Ericsson builds the radio equipment mobile networks run on, and collects patent fees from most of the world's phone makers. On 15 October its new chief executive, Per Narvinger, reported third-quarter sales of SEK 59.5bn, up 7% organically and about 7% above the LSEG consensus. The B share closed 6.8% higher. The same report said the radio market is flat and that component costs will be a headwind to gross margin next year. This issue runs Ericsson through the Six Gates and merges in the Nordic Cantillon File, so the business case and the price case sit on one page. The short answer, free: the business passes, the company holds SEK 60.7bn of net cash, and part of the quarter's beat was patent money that arrived late. What the price is asking for, and whether that is cheap or a trap, is for members.
The Six Gates run in order, and each one can stop the work. Gate 0 asks whether the market regime allows a long at all. Gate 1 tests the business. Gate 2 prices three worlds, bear, base and bull. Gate 3 lists the events that move the price, with dates. Gate 4 reads institutional flow on the IVT terminal. Gate 5 argues both sides, gives a verdict, and sets the entry, the size and the level that voids it all, before any money moves.
Ericsson also has a Nordic Cantillon File, built mechanically from its annual reports. Our rule is one analysis per company, so the File's findings sit inside this issue as their own chapter, after Gate 1. Everything here is priced on the B share, the liquid one. The A share carries ten times the votes and the same dividend.
Longs need a tailwind of 0 or better. It reads +2 on the weekly and exactly 0 on the daily, so Ericsson goes through with nothing to spare on the short clock. Risk appetite is off on the weekly, so any trade here is half size, the same rule the daily brief runs. Both clocks are labelled improving, and the breadth panel confirms at +8 of 8.
These layers are the terminal's market-wide proxies read on the Ericsson chart. They say whether the tide is in. They say nothing about what memory chips will cost Ericsson next year, which is Gate 1's job.
Ericsson sells three things. Networks, mostly radio base stations for 4G and 5G, was 65% of third-quarter sales. Cloud Software and Services, the core network and the people who run it, was 27%. Enterprise, which holds Vonage and Cradlepoint, was 8%. Inside Networks sits the best business Ericsson owns: patent licences, running at about SEK 13.5bn a year, paid by handset makers for using standards Ericsson helped write. That money has almost no cost attached.
The market around it is the problem. Only four vendors make radio networks at scale, which is good. But operators spend in waves, every ten years or so, and between the waves the market goes nowhere. Revenue was SEK 227bn in 2019 and SEK 237bn in 2025. Management and the research house Dell'Oro both call 2026 flat.
The third quarter, July to September 2026: sales SEK 59.5bn against an LSEG consensus of SEK 55.4bn. Organic growth 7%, the first strong quarter after minus 1% in the second. Adjusted gross margin 47.7%. Adjusted operating profit before amortisation, EBITA, SEK 10.1bn, a 17.0% margin. Last year's 28.1% included a SEK 7.6bn gain on selling iconectiv; without it the year-ago margin was 14.6%. Net income SEK 5.8bn, or SEK 1.77 a share.
| Segment, Q3 | Sales, change | Adjusted gross margin | Adjusted EBITA | What moved it |
|---|---|---|---|---|
| Networks | SEK 38.7bn, +9% | 50.3% | SEK 9.0bn, 23.1% | Sales up 17% on the second quarter against a usual 1%. Patent money helped the margin. |
| Cloud Software and Services | SEK 15.8bn, +3% | Stable | SEK 2.0bn, about 13% | Steady demand, no drama. |
| Enterprise | SEK 4.6bn, -10% | Lower | SEK -0.5bn | iconectiv is gone. Organic sales rose 1%. Still loses money. |
| Check | Reading | What it says |
|---|---|---|
| Where the beat came from | Networks, plus late patent money | Patent revenue was SEK 3.9bn, up SEK 0.8bn, partly back payments for periods that had not been licensed. The amount was not given. One secondary source puts patents at about a third of the Networks profit increase. |
| Has that patent money arrived | Not yet | The CFO said none of the catch-up was collected in third-quarter cash. It is due in later quarters. Year to date, patent revenue is SEK 10.4bn, 7% below last year. |
| Where sales grew | Asia, not the Americas | Americas, the largest area at a third of sales, fell 1% organically. South East Asia, Oceania and India rose 25%, North East Asia 22%. North America grew slightly. |
| Component costs | A 2027 headwind, not sized | Management said memory and other parts are rising "in the wake of the AI boom", mainly in Networks, and called 2027 likely a transition year. Our arithmetic: each point of gross margin is about SEK 1.7bn after tax. |
| Inventory | SEK 23.5bn to 35.4bn | Up 51% in nine months. The company points to delivery timing and higher component prices. Nokia spent EUR 560m of working capital in its own third quarter securing supply for 2027. |
| Cash | SEK 11.2bn in nine months | Free cash flow before acquisitions. Over the last four quarters it is 11% of sales, inside the 9% to 12% target. |
| Balance sheet | Net cash SEK 60.7bn | Gross cash SEK 98.0bn against SEK 37.3bn of borrowings. A EUR 500m bond was raised in the quarter ahead of maturities. |
| Costs | Falling | Operating expenses before restructuring SEK 18.4bn, down SEK 0.7bn. Staff 85,417, down 5.0% in a year. |
| Restructuring | SEK 5.3bn in nine months | Against SEK 1.2bn a year ago, and still "elevated" for the rest of 2026. One-offs that come every year are a cost. |
| Money back to owners | SEK 8.1bn of buybacks so far | 78.8m shares bought this year under a SEK 15bn programme that runs to 31 March 2027. Dividend SEK 3.00 a share, paid in two halves. |
Our red flag score is 4 of 10, against a kill line of 7. Four things carry it: a gross margin headwind for 2027 that management will not size, a beat that leaned on patent back payments not yet in cash, inventory up by half, and restructuring that has become a running cost while Enterprise still loses money. None of these breaks the business. The patent portfolio, the installed base and the short list of competitors are intact, and the balance sheet can absorb a bad year without blinking.
The numbers behind the business, from its own annual reports. The full File updates with every report.
Members unlock Gates 2 to 5 on Ericsson: what today's price assumes and the three valuation worlds with their price zones, the adjusted net cash bridge, what the analysts are split on, the dated catalyst calendar, the full IVT read on the weekly and daily with every chart level, the committee debate, the verdict, the three plans with size, exits and targets, and the source of every number.
Every Friday the complete daily brief ships open, with all levels and every terminal capture.
Ericsson stays on the Six Gates watchlist. Gate 0 is read every morning in the daily brief. Gate 4 is re-read after every daily close while price sits on its daily cost line, and after every weekly close. Gates 1 to 3 run again at the fourth-quarter report in late January 2027, when Ericsson gives its first word on what component costs will do to 2027, and the Nordic Cantillon File rebuilds after the 2026 annual report. If an entry plan triggers before then, members get it in the brief, with the size and the exit set before the trade, not after.
Organic growth growth excluding acquisitions, divestments and currency · EBITA operating profit before amortisation of acquired intangibles · Gross margin what is left of each krona of sales after the cost of making and delivering it · Patent licensing (IPR) fees phone makers pay to use standards Ericsson helped write · RAN radio access network, the base stations and antennas · Net cash cash and interest-bearing securities minus borrowings · Owner earnings cash from operations minus investment and lease payments · Implied growth the yearly growth in owner earnings, for ten years, that today's price needs to return 9% · Cost line the anchored VWAP, the average price large money has paid since a chosen point · Entry window entries only while price is within 3% of the cost line · Tailwind for longs the Cantillon macro bonus; longs need 0 or better · S2, S3 signal tiers; higher means more independent confirming factors · R one standard unit of risk.