Ajinomoto’s Q3 beat is an ABF repricing event, not a food recovery story

The first cut is straightforward: what was priced in was revenue of ¥418,880.0 million and EPS of ¥35.99, so consensus already expected sequential improvement from Q2 FY2026 revenue of ¥374,873.0 million and diluted EPS of ¥19.56. What actually surprised was not the existence of a better quarter, but the degree of earnings conversion: revenue came in at ¥426,670.3 million for a +1.9% surprise, while EPS came in at ¥39.64 for a +10.1% surprise. That separation is the core message of the print. A +1.9% revenue surprise alone would not normally force a reset in a company whose reported quarterly history has ranged from Q1 FY2026 revenue of ¥364,008.0 million to Q3 FY2026 revenue of ¥425,287.0 million, but a +10.1% EPS surprise alongside a gross margin of 39.0% does. The market may be mispricing this as a cyclical relief quarter in foods; the data say it is a mix and margin quarter, with ABF pulling the center of gravity away from lower-conviction food volume recovery.

That distinction matters because the sequential financial trajectory shows a business that finally converted seasonal revenue acceleration into margin expansion rather than giving it back to costs. Q3 FY2026 revenue was ¥425,287.0 million, up +13.4% QoQ and +4.5% YoY, and gross margin reached 39.0%, the highest point in the quarterly history provided. The immediately prior quarter had revenue of ¥374,873.0 million, revenue QoQ of +3.0%, revenue YoY of -1.0%, gross margin of 37.5%, and diluted EPS of ¥19.56. Q3 therefore did not simply lap an easy quarter; it paired the best gross margin in the series with revenue above the street’s ¥418,880.0 million estimate. The issue for PMs is that the Q4 FY2026 line already embeds some fade: revenue of ¥419,551.0 million, revenue QoQ of -1.3%, gross margin of 36.2%, and diluted EPS of ¥46.53. If investors anchor on that gross-margin step-down, they will miss the higher-quality signal inside Q3: management is identifying the margin upside in businesses tied to overseas seasonings, coffee recovery, functional materials, pharmaceutical and food-use amino acid, and Bio-Pharma Services CDMO, rather than relying on a single Japanese food pricing story.

The call’s own accounting basis supports that view, but it also shows why the headline sales beat should not be overread. Masataka Kaji framed cumulative performance as sales of “JPY 1,164.1 billion, representing 101% of FY 2024 and 101% excluding currency translation,” while business profit was “JPY 145.9 billion, 105% of FY '24 and 105% excluding currency translation.” Those company-basis figures are not the street-comparison revenue and EPS basis, and they should not be mixed with the ¥426,670.3 million and ¥39.64 beat metrics. Used properly, they say the same thing as the print: sales are moving only slightly ahead of the prior year on the company’s cumulative view, while profit is outgrowing sales. The bridge reinforces the operating mix point. Management attributed a ¥4.8 billion profit increase from gross profit due to change in sales to higher sales in overseas seasonings, domestic coffee, functional materials and others, while a much larger ¥18.6 billion profit increase from gross profit margin came from overseas seasonings, domestic coffee, functional materials, pharmaceutical and food-use amino acid, and Bio-Pharma Services CDMO. The magnitude gap between ¥4.8 billion and ¥18.6 billion is the clearest evidence that this quarter was about margin architecture more than volume.

The food business does not invalidate the thesis, but it prevents a clean re-rating if investors treat all of Ajinomoto as one homogeneous recovery. Kaji said cumulative results through the third quarter showed sales at 110% year-on-year, volume at 95%, and unit prices at 115%; that is pricing-led growth with volume still below the prior-year base. In another cut, sales reached 113% year-on-year with volumes 97% and prices 116%, which again shows that price is doing more work than units. The cleaner volume signal is narrower: excluding coffee products, sales were 108% year-on-year with volume at 106% and prices at 102%. That split is why the food story is investable only if one is precise about the mix. Coffee remains a cost and pricing negotiation problem, not a simple demand rebound; Hiroki Watanabe’s question explicitly pointed to negative ¥3 billion relating to material cost increase and mostly coffee. Kaji’s answer matters because it commits to a pricing mechanism rather than promising internal offset: “When material cost increases rapidly, and we cannot absorb this cost increase internally, then we will start negotiations with the customers regarding price increase.” That language is useful because it is conditional; it says the margin floor depends on customer negotiations when cost inflation outruns internal absorption.

The overseas seasoning data are better, but still not enough to make foods the variant perception. Kaji said the overseas business was 104% year-on-year for third-quarter cumulative results, with volume at 102% and unit price at 102%, and that October to December growth accelerated to 106% sales growth, with 104% from volume and 102% from unit price. Thailand, however, recorded only 2% growth. This mix gives investors a food business that is improving in aggregate but uneven by geography and still exposed to commodity pressure. Management also said that, based on performance up to the third quarter, it reviewed the full-year outlook and revised the fiscal year ’25 forecast, with sales revised slightly downward overall while food business profit, including Seasonings and Foods and Frozen Foods, is expected to meet the initial plan. The phrase “sales revised slightly downward” is not a throwaway; it tells PMs the revenue beat in the print should not be annualized across the group. The better read is that management is preserving profit despite a softer sales frame, which again directs attention to gross margin and mix.

The reason the stock should react to ABF more than to coffee is that management gave the market a number large enough to matter to the semiconductor substrate chain. Kaji said sales of ABF for high-performance boards used in AI servers and networks led management to revise its initial growth plan upward and expect a 28% increase in sales for the full year. That is the single most important sentence in the event for semiconductor investors. It ties Ajinomoto’s Functional Materials to AI-server and network demand, not to a generic electronics recovery, and it explains why a company with Q3 FY2026 revenue YoY of +4.5% can print gross margin of 39.0%. Hiroshi Saji’s question put an even sharper frame on the segment: Functional Materials third quarter, 42% sales increase, ¥8.6 billion and ¥5.3 billion increase, and 57%. The transcript fragment is not clean enough to overinterpret each item, but the 42% sales increase is directionally consistent with the full-year ABF upgrade to 28% sales growth. This is the variant perception: ABF is not just a helpful product line inside a food company; it is becoming the marginal driver of earnings surprise and should be analyzed against substrate and AI packaging supply chains.

That second-order read-through is clearest for the named ABF customers, because Ajinomoto’s upgrade is a demand confirmation for the advanced packaging substrate ecosystem. Ibiden, TSMC, Intel, Samsung, Unimicron, Nan Ya PCB, Simmtech, Compeq Manufacturing, ASE Group, and ZACROS all sit in the supplied customer map for ABF insulating film, ABF substrate film, or ABF resin varnish. The magnitude to attach is Ajinomoto’s 28% expected full-year sales increase for ABF used in AI servers and networks, plus the Functional Materials third-quarter 42% sales increase cited on the call. For Ibiden, Unimicron, Nan Ya PCB, Simmtech, and Compeq Manufacturing, that supports sustained substrate material pull rather than a one-quarter inventory refill. For TSMC, Intel, Samsung, and ASE Group, it is a packaging capacity and mix signal linked to high-performance boards used in AI servers and networks. For ZACROS, which is listed as toll coating for ABF resin varnish, the implication is a direct workload read-through from a supplier-customer chain in which Ajinomoto has raised ABF’s full-year growth plan to 28%. The data do not provide unit capacity or backlog for those names, so the defensible claim stops at demand validation and does not extend to earnings estimates.

The peer comparison makes the margin signal harder to dismiss as merely a Japan chemicals rebound. In the latest reported quarter, Ajinomoto’s Q3 FY2026 gross margin of 39.0% sits below 4901.T at 40.6% but above 6367.T at 32.9%, 4188.T at 29.9%, 3407.T at 32.3%, 3402.T at 20.6%, SHECY at 31.5%, 4005.T at 22.4%, and 5201.T at 24.2%. Revenue YoY of +4.5% is not the fastest in the peer set, with 6367.T at +16.4% and 5201.T at +7.7%, but it is comfortably ahead of 4188.T at -10.1% and 4005.T at -11.3%. That matters because the stock’s debate should not be framed as “growth versus chemicals peers.” Ajinomoto is showing a differentiated combination: mid-single-digit reported revenue YoY at +4.5%, high gross margin at 39.0%, and explicit ABF exposure to AI servers and networks with a 28% full-year sales growth expectation. If the market values it as a food and ingredients compounder with some electronics optionality, the print argues the optionality is already contributing to reported margin.

The call delivery adds one complication: management sounded more positive in prepared remarks, but less clean in guidance tone and more uncertain in Q&A. The tone history shows Q3 FY2026 sentiment at 0.35 versus Q2 FY2026 at 0.23, prepared_sentiment at 0.56 versus 0.03, and ai_optimism at 0.94 versus 0.26. Those are large call-over-call improvements, with deltas of sentiment +0.13, prepared_sentiment +0.53, and ai_optimism +0.67. But guidance_tone fell to 0.21 from 0.33, tone_confidence fell to 0.52 from 0.63, uncertainty rose to 63.6 from 37.2, and qa_evasiveness rose to 52.6 from 37.8. The conflict is important rather than disqualifying. Prepared remarks had the evidence for the ABF upgrade and profit bridge; Q&A had investor pressure around Functional Materials sustainability, umami seasoning profit decline, downward revision in Solutions & Ingredients, coffee material costs, and headquarters asset-sale cash. The tone data say the management team is more upbeat about what happened, but less definitive about guidance delivery.

That divergence between prepared confidence and Q&A uncertainty is exactly why the stock reaction should be selective, not indiscriminate. Investors should pay for the part of the beat tied to a named end market and a raised full-year growth plan, while discounting the food pricing story until volumes and coffee costs become less mixed. The Umami seasoning profit concern on the call was not trivial: Hiroshi Saji referenced a ¥2.8 billion decline of profit and said the decline appeared to be expanding from ¥1.8 billion. He also asked whether the downwardly revised 11% full-year number for total Solutions & Ingredients was achievable. Meanwhile, Manabu Sumoge pointed to Japan profit increase of ¥3.9 billion, with third quarter ¥3.3 billion and fourth quarter nearly ¥6 billion. These snippets show why the market can remain divided: food and ingredients have pockets of profit improvement, but also product-level erosion and revised targets. Functional Materials, by contrast, has the cleaner set of numbers: 42% third-quarter sales increase in the call discussion and 28% expected full-year ABF sales growth for AI servers and networks.

The balance sheet and capital-allocation angle is supportive but should not be allowed to become the main thesis. Watanabe referenced ¥45.1 billion as the selling price for the land and building of headquarters, with delivery on 27th of February, and tied it to a target of meeting ¥90 billion cash level by the end of 2026. Those numbers can matter for optionality, but they are not the reason EPS beat by +10.1%. The earnings-quality question is whether gross margin can remain structurally above the older range, not whether one asset sale boosts cash. In the quarterly history, gross margin was 36.5% in Q1 FY2024, 34.6% in Q2 FY2024, 36.8% in Q3 FY2024, 34.2% in Q4 FY2024, 36.5% in Q1 FY2025, 36.4% in Q2 FY2025, 37.0% in Q3 FY2025, 33.9% in Q4 FY2025, 38.2% in Q1 FY2026, 37.5% in Q2 FY2026, and 39.0% in Q3 FY2026. The Q4 FY2026 line at 36.2% is therefore the immediate test: if ABF strength and food margin repair cannot prevent a reversion toward 36.2%, the quarter will look more like peak mix than reset earnings power.

What to watch next quarter is therefore precise. First, compare Q4 FY2026 revenue of ¥419,551.0 million and gross margin of 36.2% against the Q3 FY2026 baseline of ¥425,287.0 million and 39.0%; the thesis is confirmed if the gross-margin step-down is materially less severe than that printed Q4 level, and broken if margin falls back toward 36.2% while revenue also follows the -1.3% QoQ path. Second, listen for whether ABF’s 28% expected full-year sales increase is maintained or raised, because that is the cleanest semiconductor read-through for Ibiden, TSMC, Intel, Samsung, Unimicron, Nan Ya PCB, Simmtech, Compeq Manufacturing, ASE Group, and ZACROS. Third, track whether food volume can move beyond the mixed base of cumulative volume at 95%, another sales cut with volumes 97%, and ex-coffee volume at 106%; without that, food profit depends on prices at 115%, 116%, or 102% depending on the cut. Finally, 27th of February matters for the ¥45.1 billion headquarters asset delivery, but the earnings thesis will be made or broken by the next quarter’s gross margin and ABF language, not by cash optics against the ¥90 billion end-2026 target.