Drugstores Dashboard.
A reference hub on Japan's six listed drugstore and dispensing-pharmacy operators.
Six names inside the same TSE bucket, six economic archetypes, and one metric that sorts them all. This sector does not read on the income statement: the EBIT margin band is narrow, from 3.85% to 7.79%, and free-cash-flow conversion is uniformly good, so neither discriminates quality. The only line that does is the lease-adjusted ROIC–WACC spread — five compounders against one value-destroyer. The dislocation is inverted: the market prices the four best dossiers at their decade floor and holds the two weakest at mid or high-cycle multiples. The universe is 100% domestic, so there is no yen effect to normalise. The single re-rating lever available is capital return under TSE pressure, and it is un-priced across the whole bucket. Long asymmetry is modest in four names; the sharper convictions sit on the short side. The consolidated sector multiple is no longer the right tool.
The six names sit inside the same TSE bucket but no longer trade on the same logic. MatsukiyoCocokara (3088.T), Cosmos Pharmaceutical (3349.T), Sugi Holdings (7649.T), Sundrug (9989.T), Kusuri no Aoki (3549.T) and Tsuruha Holdings (3391.T) span six economic archetypes — a margin monetiser, a land-owning compounder, two consolidators of the dispensing pharmacy, a balanced two-engine discounter, and a debt-financed roll-out. Welcia, absorbed into Tsuruha, is delisted and excluded. The dispersion inside the bucket is itself the first read.
The decisive fact settled by the sector work is that this bucket does not read on the income statement. The EBIT margin band is narrow — 3.85% to 7.79%, roughly 3.9 points between the best and the worst — and free-cash-flow conversion ex-growth is uniformly good, from 52% to 137%. Neither separates a compounder from a trap. The one metric that does is the lease-adjusted spread of return on invested capital over its cost, and it isolates five compounders from a single value-destroyer: Matsukiyo +6.4, Sugi +6.0, Sundrug +5.8, Cosmos +3.6, Aoki +1.9, against Tsuruha at −3.1. That single line governs the entire hierarchy of conviction.
The economic fracture beneath the spread is land ownership, not product mix. Capital expenditure runs from 1.19% of sales at Matsukiyo, a tenant that leases and returns cash, to 6.41% at Cosmos, an owner that possesses and reinvests. Choosing a demand base is choosing a capital structure: the cosmetics monetiser leases its floor space and hands cash back ; the food-frequency owner buys the land under its stores and puts the cash back to work. Reported free cash flow ranges from +71% of EBIT to −28% for exactly this reason.
The central dislocation is inverted. The four best dossiers — Matsukiyo, Sugi, Sundrug, Cosmos — trade at the floor of their own decade valuation corridors, while the two weakest — Aoki and Tsuruha — hold mid or high-cycle multiples. The market is pricing quality backwards. The generalised de-rating from the 2016–2020 peaks toward the 2026 floors probably marks the end of the sector's structural growth premium ; the only re-rating from here comes from capital return under Tokyo Stock Exchange pressure, or from a re-discrimination of quality — not from revenue growth, which in this bucket is now bought rather than earned.
What follows below sits in layers. The economic engine and the cross-operator inputs describe what is shared across the six. The archetype map and the names section sort them. The mispriced variables and the structural watchlist track what each consensus is reading wrong and what would force a reframing.
The most decisive distinction across the bucket is between pricing power and mix power, because only one name in six has the former. Matsukiyo is the sole operator with a real, durable domestic pricing edge — roughly +6.2 points of gross margin over the decade, captured by a private-label range co-developed on 169.55m customer contacts, even if that edge is blended with a cyclical inbound-tourism effect the company does not disaggregate. Every other name monetises through mix alone: Cosmos through private label at 16.8% of sales, Sundrug through discount buying scale, Sugi through the officine mix, Aoki through grocery-to-basket cross-sell, Tsuruha through an officine and private-label drift. None of them can raise price without losing the customer, because the merchandise floor is a levelled, price-matched trade and the dispensing pharmacy is administered by the reimbursement schedule.
Demand quality sorts the same way but not in the same order. The most defensible base is Cosmos — near-daily food frequency at 62.9% of the mix, mature same-store running +5.7% with positive traffic, and a demand that inflation helps rather than hurts as shoppers trade down into the low-cost format. Sugi (officine same-store +13.2%) and Matsukiyo (officine recurrence plus the private-label data layer) follow. Aoki and Sundrug are more transactional and weakly captive — roughly 90% of Aoki's revenue is exposed, with only the pharmacy contractually sticky. Beta is uniformly low across the bucket, but low beta protects only in a falling market ; in the 2024–2026 rotation the most defensive names were the worst laggards.
The critical cost tells the third story, and it is where the new macro risk enters. The best-managed critical cost is Cosmos's — an everyday-low-price gross margin near-invariant across the cycle, 1.7 points of amplitude over twelve years. The worst structural scissor is Sugi's — a pharmacist wage pulled up by a national shortage and non-passable, crossed with a dispensing price cut every two years, which turns the officine's operating leverage negative. And three names have seen an entirely new critical cost emerge at the Bank of Japan turn: the cost of debt. Cosmos posted its first net-debt year, Tsuruha runs at 2.2x EBITDA, and Aoki carries a presumed-floating book near ¥150bn — none of them immune any longer.
Cash conversion completes the picture, and it draws the sharpest internal line in the bucket. The best EBIT-to-cash bridge is Sugi's (137% conversion, ex-growth free cash flow of ¥66.7bn) and Cosmos's (a negative working-capital cycle financed by suppliers, ex-growth free cash flow of ¥34.2bn). The most broken is Aoki's — roughly 14% conversion, cumulative free cash flow close to nil over eleven years, its reinvestment financed by debt. That single contrast separates the self-funded land compounder from the destructive roll-out despite an identical food-frequency logic, and it is why the consolidated multiple becomes the wrong tool the moment format or segment dispersion widens — as it does at Sundrug (discount versus drugstore), Matsukiyo (Matsumotokiyoshi versus Cocokara) and Sugi (officine versus merchandise). The sum-of-the-parts is the instrument the market has not yet reached for.
The first cross-operator input is the one that is absent. This universe is 100% domestic — there is no overseas earnings leg, no yen-inflated margin, and no FX normalisation to run. That is a major analytical simplification relative to the export buckets in the coverage, and it removes the single most common source of headline-margin distortion. What it leaves in place as the transverse macro risk is non-passthrough wage inflation: prefectural minimum wages and the structural labour shortage add roughly 3% a year to pharmacist and floor-staff cost, and on a levelled merchandise trade and an administered officine there is no reliable way to pass it through. The bucket sorts on who can absorb it through mix and scale rather than price.
The second input is the officine fee schedule, the most concentrated and most datable regulatory risk in the sector. The Ministry of Health, Labour and Welfare revises the dispensing price and fee downward every two years, and the 2026 revision is pending in April. The exposure is graded by officine weight: Sugi is the most exposed at 30.4% of sales — a −2% revision costs roughly 12.6% of gross EBIT — followed by Aoki at 10.6%, then Matsukiyo and Tsuruha, with Cosmos and Sundrug least exposed. The market treats this risk as diffuse ; it is concentrated, quantified and on the calendar.
The third input is the rate turn, which lands unevenly. Bank of Japan normalisation reaches only the three names that have flipped into net debt or leveraged up — Cosmos, Tsuruha and Aoki — where the fixed-floating split is undisclosed and the interest-cover cushion has thinned. The modelling discipline across the bucket takes a normative cost of capital near 6–7% rather than the frequently artefactual WACC values, with a size-premium adjustment where warranted. On a ¥130/$ house convention the FX question does not arise here, but the rate question replaces it for the leveraged half of the bucket.
The fourth input is the one catalyst shared by almost the entire bucket: capital return under the Tokyo Stock Exchange's "P/B above 1" pressure. Five of the six names carry an un-priced return-of-capital option as their principal upside lever, gathered under a single common trigger. Matsukiyo, the only regular returner of capital, is also the only multiple in the bucket the market never destroyed — the clearest empirical proof that this sector re-rates on the return of capital, not on growth. It makes the bucket a basket of capital-return optionality rather than a set of unrelated single names, and it means the re-rating, when it comes, is likely to be correlated and broad rather than idiosyncratic.
| Archetype | Operator | Read |
|---|---|---|
|
Balanced two-engine
Discount growth aisle mispriced beneath a mature drugstore
|
Sundrug 9989.T | Split by format, one engine — the Direx discount business at 43% of revenue — produced all of the decade's profit growth and earns more on its assets (11.4% versus 8.3%) than the pharmacy chain it is filed beneath. The market prices both engines at the same floor multiple. Valued part by part, the drugstore on a mature multiple and the discount on a growth one, the sum reconstructs to ¥4,397 against a spot of ¥3,838 — a +14.6% composition discount, real but moderate, on a net-cash balance sheet. The best fundamental long asymmetry in the bucket, weighted fair value +10.1%. |
|
Land-owning compounder
Self-funded, handed the worst de-rating in the bucket
|
Cosmos Pharmaceutical 3349.T | The best EPS compounder in the bucket — roughly +9.9% a year for a decade — and the worst de-rating, down 53% from its multiple peak. Screened on the income statement it reads as a roll-out burning cash: reported free cash flow of −¥11.7bn and a first-ever slide into net debt. Read on the balance sheet it is a self-funded owner of ¥343.5bn of land, financed by suppliers through a negative working-capital cycle, throwing off ¥34.2bn ex-growth. The floor is an asset floor ~36% below spot ; it bounds a crisis, not the current downside. Weighted fair value +8.4%. |
|
Margin monetiser
Only real pricing power, dormant balance sheet
|
MatsukiyoCocokara 3088.T | The one name the market has correctly kept off the de-rating pile — the only regular returner of capital, the only expander of margin over the decade, the only multiple never destroyed. The consolidated 10.5% return on equity reads ordinary ; the operating capital earns 12.6% ex-cash, the best spread in the bucket, dragged by ¥116bn of dormant net cash, about 12% of the market cap. Sum-of-the-parts fair value ¥2,573 against ¥2,469.5 spot. Quality already in the price ; the upside is entirely gated by an un-taken capital-return decision. Weighted fair value +4.2%. |
|
Officine consolidator
First-rank compounder, tax-doped reported earnings
|
Sugi Holdings 7649.T | The 11–12x reported P/E is a triple illusion. A 0.43% effective tax rate — a non-cash deferred-tax credit worth ~¥13.6bn the company has guided away — doped FY2/2026 net income to ¥45.0bn against a normalised ~¥32bn ; correct the tax and the treasury divisor and the multiple resets to ~16.5x, mid-range. Underneath sits a genuine first-rank compounder, ROIC 12.0% and a +6.0pt spread. But valued part by part the sum falls below the price: the officine's +7.2pt gross-margin edge compresses to ~+1pt at contribution once the pharmacist is charged. Fairly priced, weighted fair value −4.3%. |
|
Debt-financed roll-out
Growth that no longer earns its cost of capital
|
Kusuri no Aoki 3549.T | The best revenue line in the bucket and close to the worst return on capital. A ¥135bn drugstore in 2015 has become a ¥567bn food & drug roll-out, and the market pays it the highest price-to-book of the six for that growth. Underneath, return on capital has more than halved — from 16.2% to 6.7%, now level with its cost of capital — while net debt jumped 68% in one year to ¥90.8bn to fund the expansion. The ¥64 dividend, quadrupled, is funded from the balance sheet (¥6.1bn against ¥4bn of free cash flow). A value trap whose verdict is suspended on one un-published number, the mature-estate return. Weighted fair value −7.2%. |
|
Consolidator under Aeon
The bucket's only value-destroyer, optically cheap
|
Tsuruha Holdings 3391.T | The largest drugstore operator in Japan and the only one that earns less on its capital than that capital costs — a lease-adjusted ROIC−WACC spread of −1.6 to −3.1 points against +1.9 to +6.4 for the five peers. The optical cheapness is a trap twice over: the 1.2x book multiple is 2.6x on tangible book, because 51% of equity is intangible after ¥448bn of Welcia goodwill ; the 7.2x forward EV/EBITDA is a low multiple on an EBITDA inflated in absolute yen by a low-margin base. Every method converges on overvaluation, −14% on EV/EBITDA to −54% on normalised earnings — but under Aeon's 50.3% majority no activist catalyst is possible. Weighted fair value −14%. |
The name says drugstore, but the drugstore is not the story. The Direx discount business, 43% of revenue, produced all of the decade's profit growth and earns 11.4% on its segment assets against the drugstore's 8.3%, turning them 2.15x a year against 1.44x. The market prices the group's better engine as its drag. A sum-of-the-parts by format — the drugstore on a mature multiple, the discount on a growth one — reconstructs to ¥4,397 against a spot of ¥3,838, a +14.6% composition discount.
It is moderate because the quality underneath is only moderate: ex-cash return on capital of 16.4% is healthy but has fallen roughly twenty points over the decade, and the most recently reinvested capital earns 6.5%, barely at the cost of capital. Weighted fair value sits ~10% above spot on a 1.3x up/down ratio, floored by net cash and a 3.4% dividend. The catalyst that would close it — a return of capital — has not appeared in a decade.
The best EPS compounder in the bucket, down 53% from its multiple peak. The market screens the income statement — reported free cash flow of −¥11.7bn, a first-ever slide into net debt — and files it as a roll-out burning cash. Read the balance sheet instead and a self-funded owner of ¥343.5bn of land appears, financed by suppliers through a negative working-capital cycle, throwing off ¥34.2bn once growth capex is stripped out. The de-rating is partly a screening error and partly earned: a 20.3% payout, no buyback in its history, no like-for-like disclosure.
What the reframing does not settle is the one number the file withholds — whether the marginal store in the Kantō push earns above the cost of capital. Return on capital has fallen from 16.8% to 9.7%, the spread compressed to +3.6 points. The ~¥4,000 asset floor bounds a crisis but sits ~36% below spot. Weighted fair value +8.4%, gated on mature same-store holding above +3% and on any capital-return inflection.
The one name the market has correctly kept off the de-rating pile — the only regular returner of capital, the only expander of margin over the decade, the only multiple never destroyed, at 17.6x earnings. The consolidated 10.5% return on equity reads ordinary ; the operating capital earns 12.6% ex-cash, the best spread in the bucket, dragged by ¥116bn of dormant net cash, about 12% of the market cap. The consensus reads the consolidated return, not the operating one.
The asymmetry is thin and entirely un-triggered: sum-of-the-parts fair value ¥2,573 against ¥2,469.5 spot, +4.2%, with the whole upside gated by a capital-return decision that has not been taken. A second point has to hold — the record 7.6% margin is partly inbound-fed, and the one-year total return has already turned down −19% as the tourist flow normalises. Watch the FY March 2027 margin print and any signal on the return of capital.
The reported 11–12x P/E is a triple illusion. FY2/2026 net income of ¥45.0bn was earned at a 0.43% effective tax rate, doped by a non-cash deferred-tax credit worth ~¥13.6bn the company has guided will not recur — FY2/2027 net income is guided down 27.1%. Correct the tax, correct the treasury divisor, and the multiple resets to ~16.5x normalised earnings, the middle of the decade range. The stock is not cheap ; it is fairly priced.
What the number hides is genuine: ROIC of 12.0% against a ~6% cost of capital, a +6.0pt spread and first rank alongside Matsukiyo, the best cash conversion in the sector. But the parts do not add up — the officine's 36.4% gross margin looks like a +7.2pt edge over merchandise, yet the pharmacist wage compresses it to ~+1pt at contribution, and the company does not disclose the number. Tail asymmetry close to symmetric, bull +22% / bear −22% ; the 2026 fee-schedule revision is the datable downside.
The revenue line is the best in its bucket and the return on capital close to the worst. A ¥135bn drugstore in 2015 is now a ¥567bn food & drug roll-out, and the market pays the highest price-to-book of the six for that growth. Return on capital has more than halved, from 16.2% to 6.7%, now level with its cost of capital, while net debt jumped 68% in a single year to ¥90.8bn. The premium is paying for a memory — the question is not whether the top line grows, but whether the growth still earns anything above what it costs to finance.
A capital-return pivot softens the picture — the ordinary dividend quadrupled to a guided ¥64, payout to ~32% — but at ¥6.1bn against ¥4bn of free cash flow it is funded from the balance sheet, not earned. Normalise the FY May 2024 stock-comp one-off and the underlying margin erodes, 5.81% to 4.78%. Weighted fair value −7.2% with a downside skew ; the whole verdict turns on the un-published mature-estate return, routed to Temps 2b.
The largest drugstore operator in Japan and, on the number that matters, the only one in its peer group that earns less on its capital than that capital costs. The Welcia merger that built the scale also built ¥448bn of goodwill — 51% of equity — that the return on capital does not service. The optical cheapness is a trap twice over: the 1.2x book multiple is 2.6x on tangible book ; the 7.2x forward EV/EBITDA is a low multiple on an EBITDA inflated in absolute yen by a low-margin base. The two most attractive metrics in the dossier are its two most dangerous.
Return on invested capital, lease-adjusted, runs −1.6 to −3.1 points below the 6.01% cost of capital — the only negative spread in the bucket. Reported earnings are doped ~30% by a ¥10.58bn step-revaluation gain. Every valuation method converges on overvaluation, from −14% on EV/EBITDA to −54% on normalised earnings, so the direction is not in question, only the magnitude. The one caution is timing: under Aeon's 50.3% absolute majority no activist catalyst is possible — well-founded, weakly triggered.
| Metric | Who it tests | What would change the read |
|---|---|---|
| Lease-adjusted ROIC−WACC spread | Whole bucket | The single discriminant. Tsuruha is the only negative spread (−1.6 to −3.1pt) and Aoki the thinnest positive (+1.9pt) and compressing. Aoki turning negative confirms the value trap ; Tsuruha back above zero on more than ¥30bn of Welcia synergy invalidates the short. |
| Capital-return policy — DOE, payout, buyback | Five of six | The universal un-priced catalyst. Matsukiyo is the only regular returner and the only multiple preserved. A DOE above 4.5%, a payout above 40%, or a structural buyback at Cosmos, Sundrug, Sugi or Matsukiyo mobilises dormant capital and opens the re-rating the sector pays for. |
| 2026 officine fee-schedule revision — MHLW, April 2026 | Sugi · Aoki | The most concentrated, most datable regulatory risk. On Sugi's 30.4% officine a −2% revision costs 12.6% of gross EBIT ; worse than −3% uncompensated by volume feeds the bear on the officine-heavy names and misses the operating guidance. |
| Mature-estate / cohort return on capital | Cosmos · Sugi · Aoki | The un-disclosed cardinal lever, structural to the bucket — Japanese drugstores never publish a per-store P&L. A mature same-store return proxied above the cost of capital settles the compounder reading ; below it confirms roll-out-in-saturation. Routed to a shared Temps 2b cohort reconstruction. |
| Consolidated / segment operating-margin trajectory | Matsukiyo · Sundrug · Tsuruha | The margin-structural test, per name. Matsukiyo below 7.0% confirms an inbound peak (fair value toward ~¥2,100) ; Sundrug's discount margin below 5.0% breaks the composition discount ; Tsuruha held at or below 3.9% confirms permanent Welcia dilution. |
| Cost of debt and net-debt path | Cosmos · Tsuruha · Aoki | The newly-emerged critical cost at the BoJ turn. A cost of debt above 3% on rising net debt turns a historical non-issue into a structural constraint at Cosmos ; net debt still rising while the dividend is paid confirms balance-sheet financing at Aoki. |
| Aeon control and related-party transactions — 50.3% | Tsuruha · 3391.T | Absolute majority subordinates the minority structurally, and caps any re-rating. A distribution cut to fund Aeon-piloted M&A confirms subordination and deepens the destruction ; no activist counter-lever is available under majority control. |
The framework rests on two assumptions. The first is the sector law: this bucket does not read on the income statement, and the only re-rating lever is capital return under TSE pressure. The cleanest single invalidation is a broad, correlated return-of-capital wave — if the "P/B above 1" pressure triggers a bucket-wide return of capital, the modest long asymmetries in Cosmos, Sundrug and Matsukiyo convert together as a sector event rather than a series of idiosyncrasies, and Matsukiyo, the only current returner, is the directional proxy for the theme. This is why the bucket is better sized as a basket of capital-return optionality than as isolated single-name bets.
The second runs through the un-disclosed cardinal lever that three dossiers share — the return on the mature versus the immature store, which Japanese drugstores never publish. If the shared Temps 2b cohort reconstruction shows the marginal store earning durably above the cost of capital, the compounder reading on Cosmos hardens and the value-trap reading on Aoki softens ; if it prints below, the reverse, and Aoki's short-lean is confirmed. On the short side, Tsuruha's invalidation is narrow and specific: a ROIC−WACC spread pushed back above zero on more than ¥30bn of captured Welcia synergy — improbable under four structural bottlenecks and Aeon's absolute majority, but the one development that would force a revision away from the short. Symmetrically, a distribution cut to fund further Aeon-piloted M&A would harden it.
This dashboard is the reference document for sub-industry 01a. The six single-name memos are linked below ; recurring-series coverage touching this universe will be listed here as it is published.
- 9989.T Sundrug Read the memo
- 3349.T Cosmos Pharmaceutical Read the memo
- 3088.T MatsukiyoCocokara Read the memo
- 7649.T Sugi Holdings Read the memo
- 3549.T Kusuri no Aoki Read the memo
- 3391.T Tsuruha Holdings Read the memo
None yet for this bucket.
None yet for this bucket.
The information provided on this website is for informational and educational purposes only and should not be construed as financial, investment, legal, or tax advice. All content reflects the personal opinions, interpretations, and analyses of the author at the time of writing and is subject to change without notice. Nothing contained herein constitutes, or should be interpreted as, a recommendation, solicitation, or offer to buy or sell any securities, financial instruments, or other investment products. The author is not a licensed financial advisor, broker, or investment professional. Any references to specific assets, markets, or strategies are illustrative in nature and do not constitute personalized investment advice. Investing in financial markets involves risk, including the potential loss of capital. Past performance is not indicative of future results. Readers are solely responsible for their own investment decisions and should conduct their own independent research and due diligence before making any financial commitments. You are strongly encouraged to consult with a qualified financial advisor, legal professional, or other relevant specialist before making any investment or financial decisions. By accessing and using this blog, you agree that the author shall not be held liable for any direct or indirect losses, damages, or consequences arising from the use of, or reliance on, the information presented herein. All content is provided "as is" without any warranties of completeness, accuracy, or reliability.