Oxford Instruments' Buyback Is Slow Drip of EPS Accretion, Not a Signal That the Stock Is Cheap

Oxford Instruments has spent most of the past year issuing the same quiet announcement on repeat: another small tranche of its own shares bought and cancelled. The latest, for the week to 27 August, retired 34,961 shares and left 54,940,044 in issue — one more filing in a stream that mostly scrolls past. The buyback that actually matters here is not a single week's handful of shares. It is a programme that has already put £62.2m back into shareholders' hands in the year to 31 March, funded from cash the company never had to borrow.

That distinction is the whole article. This is a story about cash-flow quality wearing the clothes of a mechanical share-count update, and reading it the wrong way — as a sign the shares are cheap — gets the conclusion backwards.

What the buyback is really funded with

Start with where the money comes from, because that is what separates a comforting buyback from a dangerous one. Oxford Instruments ended its 2026 fiscal year with £94.0m of net cash, up 11% year on year, and it converted 89% of its cash-generation into free cash. The programme is two £50m tranches: the first was completed by the end of February 2026, and by 31 March the company had finished £11.7m of the second. In March it extended that second tranche through to no later than 2 March 2027. Management also said it deployed £62.2m toward buybacks during the year.

The funding is worth noting precisely because it is comfortable. A big chunk traces back to a real asset sale — NanoScience was divested in January 2026 for £42.4m in net proceeds — and the rest to operating cash flow. There is no leverage propping this up, no debt-funded statement of false confidence, and the dividend was still raised 6.3% to 23.6p per share on top of the buyback. For a deep-value investor, this is the part of the story that is genuinely good: a net-cash balance sheet returning capital while staying net cash is a durability signal, the companies-can-do-this sort of signal.

So the buyback confirms something real about the business. It just does not confirm what the headlines imply about the stock.

Buybacks don't make a rich stock cheap

Look at the price the company is paying. The shares trade a little under £29, up sharply over the past year, against adjusted earnings per share of 100.7p for fiscal 2026. That works out to roughly 28 times trailing adjusted earnings, and close to 20 times enterprise value divided by adjusted operating profit, once the £94.0m of net cash is stripped out. That is a growth-premium multiple for a company whose reported revenue fell 4.6% and whose adjusted operating profit fell 7.3% last year. Management's stated purpose is merely to reduce the share count, which each cancellation does permanently — but when a company buys back stock at roughly 28 times earnings after a run, it is spending cash at a premium, not harvesting a mispricing.

The mechanical framing should be kept in proportion too. Every weekly "update" announces a tiny drop in a float of roughly 55 million shares — 34,961 shares is on the order of 0.06% of the company. The per-share uplift from any single tranche is negligible, and even the full programme only matters to EPS by compounding slowly as the denominator shrinks. None of this is a reason to buy or sell on the announcement itself.

Where the value, if any, actually has to come from

The buyback matters mainly as a vote of confidence in the part of the story that can genuinely move the number: the Advanced Technologies division. Its order intake jumped 28% on an organic constant-currency basis to a record £133.1m, and after a large multi-year order received in April 2026, the division's backlog now largely covers planned fiscal 2027 revenue and extends into 2028. The demand driver is real and current — datacomm, power electronics and micro-LED orders tied to AI and data-centre expansion, with datacomms-related orders up over 200% in the year. Group order intake rose 8% on the same basis, and the book-to-bill ratio was 1.07, so orders are running ahead of revenue.

That is what justifies a premium multiple, or fails to. The company's Imaging & Analysis business, the larger profit engine, posted a 23.3% adjusted operating margin but flat-to-slightly-down revenue, so the growth story leans almost entirely on Advanced Technologies converting its record backlog into revenue and, eventually, margin. In fiscal 2026 that division's adjusted operating margin was 3.0% — it collected the orders but not yet the profit, with revenue lagging on longer lead times. If the backlog converts, the ~28x multiple may prove to have been reasonable. If it does not, or if it converts at thin margins, the buyback has been executed at a price the recovery never validated.

All things considered, Oxford Instruments earns real credit for the cash-flow and balance-sheet side: high conversion, net cash, capital returned through both a dividend and a buyback without taking on leverage. That is financial strength, and it deserves to be priced as such. But "we are buying back our own stock" is not the same claim as "our stock is cheap." At roughly 28 times trailing adjusted earnings with profits down and the whole argument resting on whether a record order book turns into margin, the margin of safety is thin — so the buyback is evidence of management conviction and a solid balance sheet, not, on its own, a reason to own the shares. The condition that would change that reading is the conversion of that order book into profit, and that is the variable worth watching, not next week's share-count update.