Sundrug Co., Ltd.9989.T
The name says drugstore, but the drugstore is not the story. Split the group by format and one engine — the Direx discount business, 43% of revenue — turns out to have produced all of the decade's profit growth while earning a higher return on its assets than the pharmacy chain it is filed beneath. The market prices both engines at the same floor multiple, as if a single mature franchise were quietly declining. Valued part by part, the discount that opens up is real but moderate: a mispriced growth aisle on a net-cash balance sheet, far shallower than the sector screen advertised.
The drugstore, at ¥39.4bn of segment EBITDA on the 6.75x multiple a mature no-growth franchise earns, is worth roughly ¥266bn.
The Direx discount business — ¥26.2bn of segment EBITDA, an 11.4% return on its assets and a decade of compounding — carries a growth multiple of 8.5x and adds ¥222bn. Enterprise value reconstructs to ¥489bn.
Net cash of ¥21.6bn, long-term investments of ¥6.5bn, less the ¥2.2bn pension, bring equity to ¥514bn, or ¥4,397 per share.
The market capitalises Sundrug at ¥449bn, ¥3,838 a share. The composition discount is real — but at +14.6% it is moderate, not the deep undervaluation the sector screen implied.
Read at the consolidated line, Sundrug looks like exactly the thing the market has filed it as: a mature Japanese drugstore whose economics are slowly eroding. Over ten years the operating margin has slipped a point, from 6.56% to 5.56% ; reported return on capital has fallen from 17.3% to 9.8% ; the reported free-cash-flow margin has thinned from 4.95% to 1.54%. The share sits at the absolute decade floor on price-to-earnings, EV/EBITDA and price-to-book at once, with a dividend yield at its ten-year ceiling. On the surface, the price is pricing a franchise in structural decline, and the surface is not lying about the consolidated numbers.
The question is whether that erosion is a drift or an artefact. Because underneath the single line are two businesses that behave nothing alike. Over the decade, drugstore operating profit grew 7.5% in total — essentially flat, a 0.73% annual rate on 57% of revenue. Discount operating profit grew 159%, a 10.0% annual rate, and its share of consolidated profit rose from 22.6% to 41.3%. All of the group's profit growth, in other words, came from the aisle its name does not mention. And the discount is not the thin, dilutive aisle the label suggests: it earns 11.4% on its segment assets against the drugstore's 8.3%, turns them 2.15x a year against 1.44x, and matches the pharmacy chain on operating margin. The market is pricing the group's better engine as its drag.
That is the dislocation the dossier exploits, and the tool that captures it is a sum-of-the-parts by format rather than any consolidated multiple, because a single multiple does precisely what the market does — it averages a growing engine and a stagnant one into a floor number and calls the whole thing tired. Valued separately, the drugstore on a mature multiple and the discount on a growth multiple, the parts reconstruct to ¥4,397 against a spot of ¥3,838. The composition discount is there. But at +14.6% it is moderate — a first read of "best quality, lowest multiple in the bucket" would have you expecting something wider.
It is moderate because the quality underneath is only moderate, and honesty about that is what separates a false negative from a value trap. The moat is thin — efficiency, not a defensible asset. The return on capital, read ex-cash to strip out the balance-sheet cushion, is healthy at 16.4% but has fallen roughly twenty points over the decade, and the capital reinvested most recently earns a 6.5% incremental return, barely above the cost of capital. The drugstore that makes up most of the group has not compounded in ten years and has no monetisation layer to switch on. This is a sound operator with a mispriced growth aisle, not a hidden compounder.
The position framing is patient observation with a documented long bias, not ownership at this level. The floor is defended — a net-cash balance sheet, a 3.4% dividend, defensive demand — so the cost of being early is contained, but the weighted asymmetry is only ~10% and the catalyst that would close it, a return of capital, has not appeared in a decade. Conviction is moderate. What is worth watching is the discount same-store line and any signal on capital ; both print on the published calendar.
The decade reads in three regimes. Through FY March 2020 the two engines grew together on an over-capitalised balance sheet — revenue compounding above 5%, the margin holding near 6%, net cash building toward ¥80bn. Then the pandemic split them: the drugstore took an air pocket as the cold-medicine rush of FY March 2021 gave way to its counterblow, operating profit falling to ¥21.7bn at the FY March 2022 trough, while the discount quietly accelerated. The third regime, from FY March 2024, is the one the market has priced — capital redeployed aggressively into store construction, capex more than tripling, the cash cushion drawn down to ¥21.6bn, reported return on capital sliding to 9.8%, and the multiple falling to its decade floor. Through all three, discount profit kept climbing.
| Inflection | FY 2016Pre-cycle | FY 2020Balanced peak | FY 2022COVID trough | FY 2024Redeployment | FY 2026De-rated floor |
|---|---|---|---|---|---|
| Revenue (¥bn) | 503.8 | 617.8 | 648.7 | 751.8 | 842.5 |
| EBIT (¥bn) | 33.0 | 36.6 | 34.1 | 41.0 | 46.8 |
| EBIT margin | 6.56% | 5.93% | 5.25% | 5.45% | 5.56% |
| Return on capital | 17.3% | 13.2% | 11.2% | 11.1% | 9.8% |
| FCF reported (¥bn) | 24.9 | 22.0 | 12.9 | 4.8 | 13.0 |
| Net cash (¥bn) | −52.2 | −80.5 | −88.8 | −33.8 | −21.6 |
| Net income (¥bn) | 21.6 | 23.7 | 23.9 | 29.1 | 31.4 |
| Diluted EPS (¥) | 178.4 | 202.7 | 219.8 | 249.1 | 268.4 |
| DPS (¥) | 42.5 | 68.0 | 71.0 | 114.0 | 131.0 |
Source: Excel Income Statement / Capital Structure + Tanshin, FY March close-year basis. EBIT = reported operating income. Reported ROC (RETURN_ON_CAP) falls 17.3% to 9.8% ; ex-cash ROIC, reconstructed to strip the cash cushion, runs 36.4% to 16.4% over the same window — higher in level, steeper in decline. Normalised net income tracks reported within 1% throughout: no one-off overhang to purge.
Three decisions of capital allocation sit behind the de-rating. The drugstore was left to stagnate — 57% of revenue tied up in a franchise that stopped compounding, with no private-label or data layer built to differentiate it, while the higher-return discount was given the smaller share of investment. That misdirection is the second decision: on a rough segmentation, 60% of the decade's cumulative capex went to the 8.3%-return drugstore for barely ¥1.9bn of incremental operating profit, against ¥11.9bn from the discount on 40%. And in a sector the market re-rates only on the return of capital, Sundrug has bought back no stock in ten years, letting a net-cash balance sheet and a floor price-to-book sit untouched. The dividend has tripled and the payout has climbed to 48.8% — a real improvement — but a rising dividend is the passive half of shareholder return, and the active half has never arrived.
The engine only makes sense once you read same-store sales by format, because that is where organic demand hides behind the growth from new stores. Two pockets carry it. The discount is low-cost food frequency — near-daily traffic, and a tailwind whenever inflation pushes shoppers toward value ; its same-store line has been positive every year and beat the drugstore's every year, confirming from the demand side which engine actually works. The drugstore is health defensiveness — ageing-population dispensing and over-the-counter, stable but administered and not captive. Neither pocket is contractually sticky, and both decelerated together in FY March 2026, discount same-store to +1.7% from +3.3%, drugstore to +0.8%. The demand is defensive — the lowest beta in the bucket — but the loyalty behind it is shallow.
What the group does not have is real pricing power, and it is worth being precise about that because the gross margin has quietly improved. Gross margin rose about a point over the decade, to 25.66%, even as the lower-margin discount gained weight in the mix — which points to buying scale and mix, not the ability to charge more without losing the customer. Sundrug never publishes the volume/price/mix split that would settle it, and it carries no private-label or monetised-data layer of the kind Matsukiyo built. Its monetisation is that of an efficient standard operator, not the owner of an intangible asset.
The cost that governs the margin is labour, and it is not passed through. The bridge decomposes the decade's one-point EBIT erosion cleanly: gross margin added 1.02 points, cash costs — labour and structure — took 1.06, and depreciation, the capex ramp, took 0.96. Headcount grew 54% for an 8.5% gain in revenue per employee, so the efficiency lever that long absorbed wage inflation is thinning. If Japanese wage inflation stays above what remains of that lever, the ~5.55% margin floor is where the pressure lands first ; there is no commodity or currency cost to blame, only people.
The cash bridge is where the reported picture misleads and the underlying one reassures. Operating cash conversion is excellent — cash from operations runs 92.5% of EBIT — and the ex-growth free cash flow, the maintenance machine stripped of expansion capex, has held near ¥24–27bn a year through the decade. Reported free cash flow collapsed to ¥4.8bn at the FY March 2024 trough purely because capex tripled, not because the cash engine broke. Set against that is a balance sheet doing almost nothing: net cash of ¥21.6bn, a payout still under half of earnings, and no buyback in a decade. That un-returned net-cash balance sheet is the option the price does not hold.
This pillar carries the thesis, because the whole sum-of-the-parts rests on the discount being a genuine return engine. At the level, it is: ex-cash return on capital of 16.4% sits six to ten points above the cost of capital, cash conversion is 92.5%, and the ex-growth free-cash-flow machine is stable. But two cellular facts hold the score down. The ex-cash return has fallen roughly twenty points over the decade — a healthy spread that is compressing, not a stable one — and the capital reinvested most recently earns 6.5%, barely at the cost of capital. The engine that works is real ; the return on new capital poured into it is the open question.
Management is the second cardinal because it is the swing that decides both catalysts of a re-rating — whether the return stabilises and whether capital gets returned. The operator is good: earnings clean over ten years, guidance beaten four years running, the store base maturing gracefully. The allocator is weak. Sixty percent of the decade's capex went to the lower-return drugstore for almost no incremental profit ; the balance sheet was over-capitalised and then deployed late at a 6.5% incremental return ; and the drugstore was left to stagnate without a monetisation layer. A buyback and evidence of incremental-return discipline would move this pillar ; neither has come.
Defensive and low-beta (0.26), anchored to food frequency and dispensing, the store base maturing without dilution. But captivity is low — no monetised loyalty — and both same-store lines decelerated together in FY March 2026.
The weakest pillar, and structural. The lowest SG&A-to-gross-margin ratio in the bucket is execution discipline, not a barrier — no private-label or data layer, replicable formats, near-zero switching costs. The quality is operational, an efficient way of running a commoditised format with no rent underneath it.
No control discount — an independent issuer, unlike the Aeon-controlled names — with a tripled dividend and a 48.8% payout. But restitution is passive: no buyback, an under-deployed net-cash balance sheet, no TSE-aligned plan to exit the floor price-to-book.
Solid but not exceptional — a good operator with a thin moat, healthy but declining returns, and a capital-allocation problem. Above a value trap: the earnings are clean, the spread positive, the demand defensive. Below a first-rank compounder: the moat and the return trajectory forbid the grade. The decisive pillar is management, because it alone controls whether the return stabilises and the capital is returned — the two levers of the re-rating.
Is the discount a durable compounder, or a plateau the floor multiple is right to price ?
Is the drugstore a defensible floor, or a latent decline ?
Opinion is split between a stable mature cash cow and a slowly declining franchise, and the floor multiple prices the second. The tension is real: drugstore operating profit has been flat for a decade, carries no monetisation layer, and absorbed capex at almost no incremental return — the marks of a franchise that has stopped compounding. But its same-store line stays weakly positive (+0.8%), never firmly negative outside the COVID counterblow, and its 8.3% return on segment assets remains above the cost of capital. Stagnation, on the evidence, rather than decline.
Is the restitution a latent catalyst, or a structural absence ?
The price holds no re-rating — a floor price-to-book, a dividend yield at its decade ceiling, the optionality treated as if it did not exist. Against that sits a net-cash balance sheet, a payout still under half of earnings, and zero buybacks in a sector the market re-rates only on the return of capital. The optionality is maximal and entirely un-priced ; its realisation depends on a managerial pivot that has not come in ten years. This is the lever the bull case is built on.
At ¥3,838 and ~6.5x EV/EBITDA, the market applies 6.52x uniformly to both engines — which values the growing discount at the same multiple as the stagnant drugstore. Back out a 6.75x drugstore and the discount is left at roughly 6.17x, against the 8.5x its 11.4% return earns : that gap is the alpha. The floor on all three metrics at once — 14.2x trailing earnings, 6.52x EV/EBITDA, 1.57x book — encodes terminal stagnation of both engines, no margin expansion, and no return of capital despite a net-cash balance sheet. The clean-earnings P/E channel corroborates the parts: 15.5x, still under the 17.9x decade mean, on guided EPS lands near ¥4,260, in line with the sum-of-the-parts Base. Valued part by part, the discount is real ; valued as one line, it disappears.
The discount same-store line crosses durably below +1% (the scale lever plateauing), the drugstore tips into negative same-store, and un-passed-through wage inflation compresses the blended margin below 5.3%. The market de-rates both engines to their historical floors — drugstore 5.75x, discount 6.5x. The floor holds at ¥3,419 on a 3.8% dividend yield, a 1.40x price-to-book and net cash intact. A timing disappointment, reversible, not a permanent impairment.
Sundrug executes without surprise — the discount grows on openings and modest same-store, the drugstore holds its floor, the margin stabilises near 5.55%, capex decelerates and reported free cash flow lifts. The market keeps a floor multiple, but the sum of the parts reveals the composition discount: drugstore 6.75x, discount 8.5x, enterprise value ¥489bn, equity ¥4,397. A consolidated re-rating may or may not come ; the fair value does not need it.
The two un-priced levers fire together. Discount same-store re-accelerates above +3%, confirming the compounder, and management activates restitution — a buyback and a payout lift — triggering the sum-of-the-parts re-rating the sector only rewards on the return of capital. Drugstore 7.5x, discount 10x. The path needs both the operational lift and the allocation decision, and neither is signalled today, which is why the bull carries the thinnest probability.
| KPI | Latest value | Status | What it tells us |
|---|---|---|---|
| Discount segment same-store sales | +1.7% FY March 2026 | Cardinal | The swing variable. Decelerating from +3.3% ; positive and above the drugstore every year. Durably below +1% over two consecutive halves confirms the plateau and pulls fair value toward ¥3,419. |
| Discount segment OP margin | 5.31% FY March 2026 | Holding | The value anchor of the growth engine. Below 5.0% at a full-year print is the thesis breaker — the composition discount becomes a justified floor multiple. |
| Drugstore segment OP | ¥27.5bn FY March 2026 | Watch | The soft zone. Flat for a decade but stable. Below ¥26bn on a clean year, or same-store negative over two halves, turns the false negative into a value trap. |
| Incremental ROIC (3-yr rolling) | 6.5% FY March 2024–2026 | Priced | Barely at the ~6% cost of capital. Below 6% confirms dilutive deployment ; a lift above 10% as the recent store base matures confirms the capex was an investment peak, not a drift. |
| Capital restitution | Zero buybacks · payout 48.8% | Trigger | The main un-priced upside. A buyback, a payout target above 60%, or a TSE-aligned plan activates the re-rating the sector pays for. Absent for ten years. |
| EV/EBITDA (blended) | ~6.5x | Reference | Against a decade corridor of 5.8–11.7x. Above 9x reads re-rating toward the bull ; below 5.7x reads de-rating to the bear floor. |
| FCF ex-growth yield | 5.5% | Reference | The maintenance machine, stripped of expansion capex ; reported yield is 2.9%, depressed by the ramp. Approaches the ~6–7% cost of equity and sets the valuation floor. |
| Lease-adjusted gap | ¥6.2bn vs ¥34.9bn | Reference | non-cancellable minimum against P&L rent — a structural J-GAAP disclosure gap, not an error. The principal unresolved downside to the SOTP, pending the Yuho maturity schedule. |
The case turns positive if the discount stops merely growing and re-accelerates, its same-store line moving durably above +3% across two consecutive halves, or if a quantified multi-year capital-return policy puts the net cash and under-deployed balance sheet to work. Either would move the dossier from watchlist to long and open the bull path. Both are observable ; neither is signalled today.
The case turns negative if the narrow engine plateaus. Discount same-store below +1% over two consecutive halves, or the segment operating margin below 5.0% at a full-year print, would confirm that the group's only compounder has stopped compounding — and convert the composition discount into a floor multiple the market would then be right to hold.
The path to permanent loss, distinct from a timing disappointment, needs two things at once: a structural contraction of the discount — multi-year negative same-store and margin below 4.5% — combined with the net cash redeployed into an acquisition earning below the cost of capital. That conjunction, and only that conjunction, turns a reversible de-rating into a destruction of capital. Currently not signalled.
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.