Oxford Instruments closed at 2,566p on 10 November, the day of its interim results, 23% below the 3,318p high of 3 June and still 25% up on the year, a £1.4 billion FTSE 250 instruments group620 that most screens still tag as a quantum-computing supplier57. We should be clear about that first, because it is why this name came onto our list and it is wrong. The quantum business, the dilution refrigerators and superconducting magnets of NanoScience, was sold to Quantum Design for £60 million and completed in January910. What is left of quantum at Oxford Instruments is a plasma etch tool sold to Rigetti and some cameras, inside a £37 million "other" bucket that is less than 9% of revenue111. The reason to own the company today is different and, we think, better: its Plasma Technology division makes the etch and deposition tools under the indium-phosphide lasers in AI data-centre optics, and in the first half of this financial year its orders rose 126%17. The interim results of 10 November confirmed the pre-announced numbers, put the first hard margin figures behind them and raised the division's outlook for next year17. The shares touched 2,728p and closed 0.3% lower20. We held this piece for the print and publish it the day after.
This piece belongs to our equities section, opened in November 2026 around one frame: the next cycle is being built in silicon and in qubits, and the companies that own the fabs, the tools and the machines are where the record will be written. The analysis is as of 11 November 2026, the day after the interim results.
Where we are now
The financial year to March 2026 was a transition year with the sale inside it: orders £450 million, up 8% at constant currency, revenue £423 million, down 3%, adjusted operating profit £73.7 million at a 17.4% margin, net cash £94 million, a 23.6p dividend up 6%1. Two divisions remain. Imaging and Analysis, three-quarters of revenue, makes cameras, electron-microscopy detectors, atomic-force microscopes and NMR, runs a 22.5% margin and grew orders 1%. Advanced Technologies, a quarter of revenue, is now Plasma Technology plus X-ray tubes; its orders grew 28%, its margin was 2.6%1. By end market, semiconductor orders rose 28% to 37% of intake, while academic orders fell 10% and US academia, 11% of orders, fell 11%1. China is 22% of revenue and fell 8%; two years ago £23 million of Chinese orders were struck from the book under UK export-licence controls116.
Then the trading update of 13 October for the half year to September: group orders up almost 44%, revenue up about 8–9%, the margin up more than 100 basis points from 13.3%, Advanced Technologies orders up 125% and revenue up 26%, with about three-quarters of its orders from data-communications customers and an order book that runs well into the next financial year234. Net cash was £75 million after £83 million of a £100 million buyback2. Guidance: in line with a consensus of £444 million of revenue and £81 million of adjusted operating profit for the year to March 20273. The shares rose 3.5% to 3,070p that day and then fell for five straight sessions with no stated cause, to 2,496p, and on to 2,368p on 5 November4820.
The interim results of 10 November put figures on it. Orders were £296.0 million, up 44.2%; revenue £201.1 million, up 8.4%, or 9.0% at constant currency; adjusted operating profit £30.0 million, up 21.5%, at a 14.9% margin against 13.3%; adjusted diluted earnings 42.2p a share, up 29%17. Advanced Technologies took about £132 million of the orders, up 126%, grew revenue 26.6% at constant currency and lifted its margin to 7.5% from just over 2% a year earlier1719. The order book stood at £317 million, free cash flow was £19.1 million against £0.2 million, net cash £74.8 million, and the interim dividend rose to 5.8p from 5.4p17. Guidance was held in line with a consensus that now stands at £446 million of revenue and £81.2 million of adjusted operating profit1718. Two things moved for next year: Advanced Technologies revenue is now expected to grow about 20% in the year to March 2028, and its margin is expected to reach the 12–15% target range in that year17. The buyback stood at £83.1 million of £100 million, the rest is due early in 2027, and the board intends to extend it by £25 million to the summer17. The market shrugged: the shares had risen 5.7% the day before, touched 2,728p on the morning of the print and closed at 2,566p, down 0.3%20.
At 2,566p the stock trades at 23 times this year's consensus earnings and about 14 times EBITDA, level with Renishaw at 14.6, below Halma at 18.7 and at a premium to Judges Scientific at 10.918514. JPMorgan has a 3,500p target, Jefferies 3,000p, Berenberg 2,840p; the eight-analyst consensus is 3,237p136.
The bull case
1. The tool under the transceiver
Every AI data centre is wired with optical transceivers, and the lasers inside the fastest of them are grown on indium phosphide, a compound semiconductor that is etched and coated on tools that few companies make. Oxford's Plasma Technology is one of them. Coherent is ramping six-inch InP fabs on its tools in Europe and the United States; Applied Optoelectronics bought etch and deposition clusters for three-to-six-inch InP wafers in Texas in January12. Datacomms orders rose more than 200% last year and are three-quarters of the division's intake this half13. The largest orders in the company's history, a £10 million microLED order and a larger undisclosed multi-year optoelectronics order in April, both came in the last twelve months1. This is the same bottleneck GlobalFoundries is selling into from the fab side; Oxford sells the machine.
2. Operating leverage on a 2.6% margin
Advanced Technologies earned £2.8 million on £108 million of revenue last year, a 2.6% margin, because its order book was filling while its revenue was still falling1. Management guided the division to high-teens revenue growth this year with the margin moving significantly toward a 12–15% target that it raised in June1. The first half delivered the first step: revenue up 26.6%, the margin at 7.5% against just over 2%, and the company now expects the division to grow about 20% next year with its margin inside the target range1719. On a £108 million base, every ten points of margin is £11 million of profit against a group total of £74 million. The group inflection is in the reported numbers: revenue up 8.4%, adjusted operating profit up 21.5%, the margin up 160 basis points to 14.9%17.
3. Cash, buyback, and a stock that did not move
The sale took £42 million of net cash into the group, £75 million of it was returned in the last financial year, and the first buyback in the company's history is 83% done with the shares cancelled117. The share count is down about 5%. Net cash is still £75 million, normalised cash conversion recovered to 85% from 41%, and the board intends to add £25 million to the buyback once the first £100 million completes early in 202717. And the stock is 23% below its June high and 16% below the price it reached on the day of the October trading update, after results that confirmed every number in that update, with a consensus target a quarter above206.
The bear case
Now the argument against, and the first point is the one we opened with.
This is not a quantum company, and anyone who buys it as one will be disappointed. The cryogenics franchise is gone; what remains is R&D-scale tooling for qubit fabrication and camera sales to atom-trap labs, undisclosed and single-tool sized11. If the next bull run is in quantum computers, Oxford Instruments participates through its former customers' suppliers, not through its own P&L. We hold it for the optics, and we say so.
The growth is concentrated, lumpy and low-margin. Advanced Technologies is under a third of revenue and still earns only a 7.5% margin; two orders dominate its book; three-quarters of its intake is one end market that is itself a derivative of AI capex319. Last year showed customer-readiness delays and slow conversion of orders into revenue, silicon carbide was weak globally, and the X-ray business shrank1. If hyperscalers pause, the only growing division is the one that pauses. And the year is not yet earned: the first-half profit of £30.0 million leaves 63% of the £81.2 million consensus to the second half. Last year the second half earned 66%, so the weighting is normal, but it is still ahead17181.
The core barely grows, and the geography is awkward. Imaging and Analysis, about 70% of revenue, fell 3% last year and grew 3.1% at constant currency in the half; academic orders fell 10% last year and grew under 3% in the half with US life-science research still soft, China is 22% of revenue with an export-control precedent, and UK goods carry the US baseline tariff11716. At 14 times EBITDA the company is priced above Judges Scientific for a group whose three-year revenue growth was 3.4% a year and whose 20% margin target has been promised since 2024141.
Oxford Instruments is mislabelled, and the mislabel is the opportunity. The quantum screen that attracted buyers in the spring has nothing left to find; the AI-optics order book that arrived in the autumn has not yet been paid for in the share price. First-half orders up 44%, the growth division's orders up 126%, a book into 2028, and the first step of a margin recovery, from just over 2% to 7.5% on the way to 12–15%, at 23% below the June high.
Our view: Oxford Instruments trades at 3,100p, in a 3,000–3,200p band, by the end of Q4 2027. That is 21% above the 10 November close, back to where it stood on the day of the October trading update, and below the June high. The path runs through the 14 January trading update and the 8 June preliminaries. Conviction is moderate — about 60% — and we say why: the first checkpoint was passed and the share price did not move, which says the market wants revenue and margin rather than orders, and a datacomms-concentrated order book is a fine thing to own until it is not.
Our bet runs one way: up, on optics, not on quantum. If the shares close a week below 2,000p first, we were wrong — see below.
What would prove us wrong
One direction, one closing level, and the fundamentals that would hollow it out.
2,000p is roughly where the shares ended 2025, before the year's order inflection was known. A week closing beneath it says the market has decided the order book will not convert into revenue and margin, and the thesis is closed against us, whatever happens afterwards.
Any of these weakens the thesis while the price holds, and we will log it as it happens: a trading update or results statement moving full-year guidance below the £81.2 million consensus; Advanced Technologies revenue growth falling under 10% for a half; a cancellation or deferral of either of the two large multi-year orders; or a new export-control event removing orders from the book as in 2024.
The shares trading in the 3,000–3,200p band on the London Stock Exchange by 31 December 2027. If neither the target nor the invalidation resolves by then, the thesis is booked as it stands — unresolved and closed. It does not get extended.
When a condition triggers, we book the result and publish the piece. As always.
This is analysis, not investment advice. Do your own research. The thesis is not changed after the fact.