Packaged Foods Dashboard.

The seven names sit inside the same TSE bucket but no longer trade on the same logic. Nichirei (2871.T), Meiji Holdings (2269.T), NH Foods (2282.T), Yamazaki Baking (2212.T), Morinaga & Co (2201.T), Morinaga Milk (2264.T) and Nissui (1332.T) span five orthogonal economic archetypes — a cold-chain dual-engine, a branded confectioner with a hidden pharma arm, two protein integrators, a dairy processor with a functional-ingredient layer, and a perishable bakery. The forward P/E spread runs from 12.0x to 17.4x, the forward EV/EBITDA from 5.9x to 9.4x. Applying a single food multiple to all seven is the first-order analytical error.

The decade settled one law for this bucket, and it is unforgiving: the market durably pays return on capital above its cost converted into free cash flow, and almost nothing else. Every premium granted to something else — a rising P&L margin, a currency windfall, a balance sheet of dormant quality — has dissipated over a cycle. Morinaga Milk is the cleanest counter-example, the best ten-year share performance in the basket on the worst quality of return, free cash flow cumulatively negative across twelve years. Nichirei is the mirror, a return on capital that climbed durably above cost and was never re-rated for it. The second law sits on top of the first: under the TSE governance reform, a price-to-book re-rating fires on a dated restitution catalyst, never on dormant quality — so the hoarders wait and the names that return capital move.

The consequence, at this moment in the cycle, is that the bucket holds no actionable long at spot. Four of the seven price within roughly two percent of a part-by-part fair value; the market is now reading the quality correctly at the status quo. The edge does not live in the price — it lives in a catalyst that has not fired (a securitisation, a buyback, a segment disaggregation) or in a pullback that has not come. What follows sorts the seven by where the return actually comes from, what they share across the bucket, how each archetype is built, and what each consensus is reading wrong.

The single most decisive structural variable across the bucket is the conversion of margin into cash, not the margin itself. The discriminating pillar on the quality scorecard is the economic model — return on capital against cost, and free cash conversion — and it is the lowest-scoring pillar for five of the seven names. The reason is that a packaged-food margin in Japan is frequently a P&L event that never reaches the cash statement. Morinaga Milk is the hard case: the operating margin tripled across the decade, from 1.1% to 6.0%, and on the EBITDA line to 10.2% on ¥58.4bn — yet free cash flow was −¥4.3bn in FY March 2026 and cumulates to −¥10.5bn over twelve years, absorbed by a capital-heavy dairy base at 1.67x depreciation and a working-capital cycle that has stretched inventory from 45 to 77 days. A margin that does not become cash is not a compounder, whatever the headline says.

Return on capital sorts the bucket more cleanly than any reported margin. Nichirei is the only name where it is both above the cost of capital and rising — 6.1% to 7.3% against a 5.5% hurdle, a near-zero beta to TOPIX, the lowest drawdown in the bucket. Morinaga & Co carries the highest base return at 11.4% but has deployed roughly ¥66bn of incremental capital since FY March 2021 at about 2.6% — below any reading of the cost of capital, which caps the multiple regardless of the brand rent. Meiji is the instructive case of a reported figure misreading the business: the 4.4% return on capital the screen shows is built on impairment-depressed net income and a 43% tax rate, while the operating return is 7.8%, comfortably above the 5.52% cost. The collapse is an artefact; the operating economics create value.

The same dispersion appears as a sum-of-the-parts dislocation in three of the seven. The market applies one food multiple to dual-engine businesses and under-prices the non-food leg. Nichirei's logistics earns half the group EBITDA at an infrastructure economics the consolidated number ignores. Meiji's Pharmaceutical arm earns 13.1% with an overseas royalty leg that just compounded, blended into a single food line. NH Foods runs a baseball stadium — ES CON FIELD — at a 19.5% margin, buried inside a protein multiple. In each, the discount is real but conditional: valued part by part at the status quo, the sum reconstructs onto the market cap, and only a monetisation or a disclosure releases it.

The thread is that the consolidated multiple is the wrong tool here. A correct read of this bucket is a sum-of-the-parts that prices the cash-generative infrastructure, the patent-protected pharma and the entertainment asset at their own multiples — and that strips the currency and cost windfalls out of the cyclical legs before crediting any of it as earnings power.

The first cross-operator input is the governance reform. The mid-2026 revision of the TSE code targets idle cash, the unwinding of cross-shareholdings is under way across the market, and the Bank of Japan policy rate at 0.75% is the highest since 1995 — together they turn the balance sheet into a first-order variable. The reform is a re-rating catalyst for the names that return capital and a discount for the names that hoard it. The bucket sits across the full range: Yamazaki holds roughly ¥146bn of dormant capital, near a quarter of its equity, and is decelerating its buyback from ¥26bn to ¥2.7bn into a record cross-shareholding balance; Meiji and Nichirei sit on idle cash and cross-holdings with no active programme; NH Foods is net-levered yet runs the most material restitution in the group, a ¥40bn buyback that floors its downside. The behavioural signal dominates the static ratio — the most cash-rich name returns the least, the most levered returns the most.

The second input is the administered and windfall cost base, which sorts pricing power across the bucket. The structural floor is domestic raw milk, administered at roughly 70% above the global benchmark, which caps Morinaga Milk and weighs on Meiji's dairy. Against it sit two windfalls now partly reversing: the government wheat resale price that fell five times in a row to ¥61,010 a tonne by October 2025, which Yamazaki has kept in price, and cocoa, which tripled in 2024 and has retraced about half while Morinaga & Co held its chocolate prices up. The distinction that matters is between a windfall and pricing power. Yamazaki has held volume through repeated hikes — consumer bread grew on volume alone in the most recent quarter and absorbed a third +5.6% increase in July 2026 — which is the signature of pricing power. Morinaga Milk's pass-through of an administered cost is a margin protection it suffers, not power it holds.

The third input runs through the yen and the demographic volume ceiling together. The depreciation toward 152–160 USD/JPY inflates the translated profit of the overseas legs, and the modelling discipline takes 125–130 as a mid-cycle base before crediting any of it. Nissui is the exposed case: roughly half its operating profit is earned overseas, and stripping the currency translation and the Marine cycle back to mid-cycle removes about ¥8bn — a fifth — from a record operating-profit print. But the exposure is not uniform: Nissui's Food-overseas leg tripled its profit over the decade on volume and share, structurally, while its Marine-overseas leg is cyclical and currency-driven. Underneath all of it sits the domestic demographic drag that makes any volume-led reading of the bucket impossible — every name's organic top line is capped, and outperformance has to come from mix, price or a non-food leg, never from volume.