Yonex Co., Ltd.7906.T
The reported numbers say bargain : a real organic compounder, earnings up seven-fold in six years, trading at 16.8x after a 28% sell-off. Strip the weak yen out of the operating margin and the bargain mostly disappears — normalised earnings put the stock at roughly 19.8x, on its own five-year average rather than below it. What is left is a genuine franchise whose quality is conditional on two things it does not control : one market, China, and one currency. The question is whether that conditionality is priced correctly.
At ¥2,377 the stock trades at 16.8x reported earnings of ¥141.42 — which looks cheap against a five-year average near 18.8x and a ten-year average closer to 33x.
Normalise the operating profit to mid-cycle exchange rates — USD/JPY 130 against the ¥149.70 actually booked — and earnings power falls to roughly ¥120 a share. At that level the multiple is about 19.8x : not below its history, on it.
Valued geographically — Asia at 11x its normalised ¥10.5bn of profit, Japan at 9x, a depressed North America at 18x on recovery optionality, Europe at 10x — the enterprise reconstructs to ~¥160bn ; net cash of ¥17bn, less an underfunded pension, takes equity to ~¥163bn, or ~¥1,906 a share before any forward growth is added back.
The market capitalises Yonex at ~¥203bn. The bargain the reported multiple advertises is a currency effect, not a discount.
The interesting thing about Yonex is that the decade splits cleanly in two, and almost everything worth knowing sits on the right-hand side of the break. For the six years to FY March 2020 this was a competent, dull exporter : operating margin stuck between 4% and 7%, return on equity below 9%, earnings per share actually drifting down from a 2017 peak. Then COVID knocked the operating margin to a 2.0% trough in FY March 2021, and out of that trough the company became something else entirely — margin re-rated to ~10% and held there for five consecutive years, return on equity around 16%, earnings up roughly seven-fold. The franchise only started compounding after the pandemic. The quality is real, and it is recent.
The question the dossier turns on is what that re-rating actually rests on. The answer, uncomfortably, is one market. Asia is 60.8% of revenue, China is around 91% of the Asian segment, and applying the Asian margin to the Chinese share puts roughly 65% of consolidated operating profit inside a single sovereign jurisdiction. The engine that produced the re-rating and the risk that could end it are the same variable. So the binary is sharp : is the Chinese growth a secular participation story — more people playing badminton, buying a desired brand — that justifies a compounder, or is it a concentrated single-market dependence that the 2016-2019 punitive derating already rehearsed once ?
That historical rhyme is worth holding precisely, because the two episodes are mirror images of each other. 2016-2019 was a multiple problem : the market paid 55-63x earnings ahead of growth that had not yet arrived, then took the price down 52% over three and a half years when the earnings failed to follow. 2026 is the inverse — the earnings have been delivered, and the multiple has already been purged from the ~27x momentum peak of mid-2025 to 16.8x today. The risk this time is not a deflating multiple. It is operational : whether the Chinese earnings hold, something observable quarter by quarter.
The second thing to hold is the currency. The ~10% margin is partly borrowed. Translate the overseas profit back at a normalised yen and roughly 1.5 to 3 points of that margin go away ; normalised earnings per share fall from the reported ¥141.42 toward ¥120. The cheapness on the screen is, in large part, a monetary illusion of margin level. What it is not is a fake growth story — the +18.8% revenue growth ex-FX in FY March 2026 is almost identical to the +18.3% reported, with an FX impact of just −¥633m. The growth is organic and the level of margin is flattered, and the two facts have to be kept separate or the whole valuation goes wrong.
The position framing is patient observation at this level. At ¥2,377 the FX-normalised asymmetry is modestly negative and two cardinal variables — the isolated Chinese operating profit and the exact FX elasticity of margin — are not yet reconstructed cell by cell. The thing worth watching is the Asian segment's ex-FX growth and margin through the FY March 2027 prints. Conviction is moderate ; the dossier passes to a full modelling cycle rather than into the book.
The cleanest way to read the last decade is as three regimes with a hinge in the middle. From FY March 2015 to 2020 Yonex was a plateaued, profitable exporter — revenue grew at a ~5% compound rate, the operating margin sat between 4% and 7%, return on equity stayed below 9%, and earnings per share peaked at ¥34.82 in 2017 and then fell. COVID was the hinge : FY March 2021 saw revenue drop 16.8% and the operating margin collapse to 2.0%, the absolute trough. What came after is the company that exists today — revenue multiplied 3.2x from the trough to ¥163.6bn, the operating margin re-rated and then held near 10% for five straight years, return on equity around 16%, return on invested capital ex-cash close to 18%. The shape matters because it makes the relevant modelling base the post-2021 regime over the decade average — and because five years is short enough that durability has to be stress-tested.
| Inflection | FY 2017Pre-COVID peak | FY 2020Plateau end | FY 2021COVID trough | FY 2023Re-rating | FY 2026Current |
|---|---|---|---|---|---|
| Revenue (¥bn) | 61.0 | 62.0 | 51.6 | 107.0 | 163.6 |
| EBIT / OP (¥bn) | 4.15 | 2.42 | 1.03 | 10.06 | 16.55 |
| EBIT margin | 6.8% | 3.9% | 2.0% | 9.4% | 10.1% |
| Return on equity | 9.0% | 4.3% | 2.8% | 14.9% | ~16% |
| Return on capital | 8.2% | 4.0% | 2.6% | 13.9% | ~13.7% |
| FCF (¥bn) | 0.6 | 1.5 | 3.8 | −0.5 | ~0.1 |
| Net Income (¥bn) | 3.04 | 1.65 | 1.10 | 7.33 | 12.09 |
| Diluted EPS (¥) | 34.82 | 18.89 | 12.59 | 84.05 | 141.42 |
Source: data pack (series, FY-March vintage rewritten), Tanshin FY March 2026 (actuals), Ratios tab (margins). FY = year ended 31 March. FCF FY March 2026 is depressed by the capex peak (¥9.4bn, +74%) and working-capital absorption — a cycle artefact ; cash from operations held at ¥9.5bn. Return on capital is the own reported figure ; the ex-cash measure is ~18%.
Three management decisions shape the picture, and they are not all flattering. The first is that the pre-COVID plateau was allowed to run : five to six years of return on invested capital reported in the 2-7% range, below an ~8% cost of capital, with the strategic inflection (the Global Growth Strategy) only formalised in 2023, after the turn rather than ahead of it. The second is North America, where the company is expanding under scale : operating profit there fell 54.2% to ¥256m on a 3.5% margin even as revenue grew 15.8%, because direct-to-consumer build costs arrive before the volume that pays for them. The third is the sovereign concentration itself, left to build without a hedge. Some of that is unavoidable — badminton is played in Asia, and you cannot diversify a sport's geography by decree — but the capital plan doubles down on capacity instead of a geographic counterweight, and it is the tennis share gains, arriving late, that are doing the de-risking the strategy did not plan for.
The engine is best read as participation times brand share times attach rate, and it comes in three pockets of decreasing certainty. The first is the badminton rent — around 62% of revenue, structurally driven by the sport's secular popularity in Asia, and the home of the concentration. The second is tennis, around 13%, and it is a different animal : its growth is share-driven rather than participation-driven, because Yonex is taking dollars from Wilson, Babolat and Head rather than waiting for more people to play. US specialty-store dollar share went from 5.6% in 2019 to 18.6% in 2025, and the brand is now No.1 in rackets above $250. That makes tennis the one growth leg that is independent of badminton and of China. The third pocket is the Head-to-Toe attach — shoes, apparel, bags, around 23% of revenue under "Others" and up 30% over two years — the spillover of brand desirability from the racket into an ecosystem.
Where the word "pricing power" earns its keep is worth being precise about, because the FY March 2026 operating-profit bridge says something specific. Gross margin actually slipped — it cost ¥919m — while gross profit rose ¥11,385m, entirely on volume. So the margin level is built from scale, mix and the currency rather than a clean price increase. The pricing power is real but defensive : it holds the price and protects the gross margin, it does not expand it. And the cost that governs the margin is not a raw material at all — it is selling and marketing intensity. SG&A rose ¥8,096m in FY March 2026 (advertising +¥1,922m, personnel +¥1,562m), which is the company deliberately re-spending its operating leverage into brand-building and the growth strategy. The margin is managed at a chosen pace and capped on purpose, governed by spend rather than by an input shock.
The cash conversion looks broken in the latest year and is not. Free cash flow fell to ~¥0.1bn in FY March 2026 because capex jumped 74% to ¥9.4bn — the new Niigata tennis-racket plant, the Toyama badminton plant under construction — and working capital absorbed another ¥3.6bn of inventory. But cash from operations was a solid ¥9.5bn, and the cash conversion cycle actually improved from 159 days at the COVID trough to 95 days. This is a buildout artefact. What it does mean is that the return on this reinvested capital is the open question : the operating leverage that took the margin from 2.0% to 10.1% on 3.2x the volume is proven, but the incremental return on the 2025-2027 capacity wave is not yet.
This pillar carries the thesis, because the nature of the Chinese demand decides everything else. The quality of the demand is high — participation-secular in badminton, a desired brand, a tennis share-gain that is genuinely independent of the badminton base. But the base is narrow : ~62% badminton, ~65% of operating profit in China, and a demand that COVID proved is participation-cyclical — revenue fell 16.8% and the margin hit 2.0% when the courts closed. The tennis traction (5.6% to 18.6% US share) is the evidence that a diversification vector exists. To earn a 4, the China share of operating profit would have to fall durably below 50% by diversification rather than by Asian contraction. The base holds ; the growth depends on China not relapsing.
The second cardinal because it is what makes the compounder claim a real one. Return on invested capital ex-cash runs ~18% against an ~8% cost of capital — a +10-point spread, the highest in the bucket — and the operating leverage behind it is demonstrated : 2.0% to 10.1% margin on 3.2x volume, on a net-cash balance sheet, with the cash conversion cycle cut from 159 to 95 days. The two limits are specific. The margin level borrows ~1.5-3 points from the weak yen, and capex intensity is rising (4.2% to 5.7% of sales) into a buildout whose incremental return is unproven. To earn a 4.5, the margin would have to hold ≥10% at a normalised yen and the growth capex would have to show a ≥15% incremental return through the North American inflection.
Real at the top, soft at the bottom. The athlete-endorsement flywheel and monozukuri carbon-graphite craft are hard to replicate, and the brand converts — No.1 US racket above $250 at 29.9% share. But the consumer switching cost is soft, there is no contractual lock-in, and Li-Ning and Victor compete in the Chinese badminton core.
Recent execution is credible — the margin re-rating and the GGS rollout (tennis share, DTC, capacity) were delivered. The longer record is weaker : five pre-COVID years of sub-cost-of-capital returns, the −54.2% North American operating profit, and a strategy that arrived after the turn. The concentration is uncovered.
The weakest pillar. The dividend is progressive (¥13 to ¥25, ¥28 guided) and return on equity is ~16%, but the payout is a thesaurising 17.7%, control is the Yoneyama family, minority treatment is undocumented, and the return reporting is opaque enough that a sectoral "zero buyback" claim coexisted with a real ¥2.4bn one.
A solid franchise with strong conditional reserves, not an unconditional compounder. The operating quality is genuine — a +10-point ROIC spread reinvested — but it is doubly gated, by one market and one currency, and held back by governance. Above a value trap, below a clean compounder such as Food & Life (19-20/25). The grade is consistent with the valuation : the operating excellence is real, and the concentration and the FX both stop it short of a premium.
Is the Chinese growth secular participation, or a concentration that the market is right to discount ?
Is the ~10% margin structural, or borrowed from scale and the yen ?
The bulls call it the new normal ; the bears call it a peak flattered by a weak yen and due to regress. The bridge supports caution : the gross margin slipped while gross profit rose on volume, the GGS target of ≥10% reads like a ceiling rather than a floor, and ~1.5-3 points are on loan from the currency. The level combines scale, mix and FX, not a price increase. Whether it is durable is exactly what the modelling cycle has to settle.
Is the North America / DTC / capacity build accretive, or over-extension ?
One camp sees future growth ; the other sees margin destruction and a capex burn without return. The facts cut both ways : North American operating profit is down 54.2% on a 3.5% margin, capex is 5.7% of sales, but the US share gains are genuine. The incremental return on the growth capex — the new plants, the 1.7x tennis capacity — is the variable that decides whether this is a leg or a sink, and it is not yet proven.
At ¥2,377 — 13.9x forward consensus, 15.4x on the guided ¥154.29 — the market is pricing the deceleration. The FY March 2027 guidance of +8.8% sales, +7.6% operating profit and a 10.0% margin reads as a normalisation from the ~+18% recent pace toward the top of the 7-10% GGS band, and the derating from the ~27x mid-2025 peak to 16.8x has aligned the multiple with single-digit growth against the recent ~26% compound. What is also embedded, in the depth of the multiple, is a non-zero probability of a Chinese shock — the normalised P/E of ~19.8x is not a compounder premium, it is a conditional grower's multiple. What is under-priced is the de-risking optionality the market does not see while it reads Yonex as a pure Chinese badminton bet : the share-driven tennis runway, the Head-to-Toe attach, and a possible North American inflection. The reported 16.8x looks cheap ; normalise the yen and it is on its own five-year average.
Ex-FX Asian growth stalls — Li-Ning and Victor plus consumption nationalism erode the badminton core — at the same time as the yen normalises toward 115, removing the margin tailwind. The market, not distinguishing the stall from healthy deceleration, applies the Rule-4 punitive derating ; 2016-2019 (−52% over 3.5 years) is the reference. EPS ~¥102 on a ~13x derated multiple. The downside is bifurcated : a COVID-type participation disruption is reversible, with a ~¥1,700-1,800 floor on normalised FCF plus net cash ; a structural share loss or sovereign shock is a permanent re-rating, the floor falling to ~¥1,400-1,500.
GGS executes without surprise : Asian participation stays structural, tennis share gains continue at a decelerating pace, attach and geographic diversification progress. Growth normalises mechanically from ~+18% toward ~+8%, the reported margin holds ~10% but ~8.6% at a normalised yen, and the China concentration neither resolves nor worsens. Normalised forward EPS ~¥130 on ~17-18x with no re-rating and the sovereign discount embedded. The central target lands at ~¥2,350 — essentially on the spot.
The under-priced levers fire together. Tennis and Head-to-Toe attach accelerate, the China share of operating profit falls below 50% on diversification, North America inflects above a 6% margin, and a weak yen persists to preserve the ~10% level. The de-risking of the sovereign concentration earns a modestly higher multiple — never a base-case re-rating. EPS ~¥145 on ~19-20x, central target ~¥2,875.
Probability-weighting the three scenarios gives a fair value of ~¥2,245, an asymmetry of ~−5.6% against the ¥2,377 spot. The upside-to-downside ratio is ~0.64x — the −33% bear exceeds the +21% bull in amplitude. The 28% sell-off year-to-date purged the momentum premium but did not create a margin of safety at a normalised yen ; the stock is pricing its forward growth without a cushion. The geographic sum-of-the-parts, valued statically on normalised operating profit, lands lower at ~¥1,906 — the floor without forward growth — while the forward-P/E channel at ~¥2,350 carries one to two years of it. The truth sits in a ¥1,900-2,400 Base range, centred ~¥2,200-2,350, with the gap to spot showing the absence of any discount once the currency is normalised.
| KPI | Latest value | Status | What it tells us |
|---|---|---|---|
| Asia segment ex-FX growth | +18.8% FY3/26 (consol.) | Cardinal | The swing variable. Below +10% across two consecutive quarters at FY March 2027 reclassifies the China concentration from conditional to active risk and pulls fair value toward the ¥1,600 bear. |
| Asia segment margin | 13.9% FY3/26 | Holding | ~71.5% of segment operating profit. The value anchor. Falling below 12% would confirm de-leverage on falling Asian volume and the cyclical-vulnerability reading. |
| Tennis US dollar share | 18.6% (2025) | De-risking | Up from 5.6% in 2019 ; No.1 racket above $250 at 29.9%. Share-driven, independent of badminton and China — the main vector that could take China below 50% of operating profit. |
| OP margin · reported vs FX-normal | 10.1% / ~8.6% | Priced | The reported level borrows ~1.5-3pts from the weak yen. Holding ≥10% as the yen normalises validates structural ; below 9% confirms borrowed. |
| North America segment margin | 3.5% FY3/26 | Watch | Operating profit −54.2% on +15.8% revenue — DTC build under scale. Inflection above ~6% validates accretion ; below 5% into a second year confirms over-extension. |
| Capex / cash conversion | ¥9.4bn capex (+74%) | Watch | FCF ~¥0.1bn is a buildout artefact ; CFO held ¥9.5bn, CCC improved 159 → 95 days. Normalised mid-capex FCF ~¥7-8bn. The incremental ROIC on the capacity wave is unproven. |
| China share of operating profit | ~65% (proxy) | Trigger | Asian margin × China 90.9% of Asian revenue. Falling durably below 50% on diversification — not Asian contraction — is the de-risking the price does not hold. |
| P/E · reported / FX-normal | 16.8x / ~19.8x | Reference | Against a ~18.8x five-year average and ~33x ten-year. The reported figure reads cheap ; the normalised one sits on the five-year average — no discount. |
The case turns positive if the diversification stops being a hope and starts being a number. Ex-FX Asian growth holding above +10% with the segment margin above 12% through the FY March 2027 prints would dissipate the stall premium ; the China share of operating profit falling below 50% by diversification rather than Asian contraction, or a North American margin inflecting above 6% after several quarters of growth, would convert the de-risking optionality the price ignores into an earned multiple. Each is observable on the quarterly calendar ; none is signalled today.
The case turns negative if the narrow engine stalls. Ex-FX Asian growth below +10% across two consecutive quarters, or the Asian segment margin falling below 12%, would reclassify the China concentration from conditional to active and trigger the Rule-4 punitive derating — a stalled grower priced as a grower, the 2016-2019 reference. The distinction that matters is reversibility : a COVID-type participation disruption is a timing disappointment with a ~¥1,700-1,800 floor ; a structural loss of badminton share to Li-Ning or Victor, or a sovereign shock, is a permanent impairment with the floor at ~¥1,400-1,500.
The allocation risk is the one to watch alongside the demand. The capital plan commits ¥70-80bn to growth against ¥15-20bn to shareholder return, and it doubles on capacity rather than on geographic de-risking, placing capital onto the concentration. If the North America / DTC / capacity buildout proves non-accretive, with incremental return below cost of capital, it immobilises capital without creating value and the recovery optionality in the North American sum-of-the-parts is overstated. The full modelling cycle exists precisely to isolate the Chinese operating profit and reconstruct the FX elasticity ; the ¥120 of normalised earnings, not the ¥141 reported, is the number to underwrite.
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.