Bloomsbury's record year now rests on two novels, a Christmas TV show and an Anthropic cheque

Half-year profit was always going to be the quiet one — Academic normalised, Consumer was out of season. The whole FY27 'record' promise now hangs on an ACOTAR double-launch, an HBO tie-in and a settlement that isn't in consensus.

BMY.L · Earnings Call · 2026-10-22

A record year, narrowed to a window in January

The half-year figures are almost a distraction: revenue of £151.6m, adjusted profit of £16.8m at an 11.1% margin, EPS of 16.67p, net cash of £22.8m versus £2.4m a year earlier, and a 5% lift to a dividend that has now grown for 31 consecutive years since the 1994 flotation. What genuinely changed is the shape of the year. Bloomsbury has spent a decade selling investors on diversification — Nigel Newton's portfolio of portfolios — and this year that portfolio is being asked to work in reverse. Academic and Professional, which carried FY2026 with profit doubling to £25m, has normalised to £6.8m on £48.8m of revenue. The Consumer division, deliberately light in the first half, is now the thing that has to deliver the record. Keith Underwood is explicit that the mitigation runs one way this year and then flips back: “Last year, in FY 2026, the Academic and Professional division had strong profits doubling to GBP 25 million, which mitigated lower profits in the Consumer division.” — Keith Underwood, Chief Financial and Operating Officer · 2026-10-22 The corollary is that a £16.8m first-half profit is now a stub, and the year is a bet on the fourth quarter. “Sales in October and September were really encouraging and very strong.” — Keith Underwood, Chief Financial and Operating Officer · 2026-10-22

Two novels, and a tie-in jacket

The consumer story is unusually concrete for a publisher. Two Sarah J. Maas novels in her ACOTAR series land within a quarter of each other — A Court of Splintered Harmony days after this call, A Court of Forgotten Melody in January — and Newton is not being coy about the stakes: “This may well turn out to be the biggest-selling book of the year. Certainly, that was the case with Sarah's previous launch.” — Nigel Newton, Executive Chairman · 2026-10-22 The mechanics matter more than the hype: retailers are funding their own promotion, and Target is planning in-store boutiques built around the title. Newton notes the backlist is already moving, with one of the series back on the New York Times bestseller list in anticipation. With 16 novels behind her and 18 imminent, the catalogue cascade is the real asset — the launch title is the hook, the backlist is the annuity. The second engine is Harry Potter, and it is the cleaner test of a thesis the market tends to underrate: that screen time sells paper. HBO's series lands on Christmas Day, and Bloomsbury holds English-language rights everywhere outside the United States — Canada, Australia, New Zealand, South Africa and the wider Commonwealth. The first time around, the films arrived without a tie-in jacket; this time there is one, and the supermarkets that abandoned the brand are back.

The pump is primed. We have significant orders in the U.K. from the major supermarkets. We have doubled down on that by seeking marketing from them to go with it.
Nigel Newton, Executive Chairman · 2026-10-22

The arithmetic he offers is refreshingly unromantic: roughly 700,000 new eight-year-olds in Britain every year, most of whom have never met the boy wizard. That is a new generation of readers argument, not a nostalgia argument — and it is the kind of thing that shows up in second-half revenue rather than in first-half commentary.

The AI licence is shrinking — and that is the interesting part

The most consequential disclosure on the call was a de-emphasis. Bloomsbury's non-exclusive AI licensing agreement rolls into FY2027 and again into FY2028, but the quantum is smaller and the margin is lower, which is why the Academic division's 6% underlying growth is being reported excluding AI income. “We've always said that that would be at a lower quantum than we were able to generate in FY 2026. That's still our expectation.” — Keith Underwood, Chief Financial and Operating Officer · 2026-10-22 A year ago this was the upside surprise; now it is a headwind dressed as continuity, and management has quietly stopped leading with it. What replaces it is more interesting: a broadening of the counterparties. The first deal was training data; the pipeline is now specialist RAG deals and further training licences with other labs, which is exactly the demand the market's own curation has been flagging as frontier AI buyers hunt for licensed content. Note the contrast with where the AI keyword heat actually sat this quarter — the curated global themes were dominated by physical buildout, and the AI data center basket has in fact been de-rating over the last 30 days, with 48 decliners against 6 advancers. Bloomsbury's AI exposure is on the other side of the trade: rent collection on intellectual property rather than capex on silicon. Framed globally as AI risk for content owners, here it is a revenue line. The kicker is the Anthropic matter. A double-digit-million settlement is expected once the administrator works through the disputes process, and crucially it sits outside forecasts.

For the consensus cash forecast, Anthropic settlement is not built into those numbers.
Keith Underwood, Chief Financial and Operating Officer · 2026-10-22

Against a £505m market capitalisation and £16.8m of first-half adjusted profit, that is not a rounding error. It is also, as ever with settlements, a timing question dressed as a quantum question.

Cash, pruning, and the thing nobody asked about twice

The balance sheet is the quiet tell. Net cash moved from £2.4m to £22.8m, total net assets to £213m, and finished stock is down sharply year on year. Some of that is textbook working capital discipline — print-run sign-off, order management — but part of it is genuinely new: “starting to use AI in terms of predicting print runs and trying to drive intelligence through the use of that type of technology.” — Keith Underwood, Chief Financial and Operating Officer · 2026-10-22 A publisher using machine learning to size a first printing is a better margin story than a publisher licensing its backlist to a chatbot, because it compounds every season. That cash brings the capital allocation priorities forward, and management is doing two things at once: pruning and buying. Sale processes for Who's Who and the Games division are progressing, framed as margin repair inside Consumer, while bolt on acquisitions remain the stated use of excess cash alongside debt reduction. Investors should temper the enthusiasm with seasonality — cash always falls in Q3 before the March royalty run rebuilds it — and with the fact that a record year built on two novels and one streaming launch is, by construction, a year with less room for error than the 31-year dividend record implies. The Academic recovery Newton describes — “a recovery from some of the knocks taken by political interference, if you like, in the normal running of the academic marketplace” — Nigel Newton, Executive Chairman · 2026-10-22 — is real but early, and the Singapore hub is nine months old. The honest read: this is a well-run small cap whose diversification claim is being tested in the opposite direction for the first time in years, with a genuinely company-unique catalyst (Anthropic) bolted onto a genuinely seasonal one (ACOTAR). That combination is rarer than the headline 'record profit in line with expectations' suggests.