New Oriental's Margin Is Being Diluted by Its Own Success
Revenue up 23% and an overseas business that quietly stopped shrinking — but the group operating margin now tells you less than it used to, and management knows it.
EDU · Earnings Call · 2026-10-28
Two businesses, one blended number
New Oriental's fiscal Q1 2027 was, by the tape of the transcript itself, a record: revenue of $1.87 billion, “total net revenue grew 23.1% year over year” — Zhihui Yang, Executive President and CFO · 2026-10-28, comfortably above the top end of its own guidance, with non-GAAP operating income up 21.2% to $406.8 million. An analyst opened the Q&A with congratulations on "the best numbers in the past 33 years" — and then added, pointedly, “despite the stock's reaction, I think the underlying operating trend looks pretty solid” — DS Kim, Analyst · 2026-10-28.
That reaction is the tell. Because the thing that changed this quarter is not the growth rate; it is the composition of the growth. East Buy, the live-commerce and private-label arm, accelerated hard enough to become a materially larger share of group revenue — and in a September-quarter where the seasonal margin profile of the education business is at its best, that mix shift flattened the consolidated margin. Steven Yang was unusually direct about it:
In Q1, East Buy's rapid growth increased its share of the total revenue. And typically in Q1, because of the seasonality, core education business margins is higher than East Buy... So this shift in our revenue mix kept our consolidated operating margin at a stable level.
Strip out the hedging and this is a genuine perceptual break. For four straight quarters the story was group margin expansion — 100bps, then 130bps, then a 470bps non-GAAP surge, then 410bps. The margin profile was the single cleanest proof point management had. Now the group number is a passive residual of two businesses moving at different speeds, and the CFO is asking investors to decompose it: “we would encourage you to also look at each business on a stand alone basis. As this gives a clearer and more accurate picture of our underlying performance” — Zhihui Yang, Executive President and CFO · 2026-10-28. Both segments expanded margins individually. The blend went nowhere. That is a real analytical downgrade of the headline metric, achieved by success rather than by slippage.
The overseas turn nobody guided to
The most under-commented item in the release is the overseas franchise. Three calls running, this was the drag. In July, guidance was explicit: “we expect our overseas-related business will be flattish or low single-digit growth in the coming new year” — Zhihui Yang, Executive President and CFO · 2026-07-29. Twelve months before that, the segment was guided to decline 4-5% on macro and geopolitical pressure.
This quarter it grew. Overseas test prep +8%, overseas study consulting +7%, adults and university students +22%. For the fiscal year management now guides test prep above 5% — a reversal of the sign, delivered without fanfare. But the durability is the open question, and it fractured along geography rather than age: previously the resilience came from younger students and non-US/UK consulting, meaning the headline number is a share-gain story inside a shrinking category, not a demand recovery. The overseas business is no longer a headwind to be managed; it is not yet a growth engine either.
Meanwhile the genuinely new disclosure is New Oriental Home, the family-centric app combining education, East Buy and tourism. It has 1.68 million registered families across 69 cities, 90% cumulative participation, a 33% campaign activation rate, and — the part that matters — a cost story: “we can easily use the related to education, action. Activities that customers are attending and then easily reduce customer acquisition and cross-sell different business lines to the same group of customers” — Sisi Zhao, CFO or Financial Officer (likely) / Investor Relations · 2026-10-28. That is the first credible structural answer to the customer-acquisition question analysts have circled for a year. The activation rate and registered family counts are new to this company's disclosure vocabulary — new metrics typically precede new economics, or new narrative.
What management stopped saying
The loudest silence is cost control. It was the number-one keyword two quarters back, the thing Yang invoked on every margin question since March 2025, quantified at roughly $100 million saved in fiscal 2026 and explicitly promised as a multi-phase program with AI-driven headcount leverage. In the top of this quarter's keyword set it barely registers. The margin language has migrated from "we are cutting" to "we are levering" — utilization, retention, mix. That is a natural handoff, but it also means the easy, self-help part of the earnings story has been harvested.
More awkward: marketing expense. Sales and marketing rose 30.7% year over year to $262.2 million — faster than revenue, and a direct reversal of the promise made twice in prior calls, that “we do not need to spend, like, crazy money on marketing to acquire the new student enrollment” — Stephen Yang, Executive President and Chief Financial Officer (CFO) · 2026-01-28. Some of it is East Buy's live-streaming paid traffic, which management has flagged before as revenue-accretive. But the group-level discipline claim, so carefully built, is now in tension with the group-level cost line, and nobody in the Q&A put the two next to each other.
Then there is enrollment — down roughly 10% in non-academic tutoring and again explained as a booking-timing artifact, students registering earlier. That is the third distinct quarter in which a soft enrollment metric was reconciled away through timing. Plausible, and pre-announced in the shift to annual guidance. But repetition of a reconciliation is itself a signal worth tracking.
The gap between the quarter and the guide
Finally, the arithmetic curiosity. Q1 grew 23.1%. Q2 is guided to grow faster: “even in the coming Q2, we expect the revenue growth of the whole group will be higher than the revenue growth of the Q1” — Zhihui Yang, Executive President and CFO · 2026-10-28. Yet the full year was merely “reiterate[d]” — Zhihui Yang, Executive President and CFO · 2026-10-28 at $6.45-6.68 billion, or 14-18%. Two quarters running ahead of the top end, against a maintained range, implies a sharp back-half deceleration that management did not even attempt to explain away. Conservative guidance has been this team's explicit habit — the June call said plainly they expect to beat it — so read this as sandbagging rather than warning. Still, reiterating after a 23% print invites the question analysts asked and did not get a clean answer to.
Underneath it all sits the capital return engine: the fiscal 2026 buyback fully exhausted (56.2 million shares), a fresh $200 million authorisation untouched, and a $300 million dividend at $2.00 per ADS in two installments. That is the floor under the stock, and the reason a 23% growth quarter producing a muted share reaction is not the paradox it looks like. The core education business is expanding capacity at 10-15% — roughly 15% for K-9 — against 20%-ish revenue growth, so utilization and Learning Center leverage remain intact. The simple margin-expansion story is over; the harder, better-diversified one, with a reaccelerating East Buy as both the growth and the margin drag, is now what investors own.