest networking franchises in the industry. Nokia is winning that business on the strength of an asset it bought rather than built, which is a different kind of risk than execution risk on an organic product roadmap.

The chief executive's framing on the earnings call was notably unhedged. He said he is less concerned about an AI bubble because demand remains strong and supply constraints persist, and characterised the quarter as continued progress against the strategy set at the Capital Markets Day, with the team focused on maximising the opportunity in what he called the AI supercycle. That is a management team leaning into the narrative rather than moderating it — which is a source of upside if the demand is real and a source of considerable downside if hyperscaler capex decelerates in 2027.

The €2.8 Billion Order Book Is the Real Disclosure

Buried beneath the revenue headline is the number that should carry the most weight in any forward model: AI and cloud order intake of €2.8 billion in the quarter. Set that against €446 million of AI and Cloud net sales recognised in the same three months and the ratio is more than six to one. That is a book-to-bill figure that implies the revenue line has years of contracted conversion ahead of it before new orders are even required.

The first-quarter comparison sharpens it. That quarter produced roughly €1 billion of new AI-segment orders. The second quarter produced €2.8 billion. Order intake nearly tripled sequentially, which is a far stronger signal than the revenue growth rate because orders lead revenue and are considerably harder to manufacture through timing or accounting choices. If the revenue line is the evidence that the strategy is working, the order line is the evidence that it will keep working through 2027.

This is also the specific metric that separates Nokia's situation from the broader AI-infrastructure de-rating currently underway. The market has spent July punishing anything connected to AI capital expenditure on the argument that spending has run ahead of monetisation — the same argument that took the semiconductor index down roughly 25% since late June and pushed Meta down 9% on a free-cash-flow collapse. A contracted order book that tripled sequentially is the cleanest available rebuttal to that thesis at the company level, and it did not prevent the stock from falling 7% on the day it was disclosed.

The reason it did not is timing. Orders convert to revenue over multi-quarter delivery schedules, and Nokia has explicitly flagged supply chain constraints — memory shortages in particular — as a near-term growth risk. An order book that cannot be shipped is a receivable dated to whenever components become available. That is the gap the market is pricing: strong demand, constrained supply, and restructuring charges eating the reported margin while the backlog waits. It is a genuine risk and it is also, structurally, a good problem to have.

Where the Margin Went: €800 Million of Charges and a Negative €732 Million Quarter

The cash flow statement is where the bear case lives. Free cash flow came in at negative €732 million for the quarter, driven by approximately €800 million of restructuring costs. A company generating €434 million of comparable operating profit that burns €732 million of cash in the same three months has a reported-versus-comparable problem that goes well beyond presentational accounting.

The restructuring itself is deliberate and disclosed. Nokia is targeting €1.2 billion in gross cost savings by the end of 2026, and management confirmed on the call that the programme is on track. Further efficiency programmes are expected, particularly in Europe, which signals that the charges do not stop when this programme completes. The company also lowered its full-year capital expenditure guidance to €800 million to €900 million from €900 million to €1 billion, and backed a comparable income tax rate of 26% to 27%. Two previously separate businesses were reclassified during the quarter, which complicates year-over-year segment comparisons.

The strategic logic is defensible. Nokia is funding an expensive pivot — into optical, into data centre interconnect, into AI-RAN — by extracting cost from the legacy mobile business that no longer justifies its footprint. Every euro of the €1.2 billion savings target is a euro that can be redeployed into the segment growing 105%. Executed cleanly, this is exactly what a management team should do when the mix is shifting this fast.

The problem is that it makes the near-term financials unreadable, and unreadable financials get discounted in a market that has spent July repricing capital intensity. Investors currently punishing Meta for converting €31 billion of operating cash flow into €784 million of free cash flow are not going to reward Nokia for converting €434 million of operating profit into negative €732 million of free cash flow, whatever the stated reason. The restructuring narrative needs a quarter where the reported margin and the comparable margin converge. Management has guided to meaningful profit improvement in the fourth quarter. That is the checkpoint.

Guidance Decoded: €2.1–2.6 Billion, Above the Midpoint, and a Back-Half Skew

Nokia guided full-year comparable operating profit to €2.1 billion to €2.6 billion and said it expects to deliver above the midpoint — implying something above €2.35 billion. That range was itself raised from the €2.0 billion to €2.5 billion target set at the Capital Markets Day, so the company has moved the goalposts upward twice this year while the stock has halved from its high.

The shape of the year is the more useful disclosure. Management guided third-quarter net sales to grow 3% to 7% sequentially, with comparable operating profit roughly similar to the second quarter's €434 million, followed by meaningful improvement in the fourth quarter. Work that through: two quarters at approximately €434 million plus first-half results implies the fourth quarter has to carry a disproportionate share of the annual target. Back-half-loaded guidance is always the highest-risk configuration, because it leaves no room for a supply slip or a delivery delay.

That risk is compounded by the software timing issue management disclosed. Some software revenue recognition landed in the second quarter instead of the third, which flatters the €434 million comparable operating profit figure and makes the guidance for a sequentially similar third quarter harder than it appears at first reading. Strip the pull-forward and third-quarter operating profit has to grow underlying to hit the same headline number.

The counterweight is that the guidance was raised at all, in a quarter where the peer group was cutting. Ericsson delivered weaker-than-expected second-quarter results and explicitly warned that surging AI infrastructure demand was tightening memory chip supply and driving up costs for specialised silicon, with those pressures intensifying through the second half of 2026 and into 2027. Nokia raised its profit range into the same environment and has not issued an equivalent warning. Either Nokia has better supply agreements, a more favourable component mix through the Infinera portfolio, or it is being less forthcoming. The fourth-quarter print settles which.

The Memory Problem: Why a Rival's Warning Hit Nokia Harder Than Its Own Results

The single largest contributor to Nokia's July drawdown was not a Nokia disclosure. Ericsson reported disappointing second-quarter results and warned that AI-driven demand was tightening memory chip supply, lifting component costs across the telecom equipment supply chain. That warning triggered contagion selling across the sector, and Nokia — which had been one of the best-performing large-cap technology names of 2026 with a 157% year-to-date gain at its peak — became the prime target for profit-taking.

The mechanism was pure positioning. A stock that has run 157% in seven months attracts momentum capital that exits on the first credible negative catalyst regardless of whether it applies to the specific name. The Nasdaq Composite fell 1.85% in a single session during the same rout and Asian markets posted heavy losses as investors fled cyclical technology. Nokia fell from $13.49 to $10.12 across thirty days without a single company-specific piece of bad news, then dropped a further 7.1% when its own results arrived showing restructuring charges the market had not fully modelled.

The substance underneath is real and Nokia has acknowledged it. Supply chain constraints, memory shortages specifically, were flagged as a near-term growth risk on the earnings call. That is the same constraint Ericsson described, arriving through the same channel: AI data centre demand consuming memory capacity that telecom equipment vendors also require. The difference is that Nokia's exposure sits alongside a business that is a direct beneficiary of the demand causing the shortage. Ericsson's does not, in equivalent scale.

This is the crux of the valuation debate. If memory costs compress Nokia's gross margin through 2027, the 46% figure reported this quarter is the peak and the earnings model has to come down. If instead the optical and IP order book converts faster than component costs rise, margin holds and the €2.8 billion backlog becomes revenue at attractive economics. Management's decision to raise full-year guidance rather than flag the risk more prominently is a meaningful tell — but it is a tell, not evidence. The third-quarter gross margin line is where this gets resolved, and it reports on October 22.

Infinera Is Working: Optical and IP Networks Carry the Quarter

Network Infrastructure grew 12% in the quarter, led by Optical Networks and IP Networks. That segment is the operational vindication of the $2.3 billion Infinera acquisition, which closed in early 2025 and gave Nokia scale in the data centre interconnect market that AI capital expenditure is currently consuming. Comparable gross margin expansion of 70 basis points to 46% was attributed directly to optical growth and Infinera integration.

The strategic logic was sound before the AI cycle validated it. Optical networking sits at the physical layer between data centres, and every incremental gigawatt of AI compute requires proportionally more interconnect bandwidth between facilities. Nokia bought a franchise in that market at a price that looked full in 2024 and looks cheap now. The fourth quarter of 2025 already showed the pattern — Network Infrastructure delivered 7% net sales growth including 17% growth in optical networks, against total company revenue of €6.12 billion for that quarter and €19.88 billion for the full year, up 3%.

Integration risk has not fully cleared. Combining a mid-sized optical specialist into a company of Nokia's scale takes multiple years, and the restructuring charges currently destroying the reported margin include integration costs alongside the legacy mobile reduction. The two businesses reclassified during the quarter add further comparability noise. Investors modelling this segment need to work from the comparable figures and accept that the reported statements will not be clean until at least 2027.

The competitive position is genuinely strong and under-discussed. Nokia now competes in data centre interconnect against the established routing and switching incumbents with a portfolio that spans optical transport, IP routing and the SR Linux software stack — and it has an anchor customer relationship that most competitors cannot match, since Nvidia agreed to deploy Nokia switches, SR Linux and optical technologies across its own data centres as part of the partnership. That is a reference deployment at the most credible possible customer, and it is worth more to the sales motion than any marketing spend the company could deploy.

AI-RAN and the Nvidia Stake: A 2028 Revenue Story Priced Today

Nvidia announced a strategic partnership with Nokia in October 2025 and invested $1 billion at $6.01 per share, acquiring an equity stake alongside a commercial agreement. The partnership adds Nvidia-powered AI-RAN products to Nokia's radio access portfolio, built on the Nvidia Arc Aerial RAN Computer platform, with Dell supplying servers. T-Mobile US agreed to conduct field evaluations of the technology as part of its 6G development process, with trials beginning in 2026. The addressable AI-RAN market has been sized at a cumulative $200 billion-plus by 2030.

The commercial timeline is the part investors keep compressing. Nokia announced its first commercial AI-RAN platform on July 15. Pilot deployments begin in late 2026. Commercial availability arrives in 2027. Meaningful revenue is not expected until 2028 or later. That is a three-year gap between the narrative and the cash flows, and the stock has already run 129% off its low on the strength of the narrative. Progress has been genuine — functional tests of GPU-accelerated AI-RAN, customer integrations with several major European and Japanese operators, demonstrations at the industry's flagship trade show — but progress is not revenue.

The Nvidia stake itself is a double-edged input. At $6.01 per share it has returned a substantial multiple even after the recent drawdown, which validates the thesis and also creates a holder whose eventual disposition is a known overhang. More importantly, it anchors Nokia's identity to the AI trade at precisely the moment that trade is being aggressively de-rated. When prime brokers are issuing margin calls to funds with concentrated AI exposure and a $20 billion AI-dedicated hedge fund is unwinding positions, being the telecom name most closely associated with Nvidia is not a defensive characteristic.

The honest framing: AI-RAN is a real option with a real partner and a real addressable market, worth something meaningful in a sum-of-parts valuation, and worth essentially nothing in a 2026 or 2027 earnings model. Investors should size it accordingly rather than treating it as the reason to own the stock today. The reason to own the stock today, if there is one, is optical.

Valuation: Sixty-One Times Earnings for a Business Guiding to €2.35 Billion

At $9.145 with trailing earnings per share of $0.15, Nokia trades near 61 times trailing earnings. At the $8.39 close it was 57 times. At the June peak above $13 it exceeded 100 times, a level one prominent commentator flagged explicitly while recommending only a small position and advising investors to wait for a pullback. That pullback has now delivered roughly 48% from the high, which materially improves the entry without making the multiple cheap.

The forward picture is more favourable and requires the guidance to hold. Comparable operating profit above the €2.35 billion midpoint, taxed at the guided 26% to 27%, produces something in the range of €1.7 billion of comparable net income against a market capitalisation near €57 billion — roughly 33 times comparable earnings. That is a growth multiple for a business growing consolidated revenue 9%, and it only works if the AI and Cloud line keeps compounding fast enough to lift the blended rate materially higher over the next two years.

The bridge between trailing 61 times and forward 33 times is the restructuring. Every quarter that the reported and comparable figures converge, the trailing multiple falls without the stock moving. That is the mechanical de-rating case, and it is the strongest argument for patience here: an investor buying today is paying a growth multiple on comparable earnings while the reported statements catch up over four to six quarters.

The bear counterweight is straightforward and should not be dismissed. Consolidated revenue growth of 9% is respectable, not exceptional. Mobile Infrastructure is structurally challenged in a post-peak 5G capex environment with 6G volumes years away. Free cash flow was negative €732 million in the most recent quarter. The dividend yields 1.8%, which is not a valuation floor. And the entire premium above a mid-teens multiple rests on a segment representing 9% of revenue and on an AI-RAN business that generates no meaningful revenue before 2028. Strip the AI narrative and this is a telecom equipment vendor with a good optical franchise trading at a considerable premium to the sector.

The Chart: $8.37 Floor, $10.12 Gap, and the $17.45 Nobody Mentions

The technical structure is defined by the post-earnings collapse and today's reversal. Wednesday established an intraday low of $8.37 against an open of $8.91 and a prior close of $8.93 — a full-range down day that closed near the low at $8.39, which is textbook capitulation. After-hours trade extended to $8.23. Today's open gapped straight through that damage and the stock has held a $9.04 to $9.34 range with spot at $9.145.

The immediate levels are clean. Support now sits at $8.37, Wednesday's low and the marker for the entire post-earnings decline. Below that, the $8.23 after-hours print and then the $8.50 low-end analyst target become psychological reference points. Resistance runs first at today's $9.34 high, then at the $10.12 level where the stock sat in mid-July before the earnings gap — a level that now represents an unfilled gap and the most likely target for any sustained recovery. Above $10.12, the $13.49 late-June price is the next structural reference.

The longer-range numbers give perspective the daily chart cannot. The 52-week range spans $4.00 to $17.45, which is a 336% band and among the widest in large-cap technology. Nokia is currently 48% below the high and 129% above the low, and it sits roughly at the midpoint of that range on a logarithmic basis. Average daily volume of 94 million shares confirms genuine liquidity at these levels, which matters because it means the recent decline was distribution by real holders rather than a liquidity vacuum.

The read for a trader is that today's 8.4% move is a beta bounce, not a signal. Nokia moved because the semiconductor complex moved 7% on Microsoft's Azure beat, and because a name that fell 48% from its high in five weeks is mechanically oversold. Sustaining above $9.34 into Friday's close would be the first constructive signal. Filling the $10.12 gap would confirm the earnings reaction was an overreaction. Losing $8.37 on a closing basis would confirm the opposite and open the $8.50-and-below territory that the most bearish published target contemplates.

The Analyst Split: $8.50 to $21 Is Not a Consensus

The sell side is more divided on Nokia than on almost any comparable large-cap name, and the dispersion itself is data. One aggregation of eleven analysts shows a consensus Buy rating with an average twelve-month target of $15.02 to $15.16, a low of $8.50 and a high of $21 — a spread of nearly 150% between the extremes. A broader aggregation covering 35 analysts shows an average target of $10.43 with a 69% buy consensus, and notes the average target was revised upward 51.4% over the prior three months.

The individual actions since the print tell the story better than the averages. One large US bank raised its target to $18.50 from $18 and kept a Buy. A major European bank cut its target to €11.50 from €13.50 while maintaining Buy — a cut of 15% with the rating intact, which is the classic signal of an analyst who believes the thesis and has been forced to mark down the timeline. Another large US bank remains at Hold. A second remains at Hold. One remains at Sell. Two firms upgraded to Buy in the two weeks before earnings, on July 1 and July 15.

That distribution is not indecision. It is a genuine disagreement about whether the AI and Cloud line is a durable franchise or a cyclical windfall attached to a hyperscaler capex boom that is currently being questioned across the entire market. The bulls model optical and data centre interconnect compounding for years and get to $18 to $21. The bears model a 2027 capex deceleration hitting a company that still derives most of its revenue from a structurally challenged mobile business and get to $8.50.

Institutional positioning adds a note of caution. One large asset manager reduced its holding by more than 10.6 million shares in the first quarter, a 40.7% cut worth approximately $85 million. Against that, 378 institutions added to positions in the most recent quarter against 174 that reduced. The aggregate is net accumulation into a stock that has since halved, which means a substantial cohort of institutional buyers is underwater and represents potential supply into any recovery toward their entry levels.

The Forecast: $11 Base, $15 Bull, $7 Bear Into Year-End

The base case, at roughly 45% probability, is a recovery toward $11 by year-end as restructuring charges roll off and the reported margin converges toward the 9% comparable figure. This requires the fourth quarter to deliver the meaningful profit improvement management guided to, full-year comparable operating profit above the €2.35 billion midpoint, and no incremental memory-cost warning at the October 22 report. Under this path the stock fills the $10.12 gap during August or September, consolidates, and grinds toward $11 as the trailing multiple compresses mechanically. That is roughly 20% upside from $9.145 and it sits comfortably below the $15 average target.

The bull case, around 25%, requires the €2.8 billion AI and cloud order book to convert faster than the supply constraints suggest, with third-quarter gross margin holding at or above 46% despite memory costs. Add a clean fourth quarter, a 2027 guide that embeds continued triple-digit AI segment growth, and any early commercial traction on AI-RAN ahead of the 2027 availability date. That combination gets the stock back toward $13.49 and then the $15 consensus target, with the $18.50 high-end call requiring a full re-rating that the current market environment for AI-linked capital equipment does not support.

The bear case, around 30%, is a 2027 hyperscaler capex deceleration. If the optical order book stops growing while memory costs compress gross margin below 45%, the entire premium above a mid-teens multiple disappears. The stock revisits $8.37, then the $8.50 low-end target, and $7 becomes reachable — still 75% above the 52-week low, and roughly where a telecom equipment vendor growing 9% with a challenged mobile segment would trade on sector multiples. Weight this meaningfully: the market spent all of July pricing exactly this scenario across the AI complex, and Nokia was among the hardest hit.

The disciplined posture is to treat today's 8.4% move as beta rather than information. The stock is cheap against its own guidance and expensive against its own trailing earnings, and both statements resolve at the October 22 print. Accumulate weakness toward $8.40. Do not chase strength above $10.12 until the gross margin line confirms.

That's TradingNEWS