Price at writing:$10.09 (August 31, 2026, intraday). Highly volatile: the stock traded as low as $4.23 and as high as $17.45 in the past twelve months.
Conviction level: Speculative (size light, no stop)
Key catalyst: Third-quarter results on October 22, 2026, with the AI order conversion rate the number that matters. Four further reports fall before expiry, so no single event decides this trade.
Five chances, not one. The window covers results in October 2026, late January 2027, late April 2027, late July 2027 and late October 2027. The last of those lands roughly twelve weeks before expiry, leaving ample time to act on it. The one that gets away: full-year 2027 results, where 2028 guidance arrives, are due around late January 2028, roughly a week after this contract dies.
1️⃣ Trade Snapshot
The company. Nokia builds the equipment that carries data around the world: the fibre-optic gear that moves information between and inside data centres, the internet routers that direct it, and the radio equipment that puts mobile signal in the air. Its customers are phone companies like AT&T and Verizon, and increasingly the companies building AI data centres. Through Nokia Federal Solutions it also sells secure communications to US government and defence agencies. It owns a large patent portfolio and collects licence fees from handset makers, a small, highly profitable business that requires almost no ongoing spending.
The trade type. Re-rating plus catalyst. Re-rating means investors deciding to pay more per dollar of the same profit, which lifts the share price without the company earning a cent more. This is explicitly NOT a value play. The share price is above every cash-flow scenario in Section 4, including the optimistic one. You are not buying something cheap with a catalyst attached. You are betting that investors keep paying a premium price for Nokia's AI exposure for another seventeen months, and that by the back half of that window the AI-RAN revenue starts arriving to justify it.
What is this trade? You pay a one-off amount (the premium) for the right to buy Nokia at $12 a share any time before January 21, 2028. You do not have to wait for $12 to make money. The contract gains value as the stock rises toward the strike, and the closer it gets, the faster it gains. If Nokia sits at $10.09 or falls, the contract loses value and can expire worth nothing.
Profit-taking.
1 contract: Exit at +100%.
2 contracts: Sell one at +50%, the second at +100%.
3 or more: Sell one at +50%, a second at +100%, hold the rest until momentum fades or the catalyst plays out, then exit.
If the thesis breaks: Exit as soon as you are in profit at all, whatever the targets say.
No stop loss. This is deliberately a size-light, let-it-run position. The risk control is how small the position is, not a mechanical exit level. The defined risk is the full premium, and that outcome is accepted at the point of entry rather than managed around afterward. A 508-day contract on a stock that swings this hard would trip a percentage stop on ordinary noise and take you out of a thesis that still had a year to run.
Why the timing works, and what the extra time buys. The seventeen-month window covers five sets of results, and that is the entire reason for choosing this expiry over a shorter one. More importantly, it fixes a structural flaw in a shorter-dated version of this trade. Nokia's AI radio platform starts trials late in 2026, becomes commercially available through 2027, and produces meaningful revenue from 2028 onward. A June 2027 contract would expire before the money arrives. This one lives through the entire commercial rollout year and into the start of the revenue year. The proposed US rule on Chinese optical equipment, the new US factories, and the 2027 AI-RAN subscription launch all sit comfortably inside the window rather than at its edge.
Report
Date
Status
Q3 2026
October 22, 2026
Confirmed on Nokia's published calendar
Q4 and full year 2026
Late January 2027
Estimated from pattern
Q1 2027
Late April 2027
Estimated from pattern
Q2 2027
Late July 2027
Estimated from pattern
Q3 2027
Late October 2027
Estimated from pattern
Q4 and full year 2027
Late January 2028
Falls after expiry
Nokia publishes its following-year calendar around early December, so the 2027 dates are not yet official. The pattern is consistent: full-year 2025 results came on January 29, 2026, first-quarter on April 23, 2026, second-quarter on July 23, 2026, and third-quarter is set for October 22, 2026.
The honest limit: applying that same pattern, full-year 2027 results land around January 28, 2028. That is roughly one week after this contract expires. You pay for seventeen months and miss the single print where 2028 guidance is issued.
How far does the stock need to move? From $10.09 (Aug 31, 2026) to $12 is +18.9%. Including the premium, the true breakeven is nearer $13.60 to $14.15, roughly +35% to +40%. For scale, the stock was above $13 as recently as late June 2026, and the 52-week range is $4.23 to $17.45.
Note on the US listing. Nokia's ordinary shares trade in Helsinki in euros. What trades in New York under the ticker NOK is an ADR (American Depositary Receipt), a US-listed certificate that stands in for one ordinary Nokia share. It moves with the Helsinki price adjusted for the euro-to-dollar rate. Every dollar price in this document is a US-listed price, and every euro figure comes from Nokia's own reporting.
Key catalysts (detail in Section 5):
Q3 results, October 22, 2026: how much of the AI order book is turning into sales
Full-year results, late January 2027, carrying 2027 guidance
Three further reports through 2027, the last roughly twelve weeks before expiry
The proposed US ban on new Chinese optical transceivers, the small plug-in modules that send data as pulses of light between machines inside a data centre. Expected to be published before the end of 2026
AI-RAN commercial rollout on a subscription model through 2027
New US capacity coming online: San Jose from late 2026, expanded Pennsylvania capacity from Q3 2026, Arizona leased capacity from early 2027
A supporting signal, not a catalyst. Insiders have bought twice this year, and the second time is the interesting one. In late May 2026 senior manager Victoria Hanrahan bought 44,682 US-listed shares at an average price of $15.81 across two purchases. In late July, near the lows, three insiders bought on the same day, including board member and former Nokia finance chief Timo Ihamuotila, at prices around $8.45 to $9.55, roughly $1.4 million combined and about 45% below the May purchases. The honest reading: the May buyer is down about 36%, so this says management believes the strategy, not that they can time the stock. Read it as one small supporting data point, not a reason to size up. Small August purchases of a few hundred shares are routine plan settlements and should carry no weight at all. Detail in Section 5.
What's the most you can lose? The premium you pay. Nothing more, ever. There is no margin call and no obligation to buy anything. This is the single feature that makes the trade acceptable given everything in Section 4, and it is why no stop loss is needed.
When does this trade stop making sense?
The October or January results show the AI order book converting to sales far slower than the roughly half-in-twelve-months management has guided to.
The proposed US rule on Chinese optical equipment is shelved or watered down.
The company cuts its profit guidance, or pushes restructuring costs higher again, including any charge tied to the mainland China exit.
The AI-RAN subscription rollout slips out of 2027, which would push the revenue proof entirely beyond this expiry.
The stock breaks back below roughly $8.40, the late-July 2026 low, on heavy volume.
Cash generation fails to recover, making the full-year cash target unreachable.
Valuation Summary
Scenario
Fair value per share
vs $10.09 (Aug 31, 2026)
Cautious
$4.93
-51%
Middle
$6.27
-38%
Optimistic
$7.71
-24%
Weighted average of the three
$6.23
-38%
Analyst average (not cash-flow based)
$15.02
+49%
Caveat, and it is a large one. The $12 strike sits above every single cash-flow scenario, including the optimistic one. There is no valuation floor beneath this trade, and a longer expiry does not create one. It buys time for the story to prove itself; it does not put a floor under the price while you wait. The gap between roughly $6.23 and the $15.02 analyst average is entirely investors agreeing to pay more per dollar of profit, not extra cash arriving.
Analyst Price Targets
Eleven analysts tracked by S&P Global: consensus Buy, average $15.02, lowest $8.50, highest $21. No median is published, which is a real gap given the spread.
A low of $8.50 against a high of $21 is not a consensus. It is two groups who disagree entirely about what this company is worth.
A caution on the individual names below. Most of the analysts covering Nokia set their targets in euros against the Helsinki listing, and the dollar figures here are converted at 1.16. Several widely used data services carry targets that are months stale or currency-confused. Where a firm has not updated since its last dated note, that date is shown rather than the number being presented as current.
Notable Analyst Price Targets
Firm
Rating
Target
Notes
JP Morgan
Overweight, top pick
$21 (EUR 12 on the Helsinki listing)
Sandeep Deshpande. Raised from $14 after Q1, reiterated August 19, 2026, arguing consensus is not modelling the order book and anchoring to 2027 and 2028 estimates
BofA
Buy
$18.50
Raised August 4, 2026, on AI order intake
SEB Equities
Buy
EUR 12 (~$13.92)
Upgraded July 2026
Deutsche Bank
Not disclosed
EUR 11.50 (~$13.34), cut from EUR 13.50
July 27, 2026, four days after Q2 results. The most recent cut by any covering firm
Morgan Stanley
Overweight
EUR 11 (~$12.76)
Terence Tsui. Third raise of 2026, from EUR 8.50 after Q1, having upgraded from Neutral to Buy in January
Goldman Sachs
Neutral
EUR 8.00 (~$9.28)
As of March 27, 2026, when the firm upgraded from Sell and raised from EUR 3.50. A July 28, 2026 note carried no disclosed target, so this figure is stale and now sits below the share price
Reading: The lowest target in the S&P Global consensus is $8.50, but no covering firm I could identify publicly holds that number, and it should not be attributed to Morgan Stanley, whose actual stance is Overweight and rising. Among firms with a traceable current view, the range runs from roughly $9.28 to $21, and the only genuine skeptic is Goldman at a five-month-old EUR 8.00. Every firm that has published since Q2 sits above $13.
On data sources. For the same stock on the same day, published averages run from $10.43 to $15.30, and at least one widely used service is still serving prices from 2024. Use the S&P Global figures only, and treat any individual target without a date attached as unusable.
Where to Find More Detail
Why the conviction is Speculative and what has to go right: Section 2
Sizing, profit-taking and the pre-results exit rule: Section 3
The full valuation work and why the price has no cash support: Section 4
What each catalyst actually is: Section 5
How the contract makes and loses money: Section 6
What kills the trade: Section 7
2️⃣ Overall Assessment
The $12 strike sits above the optimistic cash-flow case of $7.71. Every scenario says the stock is expensive before you even add the strike. There is no cushion.
What the price already assumes: To justify $10.09, Nokia's cash generation has to grow roughly 17% a year for five years, then half that pace for five more. Management's own 2028 goal implies about 11% a year. The strike needs closer to 20.5%, and the breakeven needs 23%.
The offsetting fact: The order book is real and large. AI and cloud orders reached EUR 2.8 billion in the second quarter alone, with sales to those customers more than doubling. JP Morgan's August 19, 2026 argument is precisely that consensus has been too slow to model this into 2027 and 2028 numbers.
Analyst positioning is unusually one-sided. Every firm that has published a dated target since Q2 sits above $13, and Morgan Stanley has raised three times this year. The only bearish view on record is Goldman's, and it has not been refreshed since March. That is support for the re-rating case, but it is also a crowded setup: when nobody is left to upgrade, disappointment has further to travel.
The structural case for this expiry. The trade's biggest weakness in its shorter-dated form was that the contract died before AI-RAN revenue arrived. January 2028 removes that mismatch. Five reports fall inside the window, the last of them twelve weeks out, and the 2027 subscription rollout happens in full view.
What the extra time costs. A higher premium, and therefore a higher breakeven: roughly +35% to +40% rather than the +24% to +28% a June 2027 contract required. You are buying more time and a worse entry price at once. The stock also needs +18.9% to reach the strike now, up from +11.5% when it traded at $10.76 two weeks ago, purely because the stock has fallen since.
Conviction level: Speculative (size light, no stop). Not because the structure is aggressive, it is not, but because the underlying is. This stock has travelled from $4.23 to $17.45 inside twelve months. You would be buying a share price with no valuation support underneath it, with a key policy catalyst that is still only a draft, and with the company in the middle of a second major restructuring. Any one of those is manageable. Together they mean small position size, even with seventeen months of room. With no stop loss in place, size is the only risk control this trade has, so it has to do all the work.
Key dependencies: The AI order book converting on schedule, the AI-RAN subscription rollout landing in 2027 rather than slipping, the proposed US rule on Chinese optical equipment actually being published, cash generation recovering, the mainland China exit not producing another restructuring charge, and strict discipline on position size and on the pre-results exit.
3️⃣ Risk Management & Exit Framework
Position parameters
No stop loss. Deliberate. This is a size-light, let-it-run position and the defined risk is the full premium.
Profit-taking by size:
1 contract: exit at +100%
2 contracts: sell one at +50%, the second at +100%
3 or more: sell one at +50%, a second at +100%, hold the remainder until momentum fades, then exit
If the thesis breaks: Exit at the first sign of profit, regardless of the targets above.
Why size replaces the stop. A percentage stop on a 508-day contract solves the wrong problem. The premium is already the maximum loss, so a stop does not cap anything that is not already capped. What it does do is force an exit on a drawdown that a seventeen-month thesis was built to absorb, in a stock that fell 50% in eight weeks and then recovered. Set the position small enough that losing all of it is a shrug, then let the thesis run its course.
Sizing, since it is now the entire risk framework
Size so that a total loss of the premium is comfortably survivable at the portfolio level, not merely tolerable.
Watch correlated tail risk. This position breaks on the same event as other AI infrastructure exposure: investors across the board deciding to pay less per dollar of profit for anything tied to the AI buildout. Nokia, chip and networking names would fall together, so size this against the aggregate, not against this one line item.
The pre-results rule. Option prices rise in the days before results as traders position, then fall sharply once the news is out, even when the news is good. This is the single most reliable way to lose money on a correct call. If you are in profit, sell into the rise ahead of October 22, 2026 rather than holding through it. The same applies before each of the four later dates.
The October 2027 decision point. The final results before expiry land roughly twelve weeks out. That is the natural hold-or-exit moment: late enough to know how AI-RAN monetisation is tracking through its first commercial year, early enough that time decay has not begun eating the position quickly. Mark it in advance rather than deciding in the moment.
On entry timing. Option prices are cheapest just after results, not before. Two sensible routes:
Buy a partial position now and add after October 22 regardless of which way the stock moves.
Wait entirely for the period just after October 22, accepting that you give up seven weeks of exposure to the policy catalyst.
The longer expiry makes the second route cheaper in opportunity terms than it was on a ten-month contract, since seven weeks is a much smaller share of the window.
Other rules
Do not buy more if the contract is falling. Adding to a losing call whose strike is still above the share price is how small losses become total ones, and no-stop does not mean average-down.
A break below roughly $8.40 on heavy volume, taking out the late-July 2026 low, is a reassessment trigger, not an automatic exit. Check it against the thesis: if the AI order book is still converting, it is price action; if it coincides with a conversion miss or a guidance cut, that is the thesis breaking and the override above applies.
Value drains out of the contract faster in the final three weeks. Do not hold into January 2028 unless the stock is comfortably above $12.
4️⃣ Valuation Assessment
Inputs:
Share price:$10.09 (August 31, 2026, intraday)
Euro to dollar rate: 1.16, carried forward from the two listings on the same day (New York $10.56, Helsinki EUR 9.091, August 13, 2026). To re-derive. A rate of 1.13 would cut every per-share figure below by roughly 3%
Shares in issue:~5,658 million, being the 5,742,239,696 total shares in Nokia's own stock exchange release of June 30, 2026 less 83,991,222 treasury shares, meaning shares the company bought back and holds itself, which no outside investor owns and which therefore do not count when working out value per share. This replaces the ~5,500 million estimate used previously
Spare cash after all debt:EUR 2,776 million at the end of Q2 2026
Annual sales:EUR 19,889 million (2025)
Method: Ten-year projection of surplus cash in two stages (years 1 to 5, then 6 to 10), then steady growth thereafter at 2%, marked down at 9%, 10% and 11%
In plain terms: The model estimates what the business is worth by projecting the spare cash it generates each year and converting that back into today's money. Cash arriving years from now counts for less than cash today, so it gets marked down each year.
What changed and what did not. The business values in euros are unchanged from the previous version of this work. Only the share count correction and the lower share price move the numbers. The model was not rebuilt.
Historical cash generation, and why it cannot be projected
Year
Spare cash (EUR M)
Change
Source
2021
~2,065
Data provider, to reconcile
2022
~873
-58%
Data provider, to reconcile
2023
~665
-24%
Data provider, to reconcile
2024
~2,021 to 2,400
roughly tripled
Company filing gives 2,400 on one definition; providers show 2,021
2025
~1,500
-26%
Company's own figure, 72% of profit converted to cash
No trend line fits this. It runs 665, then 2,065, then 1,500. The swings come from patent licence fees landing in lumps, customers paying early or late, and the cost of job cuts. Projecting forward from any single year would be guesswork. The 2024 figure in particular is a recovery year, not a new baseline, and must not be extrapolated.
So I used the company's own guidance framework instead. Nokia expects profit of EUR 2.1 to 2.6 billion this year and says it will land somewhat above the middle, so call it EUR 2.4 billion. It expects 55% to 75% of that to become cash, giving EUR 1.32 to 1.80 billion. Restructuring costs were raised to EUR 800 million this year, with EUR 700 to 800 million leaving the bank account, far above normal. Adding back the excess gets you near EUR 1.9 billion.
Starting point used: EUR 1,600 million. Above the raw number, below the fully cleaned-up one, because the promise that job-cut costs stop is a promise, not a fact. The August 2026 decision to close almost all mainland China sites is a direct argument for keeping it there rather than raising it.
Valuation Summary
Scenario
Fair value per share
vs $10.09 (Aug 31, 2026)
Cautious
$4.44 to $5.56
-56% to -45%
Middle
$5.60 to $7.14
-44% to -29%
Optimistic
$6.84 to $8.84
-32% to -12%
Cautious (3% growth for five years, then 2%)
Marked down at
Business value (EUR M)
Per share
vs $10.09 (Aug 31, 2026)
9%
27,114
$5.56
-45%
10%
24,057
$4.93
-51%
11%
21,679
$4.44
-56%
Middle (8%, then 4%)
Marked down at
Business value (EUR M)
Per share
vs $10.09 (Aug 31, 2026)
9%
34,814
$7.14
-29%
10%
30,598
$6.27
-38%
11%
27,329
$5.60
-44%
Optimistic (12%, then 6%)
Marked down at
Business value (EUR M)
Per share
vs $10.09 (Aug 31, 2026)
9%
43,121
$8.84
-12%
10%
37,616
$7.71
-24%
11%
33,358
$6.84
-32%
Weighted average (25% cautious, 55% middle, 20% optimistic, at 10%): $6.23, or 38% below $10.09 (Aug 31, 2026).
Growth stress test: Halve the middle case to 4% then 2% and fair value falls to $5.12, or 49% below the current price.
The finding that matters most: The optimistic case at the gentlest mark-down rate produces $8.84, still below today's price. No defensible combination of assumptions reaches $10.09, let alone $12.
Sensitivity to the starting point, middle case at 10%: EUR 1.5 billion gives $5.92, EUR 1.8 billion gives $6.99, EUR 2.0 billion gives $7.70.
What the price already assumes
Since the model keeps landing below the market, the more useful question is what today's price requires. Growth needed for five years, then half that pace for five more, marked down at 10%:
Starting cash
To justify $10.09
To justify $12.00
To justify ~$13.90 breakeven
EUR 1,500M
18.3%/yr
21.6%/yr
24.5%/yr
EUR 1,600M
17.2%/yr
20.5%/yr
23.3%/yr
EUR 1,800M
15.0%/yr
18.3%/yr
21.1%/yr
EUR 2,000M
13.1%/yr
16.4%/yr
19.2%/yr
Nokia's own 2028 goal is profit of EUR 2.7 to 3.2 billion with 65% to 75% converting to cash. Take the middle and that is about EUR 2.07 billion of cash in 2028 against roughly EUR 1.5 billion in 2025: growth of about 11% a year. The current price needs roughly 17%. The strike needs roughly 20.5%. The breakeven needs roughly 23%.
Strike positioning against the value work
$12 is 92% above the middle case of $6.27 and 56% above the optimistic case at the gentlest mark-down rate.
The strike is supported only by the analyst lens. Every firm with a dated target published since Q2 sits above $13, and the highest is $21. Wall Street's targets back a move through the strike. The cash-flow math does not come close.
Longer duration does not narrow this gap. It only gives the story more time to turn into cash. If it does not, the seventeen months work against you rather than for you.
Earnings cross-check
Profit per share (EUR)
At $10.09
At $12.00
2026
0.347
25.1x
29.8x
2027
0.384
22.7x
27.0x
2028 (company goal)
0.421
20.7x
24.6x
Twenty-one times the profit the company hopes to earn in 2028. For reference, the forward price-to-earnings multiple currently published by S&P Global is 23.9x. That figure is the share price divided by the profit per share analysts expect over the next twelve months, and it is the standard shorthand for how expensive a stock is. That is a price normally paid for a fast grower, on a business where the mobile division grew 6% with profit unchanged year on year.
Why the analysts are so much higher
A model built on total company cash will understate Nokia right now, because that is not what buyers are paying for. They are paying for a change in mix. Network infrastructure grew 12%, optical grew 20%, internet routing grew 16%, and sales to AI and cloud customers more than doubled. The shrinking mobile business funds it. Analysts value the pieces separately, applying a rich price to the AI-exposed part and a separate value to the patent arm, which produced EUR 1,501 million of sales at essentially no cost and EUR 1,059 million of profit in 2025.
Running that piece-by-piece approach at ordinary multiples still only reaches about $7 a share. The distance from roughly $7 to $15 is investors agreeing to pay more per dollar of profit. None of it is extra cash. That is the bet.
Note on shares given to staff: Companies often pay part of employee compensation in shares rather than cash. This is called stock-based compensation. No cash leaves the business, so it does not reduce reported cash flow, but it is a genuine cost to you as an owner because it creates new shares and shrinks your slice of the company. Nokia has historically issued roughly EUR 150 to 250 million a year this way. Against EUR 1.5 billion of cash that is roughly 10% to 17%, sitting on the threshold at which it would deserve a separate adjusted valuation table. Because the cash-flow work here is a caution flag rather than the basis of the trade, it stays as this note. Unconfirmed. Needs the cash-flow statement in the 2025 annual report. It will not change the conclusion, only sharpen it.
5️⃣ Catalyst Thesis
Third-quarter results, October 22, 2026. The first and most important test, and the headline numbers are not the point. Management said they expect around half of the EUR 2.8 billion of AI and cloud orders to become revenue over the next twelve months. That conversion rate is the number the entire trade rests on. If roughly half is converting on schedule, the growth the share price demands starts to look reachable. If it slips, the case weakens badly, because there is no valuation support to fall back on. Also watch second-half cash generation, which has to swing sharply positive, and any further increase in restructuring costs. Management guided to sales rising 3% to 7% in the third quarter with profit roughly flat, then a meaningful improvement in the fourth. That gap between a flat third quarter and a strong fourth is where disappointment tends to live.
Full-year results, late January 2027. Where 2027 guidance arrives, where the fourth-quarter improvement either materialised or did not, and where full-year cash generation is finally settled. It also brings the first full-year read on how the AI order book converted across twelve months rather than one quarter.
Three 2027 reports, and why they are the point of this expiry. Late April, late July and late October 2027. This is what a June 2027 contract could not reach. AI-RAN pilots run late in 2026, commercial availability arrives during 2027 on a subscription model, and these three prints are where subscription revenue either shows up in the numbers or does not. The October 2027 report, roughly twelve weeks before expiry, is the last usable signal and the natural decision point.
AI-RAN, the product the whole story now rests on. A RAN, or radio access network, is the layer of masts, antennas and base stations that connects phones to the network. AI-RAN means running that layer on AI chips so software can manage the radio signal more cleverly. Nokia launched what it calls the first commercial AI-RAN platform, built on NVIDIA's Aerial and accelerated computing stack. The pitch is that AI squeezes more capacity out of existing 4G and 5G radios while giving carriers a software path toward 6G. Nokia says the platform already delivers over 20% better spectral efficiency, meaning more data pushed through the same slice of radio spectrum, with a target of 50% by 2027 and more than double by 2028. Two honest caveats: the launch announcement carried no subscription pricing, contract size or revenue target, and commercial rollout is 2027 with meaningful revenue in 2028. Chief Executive Justin Hotard calls it the biggest innovation in radio in decades. That is a claim, not a result.
The proposed US rule on Chinese optical equipment. Regulators are drafting a measure to block imports of new Chinese-made optical transceivers, and officials want it published and in force before the end of 2026. Nokia is among the small group of non-Chinese suppliers at the highest speeds. Treat it carefully: it is a draft, the sources themselves said it could be modified or shelved, and at least one specialist firm does not expect it to survive. It is an option on top of the thesis, not the thesis.
Building the capacity to supply it, and the US government relationship. Nokia committed $4 billion to US R&D and manufacturing in November 2025, announced alongside the Trump administration: $3.5 billion to domestic R&D centred at Nokia Bell Labs in New Jersey, and $500 million to manufacturing and R&D across New Jersey, Texas and Pennsylvania. That follows the $2.3 billion Infinera purchase. In practice this means the San Jose plant starting production late in Q4 2026, a tenfold increase in advanced testing and packaging capacity in Pennsylvania from Q3 2026, and leased capacity at an Arizona semiconductor site acquired from NXP from early 2027. Note the direction of the money: Nokia is spending, not receiving. There is no US government stake in Nokia.
Nokia Federal Solutions, and a widely misreported number. Nokia launched a dedicated US federal unit built on its acquisition of Fenix Group, which supplies secure battlefield wireless to defence customers. In March 2026 the unit was selected as one of the approved holders on a US Missile Defense Agency programme called SHIELD. What that actually means: SHIELD is a framework contract. The agency pre-approves a list of suppliers, and those suppliers are then allowed to bid on individual pieces of work as the agency issues them. The framework carries a ceiling of $151 billion, which is the maximum the agency may spend across the entire programme, across every approved supplier, over its whole life. Several outlets reported this as Nokia winning a $151 billion contract. That is wrong. Nokia's own statement describes the award as positioning the unit to compete for future task orders. Treat it as the right to bid on future federal revenue, not as revenue.
Insider buying, and how to read it. Two clusters. In late May 2026, senior manager Victoria Hanrahan bought 44,682 US-listed shares across May 26 and 28 at an average price of $15.81, roughly $707,000. In late July 2026, near the lows, three insiders bought on the same day: senior manager Patrik Hammarén took 43,293 shares at EUR 8.4420, board member and former Nokia finance chief Timo Ihamuotila took 60,000 shares at EUR 8.4522, and senior manager Pallavi Mahajan took 62,000 US-listed shares at $9.55. Roughly $1.4 million combined at prices about 45% below the May purchases. The honest framing: the May buyer is down about 36%. This is a signal about management's belief in the strategy, not about timing, and small August purchases of a few hundred shares are routine plan settlements that should carry no weight at all.
NVIDIA's stake, and a headline worth ignoring. NVIDIA's latest Form 13F, the quarterly filing in which large US investors must list the holdings they own, showed a $2.21 billion position in Nokia, and the US-listed shares surged 8.7% in a day. But NVIDIA's original investment was $1 billion at a subscription price of $6.01 per share, about 166 million shares. Valued at the June 30 price, that same holding is worth $2.21 billion. Same shares, higher price. NVIDIA did not buy more. A 13F reports what a position is worth on the reporting date, not what was bought during the quarter, and the two get confused constantly. A meaningful part of the late-August rally rests on that confusion. The underlying partnership on AI-RAN is real and does matter; the 13F number does not.
The steady part underneath. The patent licensing arm runs at roughly EUR 1.4 billion a year at essentially no cost, and the mobile division still contributed EUR 310 million of profit in the quarter. These do not grow, but they fund everything else and make a collapse in the share price less likely than the raw valuation would suggest.
6️⃣ Strike Structure
Strike: $12.00
Premium: Estimated $1.60 to $2.15, based on the time remaining and how volatile this stock has been
Breakeven at expiry: Strike plus premium, roughly $13.60 to $14.15, or +35% to +40% from $10.09 (Aug 31, 2026)
Move to the strike:+18.9%
Sensitivity to each $1 move (delta): Estimated 0.45 to 0.55, given the distance and the long duration
Days to expiration:508
How it makes money.
Movement in the stock (delta). For every dollar Nokia rises, the contract gains roughly 45 to 55 cents at the moment. That rate accelerates as the stock climbs toward $12. The stock does not need to reach $12 for the trade to work. A run to $11.50 or $12 during 2027 could comfortably double the premium.
Nervousness about the stock (vega). When traders expect big moves, options get more expensive regardless of direction. This works for you into each results date as positioning builds, and hard against you immediately after, when that expectation collapses and the premium drops even if the news was good. This is why the pre-results exit rule in Section 3 exists, and it is the most common way a correct call still loses money. On a 508-day contract, vega is the largest single sensitivity. A general collapse in option pricing across the AI complex would hurt this position even if Nokia itself did nothing.
Time passing (theta). The contract loses a little value every day simply because there is less time left. At 508 days this is very slow, and this is the clearest mechanical advantage over a shorter expiry: the daily bleed is roughly half what a ten-month contract suffers. It accelerates sharply in the final three weeks, which is why you should not hold into January 2028 unless the stock is clearly above $12.
7️⃣ Key Risks
Dominant risk: no valuation floor. The strike sits above every cash-flow scenario including the optimistic one. Nothing underneath the price is holding it up except investor willingness to pay a premium for AI exposure. When that willingness turns, the fall can be fast, and the option has no floor at all. This stock fell from $16.77 in early June 2026 to the $8s by late July 2026, roughly 50% in eight weeks. That is what the downside looks like in practice, and with no stop in place it is what you would sit through.
Analyst positioning leaves little room to improve. Every dated target published since Q2 sits above $13, the highest is $21, and Morgan Stanley has raised three times this year. Upgrades and target raises have been a real driver of the recent rally, and that well is close to dry. The next surprise is more likely to be a cut than a raise.
You pay for time and still miss the key print. Full-year 2027 results, where 2028 guidance is issued, land about a week after expiry. If the market waits for that number before re-rating, seventeen months of premium buys you the wait and none of the payoff.
Higher breakeven than a shorter contract. The stock needs +18.9% to reach the strike and roughly +35% to +40% to break even at expiry. The extra time is not free, and paying for it moves the finish line further away.
Orders that do not become sales. The entire case rests on roughly half of EUR 2.8 billion converting within twelve months. Order intake is a promise from a customer, not money in the bank, and AI-related orders in particular can be pushed out.
AI-RAN monetisation is unpriced. The launch carried no subscription pricing, contract size or revenue target. A commercially available product with no disclosed economics is a story, and the market can stop paying for stories at any point inside seventeen months.
Restructuring keeps recurring. Nokia plans to close almost all its mainland China sites and cut most of its workforce there by year end, following the Hangzhou R&D closure and about 1,600 job cuts. Strategically defensible, but it argues against the assumption that job-cut costs are about to normalise, which is exactly the assumption the EUR 1,600 million cash starting point depends on.
Cash generation. Free cash flow was negative EUR 732 million in the second quarter amid EUR 800 million of restructuring costs, and management flagged supply constraints, particularly memory shortages, as a near-term risk. Ericsson has flagged the same pressure across the sector.
The policy catalyst may not arrive. The optical transceiver rule is a draft that may be shelved. A meaningful part of the recent rally would unwind.
Total loss is the realistic downside. With the strike above fair value on every measure, the stock stalling for seventeen months is enough to lose the entire premium. Holding shares in the same company alongside this contract does not protect the contract: a stock that drifts sideways near $10 leaves the equity merely dull while the call expires worthless.
Counter-argument (why this can work). The business really is changing. Optical grew 20%, internet routing grew 16%, sales to AI and cloud customers more than doubled, and the company took in EUR 2.8 billion of AI orders in a single quarter with supply, not demand, as the constraint. Customers placing longer-dated orders because they cannot get equipment is a strong position to be in. Nokia is one of very few non-Chinese suppliers at the highest speeds at the exact moment US policy is pushing buyers away from Chinese ones, and it is bringing US capacity online through 2026 and 2027 to serve it. JP Morgan's August 2026 argument is that consensus has simply not caught up to the order book yet, with a $21 target and 2027 and 2028 as the anchor years. This expiry is the one that actually reaches those years. Five sets of results give five separate chances for the conversion numbers to prove the case, and whatever happens, the most you can lose is the premium.
Open items to resolve
2027 reporting dates. Nokia publishes its calendar around early December 2026. Confirm the four estimated dates then, and specifically confirm whether full-year 2027 results really do fall after January 21, 2028.
Euro to dollar rate. Re-derive from a same-day New York and Helsinki pair. Every per-share figure in Section 4 and every euro-denominated analyst target scales with it.
The $8.50 lowest target. S&P Global reports it but no covering firm could be matched to it. Worth identifying, since it is the only bearish data point in the consensus.
Goldman Sachs current view. The last public target is EUR 8.00 from March 27, 2026. The firm published a note on July 28, 2026 with no disclosed target. Confirm whether a newer figure exists.
Share count. The ~5,658 million figure is total shares from the June 30, 2026 release less treasury shares at August 27, 2026. Confirm the diluted figure against the annual report.
Starting cash figure. Reconcile the 2024 discrepancy (EUR 2,021M vs EUR 2,400M) against the 2025 annual report.
Half-year 2026 free cash flow. The summary press release does not carry the line. Pull it from the full financial report.
China exit cost. Whether the mainland China closures produce a restructuring charge beyond the EUR 800 million already guided for 2026.
Staff share cost. From the cash-flow statement in the 2025 annual report.
Research for personal use. Not investment advice. Options can expire worthless and the entire premium is at risk. Verify pricing, the share count, the cash position, the annual report figures flagged as unconfirmed, and any developments affecting the catalysts through primary sources before any transaction. The valuation work is a scenario tool, not a prediction; small changes in growth and mark-down assumptions move fair value substantially.