Open in interactive viewer → charts, metric popovers & call review

Toray Posts a Record Half on AI Demand — and Quietly Halves Its Own Self-Help Promise

A ¥99.4bn first-half core profit record, a ¥70bn buyback funded by asset sales, and a Darwin Project that just admitted it will deliver less than half of what it once guided
3402.T · Earnings Call · 2026-11-10

The beat that needed no asterisk

Toray's six months to September were, on the numbers, the best first half the company has ever printed. Revenue rose 13.4% to ¥1,399.7bn, profit jumped 87.1% to ¥69.1bn, and core operating income climbed 46.4% to ¥99.4bn — with management noting it was “marking a record high for a six-month period” — Mitsuo Ohya, Representative Director and President · 2026-11-10. The core operating margin widened 1.6 percentage points, and the ¥31.5bn year-on-year core profit bridge was not one number doing all the work: performance chemicals revenue rose 13% to ¥501.4bn with core income up 46% to ¥41.9bn, carbon fiber composites grew revenue 35% to ¥183bn and income 75% to ¥16.5bn, and even the slowest leg, water treatment and healthcare, expanded core profit 33%. Guidance went up with it. The full year is now framed at ¥2,850bn revenue, ¥180bn core operating income and ¥110bn profit, with ROIC of roughly 5% and ROE of 6%, on a ¥155/USD assumption from October onward. That is a company guiding *into* strength rather than defending a number.

The AI hinge — and the tape's loud disagreement

The interesting part is *where* the incremental demand came from. Management framed the half as a deliberate mix of "expanding our core growth businesses and restoring profitability in low-profit businesses" alongside price pass-through against Middle East-driven raw material and fuel inflation — a competent but familiar conglomerate script. Then this line: “Increase in demand related to AI and data center also contributed to the business performance.” — Mitsuo Ohya, Representative Director and President · 2026-11-10 That is a genuinely fresh thematic anchor for a company historically pitched to investors on aircraft applications and autos. The data center thread shows up where you would expect a materials supplier to hide it: films for MLCCs and, more specifically, “sales of products for the power inductor applications expanded, although OLED-related materials were affected by the slow demand for display panels and intensified competition in China” — Mitsuo Ohya, Representative Director and President · 2026-11-10. Here is the tension worth flagging. Over the last 30 days the market's tape has been actively de-rating exactly this demand pool — AI data centers sits among the sharpest global decliners (9 advancers against 57 decliners), with data center capacity off by roughly a third and high bandwidth memory down more than 25%. Toray is monetizing the AI buildout at the passive-component and power-management layer, and the equity market is currently punishing the layer above it. The corroboration is real, though: AAOI spent its call talking about terabit transceiver capacity expansion, the same demand pool seen from a different rung of the ladder. A materials supplier with MLCC film exposure is a lower-beta, later-cycle way to own that theme — which is either the appeal or the trap, depending on whether the capex wave rolls through. The second growth leg, defense applications and space within carbon fiber, is also confirmed elsewhere in this reporting cycle by the aerospace and drone-adjacent names. Aircraft build rates at Toray's major customers "continued to grow steadily"; that is a pass-through of someone else's production ramp, not a Toray-specific win.

What management stopped promising

The real story of this call is not the record. It is the quiet retreat on the Darwin Project, the cost-and-portfolio program that has anchored the equity narrative.

For FY 2026, Toray had expected an improvement of around half of the JPY 25 billion targeted over the three-year period. However, due to a slight shortfall, mainly in the fibers and textiles business, the company now expects an improvement of about JPY 10 billion compared with FY 2025.

Mitsuo Ohya, Representative Director and President · 2026-11-10
Read that against the headline: record profits, raised guidance, and the self-help engine is delivering barely a third of the three-year target in its second year. The headline beat is cyclical and mix-driven — AI films, aircraft carbon fiber, strategic pricing — while the part management controls is slipping. That flips the usual read on a Japanese industrial beat. What management substituted for it is surgical. Under structural reform the company is re-sizing battery separator film capacity to order volumes and reviewing the TC woven fabric production structure — portfolio exits rather than cost squeezes. That is a slower, lumpier, more credible path than a three-year savings target, but it also means the Darwin Project is no longer the load-bearing promise.

The buyback is sized to the surgery, not to the profit

The ¥70bn repurchase resolved at the 10 November board is the detail that ties the story together. Management was explicit: “The size of this share buyback was determined based on the scale of asset sales.” — Mitsuo Ohya, Representative Director and President · 2026-11-10 Toray has been converting itself — divestments including its Soda Aromatic stake — and recycling the proceeds into its own equity, funded by asset sales rather than by the record profit. Add an annual dividend of ¥26 per share, up ¥6 including a ¥3 100th-anniversary commemorative component, and a DOE target of 3% or better, and you get a conglomerate whose most reliable near-term return driver is balance-sheet surgery. The balance sheet can carry it: interest-bearing liabilities of ¥848bn against owner's equity of ¥1,852.5bn, a D/E of 0.46, and positive free cash flow of ¥96.2bn for the half. Capital expenditure actually *fell* ¥3bn to ¥64.1bn while depreciation rose to ¥69.4bn — the capex cycle is behind them, which is precisely when a manufacturer should be returning cash.

The honest risks

Three things could unwind this quickly. First, competition in China is already biting the OLED materials line, and display-panel softness is not a Toray problem to fix. Second, the profit engine of the half was price pass-through against Middle East-driven raw material and fuel inflation; if feedstock eases, those pricing gains become a comparison headwind. Management flagged as much in its own outlook, tying the year to "trends in AI-related demand and the slowdown in the Chinese economy" under a global economy still carrying geopolitical and protectionist risk — and the ¥155/USD assumption leaves little room for a yen reversal. So: a genuinely good half, an upgraded year, a shareholder-return step-up — and one important promise that just got smaller without anyone shouting about it. The next two quarters will tell whether the AI films and carbon fiber demand are structural, or whether Toray's record half was a cyclical peak dressed in a medium-term plan.