On Holding AG (NYSE: ONON) — A Best-in-Class Brand Priced for a HOKA-Style Fade It Isn’t Showing Yet

Independent equity research. Report date: 2026-07-03. All figures in Swiss francs (CHF) unless noted; USD conversions at an assumed 1 CHF ≈ US$1.25. On reports IFRS in CHF; the shares and market capitalization are USD.

⚡ Claude’s Take

This block is the author’s own subjective opinion. It is general information, not investment advice, and not a recommendation to buy or sell any security. The analysis that follows takes no position and carries no price target — it discusses valuation only as embedded expectations.

Verdict: HOLD / accumulate-on-weakness / not-a-short — a genuinely high-quality compounder de-rated to its cheapest-ever multiple, where the price already pays for a fashion-fade the fundamentals have not yet begun to show. Conviction: medium. Framing: quality-growth-at-a-reasonable-price / abandoned momentum — explicitly not a deep-value or falling-knife call. Directional zone: I’d accumulate below ~$34 (the 52-week-low zone), am a comfortable holder into the low-$40s, and see a defensible fair-value band of ~$44–$55 (roughly 28–34× a normalized ~$1.20 of per-share operating earnings, a multiple its ~30% growth and 62.8% gross margin can carry) — with genuine bull optionality above that if the growth durability proves out. This is not a table-pounding buy, because the moat is real but narrow and the absolute multiple isn’t “cheap”; it is a business I’d rather own on weakness than short into strength.

The market has done something unusual to On: it halved the multiple (from ~7–13× EV/Sales to ~2.8×; the 2.9th percentile of its own post-IPO history) while the business accelerated — FY25 revenue +30% (all organic), operating income +78%, gross margin to a record 62.8%, net cash ~$1.3B, and the first GAAP operating profits scaling. The de-rate is a multiple event, not an earnings event — the DECK/HOKA “fallen-darling” template, not the Nike (earnings trough) or Lululemon (earnings decline) one. My reverse-DCF says ~$10.8B EV embeds only a ~10–11% five-year revenue CAGR — roughly a third of today’s run-rate — so a HOKA-style fade is largely pre-paid. Three things keep me at HOLD rather than BUY: (1) the moat is Greenwald’s weakest form — a single-brand, ~93%-footwear, zero-switching-cost intangible untested through a full fashion cycle, with HOKA’s +58%→+16% deceleration as the live cautionary comp; (2) ~32× normalized earnings is not absolutely cheap, only cheap-vs-its-own-bubble; (3) beta 1.46 means it falls harder than the group in any risk-off tape. But the tells lean constructive: a co-founder (Olivier Bernhard) bought ~$2.2M in the open market near the lows, the FY26 CHF→USD functional-currency switch mechanically removes the FX optics that manufactured the “earnings are falling” scare, and growth is still ~26% constant-currency in every region with apparel/China/DTC barely penetrated. What flips me bullish: two more quarters of 25–30% constant-currency growth with gross margin held ≥62% (fade disproven → re-rate). What flips me bearish: constant-currency growth breaking below the high-teens with an inventory build — the first quarter On becomes HOKA. Tag: “the best-priced growth in footwear — if it isn’t the next Under Armour.”

📈 Stock Price Action — Five-Year Event Map

Text-only by design. Price moves are FACT (AZI daily CSV, IPO 2021-09-15 → 2026-07-02); attributed drivers are INTERPRETATION. No price target, no recommendation, no chart-pattern language.

The arc. On IPO’d on 2021-09-15 priced at $24, opened near $35, and spiked to ~$55 within two months — then round-tripped the entire 2022 growth-stock bear market to a low of $16.38 (2022-05-11). From that trough it compounded roughly 4× on a genuine profitability inflection to an all-time high of $63.62 (2025-01-30), before a tariff shock, a strong-franc reported drag, and a high-beta/growth-momentum unwind cut it to a 52-week low of $31.88 (2026-03-30). It has since bounced to $36.83 (2026-07-02) on record Q1-26 margins — leaving it ~42% below its ATH, inside a 52-week range of $31.88–$54.36.

Cycle narrative. (1–2) The IPO landed at the top of the 2021 growth bubble and gave it all back as rates rose — a valuation event, not a business event. (3–4) The 2022→2025 quadruple was earned: On turned FCF-positive, pushed gross margin past 60%, and built demonstrable brand momentum, culminating in the January-2025 ATH. (5) The February–April 2025 collapse was macro/policy — the “reciprocal tariff” regime and a high-beta rotation hit ONON harder than most given its 1.46 beta. (6) The May-2025 spike proved the fundamental intact — a clean beat re-rated it back near the highs. (7) The long slide to March-2026 is the crux: even a strong Q4-25 print couldn’t arrest a de-rating driven by the guided growth step-down, the optical strong-franc revenue drag, and tariff whipsaw, with the sharpest leg (−19% on 25–26 Mar) a broad macro/tariff sell-off, not company-specific. (8) The mid-2026 bounce is quality-led — record gross and EBITDA margins plus a raised profitability guide — but leaves the stock still ~42% below its own peak.

1. Executive Summary

On Holding AG is a ~15-year-old, Zurich-based premium performance-running and sportswear brand (“On” / CloudTec) that has scaled from CHF 425M (FY20) to CHF 3,014M (FY25) — a ~47% five-year CAGR, entirely organic — while earning the best gross margin in the athletic-footwear cohort (62.8%, versus Nike ~44%, Deckers/HOKA ~58%, Lululemon ~57%). It is an asset-light, design-and-demand-creation house: it outsources essentially all production to fewer than 30 suppliers (~90% of footwear from Vietnam), sells through wholesale (58%) and a fast-growing, higher-margin direct-to-consumer channel (42%), and carries net cash of ~CHF 1.0B with no financial debt.

The investment tension is unusually clean. The business is high quality and still hyper-growing — FY25 revenue +30% (+35.6% constant-currency), operating income +78% to CHF 377M (12.5% margin), adjusted EBITDA CHF 567M (18.8%), ROIC ~18% and rising, growth double-digit in every region (Americas +23% cc, EMEA +35% cc, APAC +107% cc) and every category (shoes +33%, apparel +76%, accessories +135% cc). Yet the stock has de-rated to the cheapest multiple of its public life — ~2.8× EV/Sales and ~15.6× EV/adjusted-EBITDA (2.9th percentile of its own post-IPO history), down ~42% from its January-2025 all-time high.

Two things make the reported numbers misleading, and both matter. First, GAAP net income is dominated by non-cash FX: FY25 net income fell to CHF 203.7M (from CHF 242.3M) even as operating income near-doubled, purely because a strengthening franc produced a CHF 173.2M foreign-exchange loss below the operating line — so any P/E on On is noise, and operating income / adjusted EBITDA are the clean lenses (a FY26 functional-currency change from CHF to USD mechanically removes this distortion going forward). Second, the share structure is a valuation trap: Class B “voting rights shares” carry only one-tenth the economics of Class A, so the correct economic-equivalent share count is ~331M (not the ~638M raw total) and the true market cap is ~$12.2B — roughly half what the raw count implies. On corrected numbers, On is the fastest-growing, highest-margin, net-cash name in footwear, trading at a PEG (~1.1) below Deckers’ — not the expensive outlier some cross-reads suggest.

The bear case is intellectually honest, not a strawman: On’s moat is a narrow, single-brand, ~93%-footwear intangible with zero switching costs, untested through a full fashion cycle, in a structurally mediocre industry (low entry barriers, contested wholesale shelf, tariff exposure, a documented history of hero-brand boom-and-fade). HOKA’s 2025 deceleration (+58%→+16%) is the live cautionary comp. A reverse-DCF shows the current ~$10.8B EV embeds only a ~10–11% five-year revenue CAGR — a HOKA-style fade largely pre-paid. The debate the market is really having is not whether the business works today (it plainly does) but how durable the brand and its growth are — and the price has already positioned for the pessimistic answer. This memo takes no position (see Claude’s Take for one); it lays out the evidence on both sides.

2. Business Overview

On Holding AG designs, markets, and sells premium performance running and sportswear under the On brand. Founded in Zurich in 2010 by an ex-professional triathlete (Olivier Bernhard) and two partners (David Allemann, Caspar Coppetti), it reached a CHF 3,014.0M (FY2025) net-sales run-rate (~$3.9B; ~$3.9B TTM through Q1-26) — a rate of scaling that is itself the central fact of the story and the central risk. On is a brand-and-demand-creation business, not a manufacturer: it designs in Zurich and outsources essentially all production to a focused network of fewer than 30 suppliers, with 6 partners accounting for ~70% of 2025 production; footwear is made by 10 suppliers (8 in Vietnam, 2 in Indonesia), apparel/accessories by 16 (Vietnam, Turkey, China, Indonesia). The only owned production is a small Zurich/Busan LightSpray operation (a robotic sprayed-upper technology — nascent optionality, not yet material). This asset-light structure is why the business earns high returns on ~2.6%-of-sales reported capex.

Product mix (FACT, 20-F net sales by product, CHF M):

The story compresses to one line: On is ~93% a running-shoe company. The technology narrative — CloudTec cushioning (the hollow “cloud” midsole pods) and the Speedboard propulsion plate — anchors a premium positioning at $150–330 price points held with minimal discounting. FY25 growth was led by the Cloud/Cloudtilt “all-day” franchises, Cloudsurfer/Cloudmonster performance running, and the Roger tennis line. Apparel (+68%) and accessories (+124%) grow off tiny bases and are the explicit “toe-to-head” diversification lever.

Channel mix (FACT, CHF M):

DTC is the higher-margin, faster-growing channel, mixing up ~1 point per year (40.7% → 41.8%). DTC = owned e-commerce plus 67 owned retail stores (14 Americas, 10 Europe, 5 APAC ex-China, 38 China). Wholesale spans run-specialty, sporting goods, outdoor, luxury and street-fashion, anchored by global key accounts Dick’s Sporting Goods, JD Sports and Foot Locker — with no long-term wholesale contracts (orders are cancellable). The standard tension applies: DTC protects brand and margin, wholesale drives awareness and reach, and the two partly cannibalize.

Geography (FACT, CHF M):

This table is the operational spine of the growth debate. The Americas is the largest region (57.7%) but the slowest-growing — the first sign of maturation in On’s proven core, and the direct analog to HOKA’s flattening US. EMEA — the Swiss/European home — is re-accelerating (+35% cc), a healthier signal than a topped-out home market. APAC (+107% cc, led by China: 38 stores, WeChat/Tmall/JD.com) is the hyper-growth engine off a low 17% base. Unlike HOKA (whose growth is now entirely international against a flat US), On is still growing double-digit everywhere — a materially better-diversified growth base.

Business model / recurring nature. On makes money by creating brand demand and selling physical product at premium prices. Revenue is discretionary and non-contractual — no subscriptions, no installed base, no switching costs; “recurring” exists only as infrequent, fashion-sensitive repeat purchase that must be re-won each season. Marketing (12.5% of sales) funds a roster of athletes and cultural figures — Roger Federer (investor-collaborator and halo, the “On the Roger” tennis line), Zendaya, FKA Twigs — the demand-creation flywheel that substitutes for any structural moat.

Verdict. A clean, transparent, asset-light, single-brand premium-footwear house with best-in-class gross margin and an unusually well-diversified growth base by geography and channel — but ~93% concentrated in one product category under one brand. The model is simple to understand; the durability question (§4) is the whole game, because there is no second brand, no contractual revenue, and no switching cost to cushion a fashion stumble.

3. Industry Dynamics

On competes in the global athletic/performance footwear and sportswear market — footwear alone a ~$150B+ category growing mid-to-high-single-digit, inside a broader sportswear/athleisure market exceeding ~$400B. The demand side is structurally healthy: secular tailwinds from the running/wellness boom, casualization of dress (“athleisure”), premiumization ($150–330 performance shoes), and more pairs per capita across more occasions. On has ridden exactly this wave.

The supply side earns the industry a mediocre structural grade — the crux of whether On is a compounder or a fashion-cycle trap. Three structural features:

1. Barriers to entry are low and falling. Design, contract manufacturing in Vietnam/Indonesia, and digital-plus-wholesale distribution are all rentable. Nike’s own filings concede “a reduction in barriers to starting new footwear and apparel companies,” and On itself is the proof — a 2010 startup that reached ~CHF 3B and a 62.8% gross margin in fifteen years. HOKA, Birkenstock’s relisting, and Alo are the same phenomenon. What On did to Nike, a well-funded challenger can attempt against On.
2. Switching costs are zero and the customer is promiscuous. The next running-shoe purchase is a free choice re-made every 6–12 months. Wholesale shelf space (Dick’s, Foot Locker, JD Sports, REI, run-specialty) is finite and contested, and the channel periodically destocks, transmitting volatility to brands. On’s own 20-F flags that partners “may decide to emphasize products from our competitors” or “redeploy retail floor space.”
3. The competitive set is deep-pocketed and re-arming. On names Nike, adidas, Under Armour, Brooks, HOKA (Deckers), ASICS, New Balance, lululemon, Alo, Patagonia, Arc’teryx, Anta and Li-Ning. Nike and adidas are far larger and have re-prioritized running (Nike’s “Sport Offense,” adidas’s resurgent Samba/Gazelle and running reset); specialty runners (ASICS, Brooks, New Balance, Salomon) are all adding max-cushion/super-shoe capacity.

Marathon capital-cycle read (INTERPRETATION): performance running is in a late-boom, capital-influx phase. The supernormal returns and growth earned by On and HOKA have pulled capital and capacity into the category — On IPO’d and is reinvesting aggressively (automated warehouses, retail rollout, LightSpray), while Nike, adidas, New Balance and ASICS all add running capacity. The capital-returns framework predicts abnormally high returns attract supply that competes them back toward the cost of capital over time. This does not mean On’s economics collapse next year; it argues strongly against capitalizing a 62.8% gross margin and ~30% growth into perpetuity — the very mistake the market made pricing On (and DECK) at the 2024–25 highs.

The fashion/fad-cycle risk is the defining sector hazard. Performance-footwear brands built on a hero silhouette have a documented history of boom-and-fade: Under Armour (peaked ~2016), Skechers (repeated cycles; taken private 2025), Crocs (boom, near-death, revival), Vans (VF’s roll-over), and — most instructive — UGG and now HOKA within Deckers. HOKA’s deceleration from +58% → +28% → +24% → +16% with a flat US is the live cautionary tale for On, whose aesthetic is similarly concentrated in a recognizable running silhouette (the CloudTec sole) that could rotate out of favor. On’s own risk factors concede the brand’s appeal partly “stems from the relative novelty and scarcity” that wider distribution erodes.

Regulatory / tariff landscape (FACT, live 2025–26 issue). Because On sources predominantly from Vietnam (~90% of footwear) and Indonesia (~10%), it carries direct exposure to US tariffs on Southeast-Asian footwear. In August 2025 the US imposed a fixed 20% reciprocal tariff on Vietnam atop existing duties; in February 2026 the US Supreme Court ruled the IEEPA-based tariffs invalid, adding uncertainty and raising the prospect of refunds. This is a real, active headwind to gross margin and a live FY26 variable — though (see §6/§8) On has so far absorbed it via premium pricing and supply-chain efficiency.

Verdict: a structurally mediocre industry with a genuine fashion-cycle trap embedded. Demand growth is healthy and durable, but low entry barriers, zero switching costs, contested shelf space, periodic destocking, a capital cycle flooding the highest-return running sub-category, tariff exposure, and a documented history of hero-brand boom-and-fade mean the industry confers no durable protection. A great operator can earn great returns here for a long time — On is doing exactly that — but it must manufacture and continuously defend its own brand moat; the sector will not do it for them.

4. Competitive Position

The moat is a brand intangible — and only that. In Greenwald’s taxonomy, On has no scale-economy advantage (at ~CHF 3B it is a fraction of Nike’s ~$50B or adidas’s ~€24B, buying media, capacity and shelf leverage from relative weakness), no switching costs, and no network effects. What it has is a genuinely strong, fast-built brand — “On,” CloudTec, “Running on Clouds,” the Federer halo — plus a portfolio of design/technical patents and trademarks (CloudTec, Speedboard, Helion, LightSpray, and dozens of Cloud-prefixed model marks). Brand is a real moat, but it is Greenwald’s weakest and least durable form: it must be re-earned every season and erodes fast when a brand becomes over-distributed, “basic,” or unfashionable.

The brand moat is real today — and the financials are its fingerprint. A moat claim only counts if it ties to a financial outcome that would deteriorate without it. On’s does, emphatically:

• Gross margin of 62.8% (FY25), 63.9% in Q4-25, 64.2% in Q1-26 — the best in the peer set, above Deckers (~57.7%), Lululemon (~57–59%), adidas (~50%) and far above Nike (~44%). Direct evidence of pricing power: consumers pay $150–330 largely at full price.
• DTC at 41.8% of sales and mixing up, giving control of price, data and brand presentation.
• Premium price points held with disciplined, minimal markdown — inventory allowances actually fell in FY25.
• ~30% constant-currency growth taking share from Nike/adidas across every region.

If the brand were not real, a 15-year-old outsourced shoemaker could not command a 62.8% gross margin. That number is the moat, quantified — and it is the first line item that would deteriorate if the brand cooled.

But the durability is genuinely contestable — pressure-testing each pillar:

• The brand is <15 years old and unproven across a full fashion cycle. UGG (decades) and even HOKA (~15 years of credibility) have longer track records; On has never been tested through a downturn in its own popularity. Its equity is concentrated in one instantly-recognizable design language (the CloudTec sole) — precisely the hero-silhouette concentration that saturates.
• HOKA’s 2025 deceleration is the explicit cautionary template. HOKA — the most structurally similar comp (premium, performance-running, DTC-plus-wholesale, hero franchises) — went +58% → +16% and a flat US in a single cycle while still an excellent business. On’s Americas already decelerated to +17.6% (reported) — the first, faint version of the same pattern.
• Zero switching costs; the incumbents can re-compete. Nike and adidas have vastly larger R&D, media and distribution budgets and are re-prioritizing running. On’s defense is brand heat and product cadence, not structural protection.
• Single-brand, ~93%-footwear concentration. Unlike Deckers’ two-brand structure, On has no second brand and no meaningful non-footwear leg (apparel 5.6%). A stumble in the On running franchise hits the entire company with no cushion.

Head-to-head (FACT/INTERPRETATION):

• vs. HOKA/Deckers: On grows faster (~30% cc vs HOKA +16%) at a higher gross margin (62.8% vs ~57.7%), and is earlier in international penetration — arguably the stronger franchise at this moment, but the less-proven one. (On the corrected share count, the valuation gap is much narrower than cross-reads suggest — see §10.)
• vs. Nike/adidas: On has no scale advantage and has been a share-taker during Nike’s multi-year stumble; the durability test arrives as Nike’s running reset and adidas’s franchise resurgence hit shelves in 2026–27.
• vs. New Balance / lifestyle “retro” cycle: New Balance and adidas Samba/Gazelle show how fast the lifestyle-running aesthetic rotates between brands — a tailwind for On now, a reminder that fashion favor is lent, not owned.

Greenwald verdict — name the moat and tie it to a number. On has a brand/intangible-asset moat and nothing else — no scale, no switching costs, no network effects. It is real today (the 62.8% gross margin and disciplined full-price DTC prove it) but narrow, single-brand, fashion-cyclical, and untested through a full cycle. On the share-stability test, On has been gaining share rapidly (a live-advantage signal) — but rising share in a low-barrier, zero-switching-cost category is exactly what precedes mean-reversion once capital and competition respond (Marathon). The moat ties cleanly to a financial outcome — strip the brand heat and the 62.8% margin, full-price selling, and ~30% growth all deteriorate together — confirming it is genuine but fragile.

Verdict: a real but narrow, single-brand, contestable brand-intangible moat — a hot brand executing superbly, not yet a proven fortress. The economics validate a genuine advantage today, and On looks earlier in its cycle than HOKA (double-digit growth in every region, home-market re-acceleration, low apparel/APAC penetration), which tempers the mean-reversion worry. But On is one fashion cycle away from proving whether “On” is the next Nike or the next Under Armour — and the absence of a second brand or any switching cost means the durability is an open, evidence-dependent question, which is precisely why the valuation (§10) is the whole debate.

5. Growth History and Forward Opportunities

History — one of the fastest clean scale-ups in consumer (FACT). On compounded net sales from CHF 425.3M (FY20) to CHF 3,014.0M (FY25), a ~47% five-year CAGR, in a healthy, still-hyper-growth deceleration (FY25 +30.0% reported, +35.6% cc). Three features make this high-quality growth:

1. It is entirely organic — no acquisitions; every franc is brand-built, not bought. The opposite of “growth-by-M&A” that hides poor economics.
2. It is broad-based across geography, channel and category simultaneously (FY25 cc): Americas +23.4%, EMEA +34.7%, APAC +106.7%; Wholesale +32.6%, DTC +39.9%; Shoes +32.9%, Apparel +75.5%, Accessories +135.1%. No single-point dependency — materially healthier than Deckers, whose FY26 growth is entirely international against a flat US.
3. Constant-currency growth exceeds reported growth (strong CHF is a translation headwind, not a demand problem) and came with expanding gross margin (60.6% → 62.8%) and operating margin (9.1% → 12.5%) — the demand-pull signature, not promotion-driven volume.

Unit vs. price/mix (INTERPRETATION): the 20-F does not disclose pairs shipped, but the qualitative bridge — “new product launches, updates to existing models, and continuity of successful products,” plus door and store expansion — points to growth that is predominantly volume/door-driven with a favorable premium-mix overlay, not price-hike-led. Higher-quality than price-led growth (which exhausts faster), but it also means volume is exposed to any cooling in brand demand.

Forward drivers (the bull’s runway):

• (1) Apparel — the biggest untapped category lever. Apparel is only 5.6% of sales but grew +75.5% cc, positioned as a lower-cost customer-acquisition entry point; the stated “toe-to-head” ambition builds apparel toward ~10%+ of sales into categories where Lululemon and Nike earn strong margins. Accessories (+135% cc) is a smaller version.
• (2) APAC / China — the geographic hyper-engine. APAC is 17% of sales growing +107% cc, China running a 38-store network plus WeChat/Tmall/JD.com. Off this low base, plausibly years of high growth — the single largest geographic lever (caveat: China consumer softness and geopolitics raise volatility).
• (3) DTC + owned-retail rollout. DTC (42%, +40% cc) is the higher-margin channel and still under-penetrated; the store network (67 locations, new stores ~40% larger with ~+20% productivity) is early, and each store lifts brand presentation, data and margin. Automated warehouses (Atlanta/Belgium) support the omnichannel scale-up.
• (4) Category adjacencies — tennis (Federer/“The Roger”), training, trail/outdoor, all-day lifestyle (Cloud/Cloudtilt) — smoothing seasonality and expanding the base.
• (5) Women’s and under-penetrated EMEA markets.
• (6) Manufacturing IP — LightSpray. A robotic arm sprays a continuous filament into a one-piece upper in ~3 minutes, collapsing ~200 assembly steps to one at a 170g shoe with ~75% less CO₂; the Busan, South Korea factory (opened Feb-2026) lifted capacity ~30×, and the LightSpray Cloudmonster Hyper sold out quickly. Combined with the On Labs Zurich superfoam work (Surreal on Cloudsurfer 3, Oct-2026), this is the closest thing On has to a durable, hard-to-copy edge — funded out of a 62–64% gross margin, not dilution.

FY2026 guidance (management commentary — treat as hypothesis, not evidence). With FY25 results (6-K, 2026-03-03) and reiterated with Q1-26 (6-K, 2026-05-12), management guided continued hyper-growth: constant-currency net-sales growth in the mid-to-high-20s% (Q1-26 printed +26.4% cc; +14.5% reported, the ~12-point gap being pure FX translation), a record gross margin guided to ~64%+ (Q1-26 actual 64.2%), and an adjusted-EBITDA margin around 18% (Q1-26 ran ~21%). Management is deliberately guiding H2 wholesale slower to keep partner inventory clean ahead of a 2027 innovation wave. The load-bearing read: On is guiding to roughly double Deckers’ growth while holding a record margin — the market is not being asked to pay for a decelerating brand yet.

The bear’s read of the same facts (INTERPRETATION): the ~47% five-year CAGR has already stepped to ~30% (FY25) and a guided high-20s% (FY26); the Americas — the proven core — decelerated to +17.6% reported, the first crack; the ~93%-footwear concentration means apparel/APAC must deliver; and a single-brand, zero-switching-cost, fashion-cyclical business is priced for the deceleration not to accelerate — the exact setup that punished HOKA when growth halved.

Verdict: genuinely high-quality growth — organic, broad-based across every geography/channel/category, margin-accretive, and still hyper (high-20s%) — but decelerating, with the proven US core showing the first sign of maturation. The quality of On’s growth is clearly higher than a typical hot-brand’s (no M&A, no single-point dependency, home-market re-acceleration, real apparel/APAC levers off low bases), arguing On is earlier in its cycle than HOKA and not yet a decelerating fad. But the rate is unambiguously coming down, the single-brand concentration is unresolved, and the valuation embeds much of the runway — so the growth is high-quality today with the durability question deferred, not answered.

6. Financial Quality

On’s reported income statement is one of the most misleading in the sportswear cohort — not because the business is weak, but because a large, non-cash foreign-exchange line drowns out a genuinely excellent operating engine. This section separates the two.

Revenue: fast, broad-based, structurally high-quality (FACT). Net sales compounded to CHF 3,014.0M (FY25, +30.0% reported, +35.6% cc) — reported CHF growth ~560bps understated by CHF strength. Growth is not concentrated (see §2/§5 tables): DTC (the higher-margin channel) is growing faster and lifting mix; APAC is near-doubling; Americas dependence is falling (63.9%→57.7%). The company is de-risking geographically as it scales.

Gross-margin structure: best-in-class and still rising (FACT). Gross margin ran 59.6% → 60.6% → 62.8%; gross profit CHF 1,893.6M (+34.7%, faster than sales). Management attributes the FY25 step to freight/air-freight normalization, favorable FX, DTC mix, and pricing.

On earns a ~300–400bp premium over even Lululemon and Deckers, and ~1,900bp over Nike — direct evidence of pricing power and a premium, largely full-price, DTC-weighted model. (Note: cost of sales is majority USD-sourced finished goods; the FY25 CHF/USD tailwind that flattered the gross line is the mirror image of the FX loss below the line — a point most gross-margin extrapolations miss.)

Operating leverage: the clean earnings story (FACT). Operating result: CHF 180.2M (FY23, 10.1%) → CHF 211.6M (FY24, 9.1% — a dip on an SBC spike and brand/retail investment) → CHF 377.0M (FY25, 12.5%, +78%). SG&A ex-SBC held roughly flat as a percentage of sales (48.4%→48.2%) while absorbing aggressive retail, marketing and headcount expansion — so the 340bp op-margin gain came from gross margin plus an SBC step-down. Real, sustainable operating leverage, not a one-year cut.

The quality-of-earnings issue: GAAP net income is FX noise (FACT → INTERPRETATION). Net income fell 15.9% (CHF 242.3M → 203.7M) in a year operating income rose 78%. The reconciliation:

The entire decline is a CHF 240.9M unfavorable swing in the FX line as the franc strengthened and On re-measured USD-denominated monetary balances; CHF 116.0M of the FY25 FX loss was unrealized (a non-cash balance-sheet mark). Because On reports in CHF but earns and holds cash largely in USD, this line oscillates violently every year (FY23 −111.4, FY24 +67.7, FY25 −173.2) with no bearing on the operating business. Effective book tax was a lumpy 0.7% in FY25 (deferred-tax-asset recognition, not cash savings — cash taxes paid were CHF 64.0M), so the reported tax line is equally uninformative.

Conclusion: GAAP net income, GAAP EPS, and any P/E on On are unreliable and should be discarded. Operating income (CHF 377.0M, 12.5%) and Adjusted EBITDA (CHF 567.0M, 18.8%) are the clean lenses. Even On’s own “Adjusted net income” (CHF 266.5M) is still FX-polluted (it removes only SBC), so do not anchor on it either. A structural fix is coming: effective 2026-01-01 On changed its functional currency from CHF to USD, so these re-measurement swings now flow to other comprehensive income, not the P&L — GAAP EPS should converge toward the operating reality in FY26.

FCF quality: real but working-capital-whipsawed (FACT). Operating cash flow CHF 232.1M → 510.6M → 359.5M; capex CHF 78.6M (2.6% of sales); FCF CHF 185.0M → 445.7M → 280.9M. The FY24→FY25 decline is almost entirely working-capital reversal, not deterioration: FY24 was flattered by a payables/other-liability build (+~142M inflow); FY25 was dragged by receivables/inventory growth (−~163M outflow). Through-cycle FCF conversion normalizes to roughly 65–75% of adjusted EBITDA. Two positive quality tells: reported balance-sheet inventory was flat (CHF 419.2M → 419.8M) despite 30% sales growth (no channel-stuffing; the cash-flow inventory figure is gross-of-FX), and the inventory allowance decreased.

SBC, returns, balance sheet (FACT). SBC total CHF 66.6M = 2.2% of sales (genuinely low for a 30% grower — dilution is not a material shareholder tax here). ROIC ~18% and rising, comfortably above a ~9–10% WACC; ROE (64.7%) is distorted by the modest equity base and de-emphasized. Balance sheet: cash & equivalents CHF 1,019.9M (+ ~CHF 59M other current financial assets), no financial debt — the ~CHF 521.5M of “debt” is entirely IFRS-16 lease liabilities — plus an undrawn CHF 700M revolver. Total equity CHF 1,632.4M; current ratio 2.7×; net cash ex-leases ≈ CHF 1.02B (~$1.28B).

Share count / EPS basis (FACT — defuses the 638M trap). Class A 296.87M (full economics) + Class B 341.24M (1/10th economics — distributions by par value CHF 0.01 vs 0.10; 10 B convert to 1 A). Economic-equivalent ≈ 331M shares. On reports Basic EPS Class A of CHF 0.62 (Class B CHF 0.06) — already the correct per-economic-share figure (203.7 ÷ ~329.5 weighted economic shares). Dividing net income by the raw 638M A+B count produces a phantom CHF 0.32 and must not be used. Adjusted diluted EPS Class A is CHF 0.80.

Verdict: yes — a genuinely high-quality P&L, and economics clearly improve with scale. Best-in-cohort and still-rising gross margin (62.8%), 340bp of operating-margin expansion, ROIC ~18% > WACC, modest SBC, and a net-cash balance sheet all point one way. The only real “quality” issues are presentational, not economic: GAAP net income/EPS are dominated by non-cash FX and a distorted tax line, and FCF is whipsawed by working capital. Anchor on operating income, adjusted EBITDA (treating SBC as a real cost), and normalized FCF — on those measures, On screens as one of the highest-quality businesses in the group.

7. Capital Allocation

Financing posture (FACT). On raised ~$746M gross in its September-2021 IPO and has never taken on financial debt. Proceeds funded organic build-out — automated warehousing, retail, product, working capital — and it today sits on ~CHF 1.02B net cash with an undrawn CHF 700M revolver. It self-funds 30% growth; capital-structure risk is effectively nil.

M&A: zero, by design (FACT → INTERPRETATION). No acquisitions. Intangibles are just CHF 54.2M (mostly internal-use software) and there is essentially no goodwill. Through a Marathon lens this is a virtue: a single-brand operator earning ~18% incremental ROIC should compound internally rather than pay up into a hot consumer-M&A market. No integration risk, no impairment overhang, no capital-cycle overpayment.

Reinvestment intensity — the capex figure understates it (FACT). Reported capex is CHF 78.6M (2.6% of sales), but On’s warehouses, stores and Zurich HQ are leased, not owned, so growth spend lands in right-of-use assets and lease liabilities, not capex. ROU additions were CHF 291.2M in FY25 (the automated Atlanta warehouse plus store leases); lease liabilities are CHF 521.5M. The “asset-light” optics are real for the balance sheet, but the underlying infrastructure commitment is heavy and long-dated — worth flagging for anyone modeling On as a low-capital-intensity compounder.

R&D and marketing (FACT → INTERPRETATION). Formally reported “research, design and development” is only CHF 10.6M (0.35% of sales) — a definitional artifact; true product-innovation investment sits in design/product headcount within SG&A (the On Labs group is ~1,168 people), and marketing likewise sits in SG&A (ex-SBC SG&A 48.2% of sales). The return on that brand spend is visible in the outputs: 62.8% gross margin, +30% growth, APAC +96%.

Shareholder returns: none, and increasingly a question (FACT → INTERPRETATION). On has never paid a dividend, has no buyback, and intends to retain all earnings for the foreseeable future. At 30% growth with ~18% ROIC ≫ WACC, full retention is the right call today. But the net-cash pile (~CHF 1.02B, ~11% of market cap) earns only ~3% and, held in USD by a CHF reporter, is itself a source of the FX drag that mangles reported earnings. With growth mathematically certain to decelerate and no capital-return framework in place, “what happens to the cash” is a legitimate — if not-yet-urgent — allocation and governance watch-item.

Compensation and incentive alignment (FACT → INTERPRETATION). FY25 aggregate executive compensation was CHF 19.3M (CHF 12.7M share-based, only CHF 1.7M base+bonus) — heavily equity-weighted, aligning management with the share price. The LTIP grants time-vested RSUs and performance PSUs whose multiplier is tied to relative total shareholder return vs. a broad market index; a March-2025 amendment lengthens the PSU horizon (100% three-year cycle from 2026). The gap: there is no returns-on-capital (ROIC/ROCE) or margin metric in the LTI — a mild negative by the Greenwald/capital-returns standard, since relative-TSR can be met by multiple re-rating rather than operating performance. Offsetting it, founder alignment is strong (13.8% of economics, wealth overwhelmingly the stock), and co-founder Olivier Bernhard was an open-market buyer in May 2026 (see §8).

Verdict: broadly intelligent and disciplined capital allocation. Management reinvests at high incremental returns, has avoided value-destructive M&A entirely, keeps dilution modest, and runs a fortress balance sheet. Two watch-items keep this short of an unqualified positive: (i) a growing, low-yielding net-cash balance with no articulated return-of-capital plan as growth decelerates, and (ii) the absence of a returns-on-capital metric in executive incentives. Neither is a red flag today; both are things to monitor as On matures.

8. Changes and Headwinds — Last Two Years

The past 24 months carried On through a full sentiment round-trip — from an early-2025 growth darling to a de-rated, tariff-and-FX-pressured name at a ~42% drawdown by mid-2026 — while the underlying business accelerated the entire time. Six developments matter.

1. Leadership: the founder re-consolidation (2025–2026). The most consequential governance change is the CEO transition. Martin Hoffmann, who ran the company as combined CEO & CFO for five years, stepped down (final earnings call Q1-26, reported 2026-05-12; stays as an adviser into 2027). Co-founders David Allemann and Caspar Coppetti assumed co-CEO roles, Frank Sluis became CFO on 2026-05-01 (On’s first CFO hired from outside the founding orbit, with large-cap global-consumer experience), and Scott Maguire expanded to President & COO. This follows the earlier, cleaner exit of ex-co-CEO Marc Maurer, who left the founders’ shareholders’ agreement after the 2025 AGM. Management frames it as continuity. Interpretation: the substance is a re-concentration of operating control into the founder team just as the business crosses CHF 3B — a mild execution/key-person watch item, offset by the genuine outside CFO hire. The risk is that a dual-CEO structure, a brand-new CFO, and a departed long-tenured operator all land in the same six months.

2. The strong-Swiss-franc headwind — the single biggest distortion. A sharply stronger franc in 2025 turned a CHF 67.7M FX gain (FY24) into a CHF 173.2M FX loss (FY25) (CHF 116.0M unrealized), dragging GAAP net income down even as operating income near-doubled (the §6 QoE trap). It also opens a wide reported-vs-constant-currency wedge: Q1-26 grew +26.4% cc but only +14.5% reported; the Americas +17.1% cc but just +3.1% reported. Interpretation: FX is optical, not operational — it flatters nothing and understates the true trajectory — but it clearly weighed on the print and, plausibly, the multiple. The FY26 functional-currency change to USD removes it prospectively.

3. US tariffs on Vietnam/Indonesia footwear. Sourcing is concentrated: ~90% footwear from Vietnam, ~10% Indonesia; zero footwear from China. In August 2025 the US imposed a 20% reciprocal tariff on Vietnam; in February 2026 the Supreme Court invalidated the IEEPA-based tariffs, adding uncertainty and refund optionality. The striking part is On’s absorption: FY26 gross-margin guidance was raised to a record despite the tariff drag, with tariffs only ~−100bps in the Q1-26 gross-margin bridge — swamped by ~250bps of supply-chain efficiencies, ~150bps of full-price mix, and ~100bps FX. Interpretation: premium pricing power plus a best-in-class gross margin gives On unusual room to eat trade friction that would gut a mass-market peer — a genuine relative strength, though the sourcing concentration itself remains a real tail risk.

4. Growth deceleration — real, but off a torrid base. Constant-currency growth stepped from +35.6% (FY25) to a guided high-20s% (FY26), with Q1-26 at +26.4% cc, and management explicitly building H2 “cushioning” and cleaner wholesale inventory ahead of a 2027 innovation wave. No channel or region is rolling over (APAC +61% cc in Q1, EMEA in its sixth straight quarter >+25% cc, apparel +57.5% cc, wholesale still in only ~50% of doors at key accounts). Interpretation: deceleration-by-arithmetic, not demand fatigue — but a high-beta name guiding growth lower is more exposed to any genuine miss.

5. The reinvestment — apparel, LightSpray, On Labs, stores. Apparel grew +75.5% cc and now exceeds 10% of DTC; LightSpray (robotic sprayed-upper) scaled ~30× on the Feb-2026 Busan factory; On Labs Zurich (400+ R&D staff) delivers the Surreal superfoam on Cloudsurfer 3 (Oct-2026); On added 18 net stores (67 total, new stores ~40% larger, ~+20% productivity) and set an Investor Day for 2026-09-21/22. Interpretation: credible offense — vertical manufacturing IP (LightSpray) is the closest thing On has to a durable, hard-to-copy edge, funded out of the gross margin rather than dilution.

6. The price action itself — see the Five-Year Event Map above.

Verdict — net thesis impact: mildly STRENGTHENING, obscured by optics. Strip away FX (non-operational) and tariffs (largely absorbed), and the two-year record is a business that crossed CHF 3B, expanded gross margin to a record 62.8% and adjusted EBITDA to 18.8%, built a differentiated manufacturing capability, and diversified across region/category/channel — while the multiple halved. The offsetting negatives are real but second-order: a founder-reconsolidated C-suite with an untested configuration, a decelerating (if still high-20s%) top line, and acute Vietnam sourcing concentration. On balance these developments strengthen the operating thesis and widen the gap between fundamentals and price.

9. Risk Analysis

The dominant risk here is not solvency or execution — On is net cash with best-in-class margins — it is the durability of brand heat and the valuation that heat commands. On is a hot premium brand trading at a premium (if de-rated) multiple; the two can compress together. HOKA (Deckers) — On’s closest factor and product comp — is the live cautionary precedent.

Catastrophic-loss / permanent-impairment assessment. The probability of a total or catastrophic permanent loss is low. On carries no financial debt (all ~CHF 522M “debt” is IFRS-16 leases), holds >CHF 1.02B cash (net cash ~$1.28B), a ~2.7× current ratio, positive/growing FCF, and a 62–64% gross margin — no solvency, liquidity, or dilution-spiral path to zero. The realistic severe downside is not bankruptcy but de-rating: if brand heat cools toward a HOKA-style deceleration (risk #1) while the multiple normalizes toward a mid-teens grower (risk #3), the equity could retrace toward — or below — its 52-week low (~$32) and stay there (a 30–50% drawdown from richer entry points) without the business ever breaking. The tail that would genuinely impair intrinsic value is a structural brand collapse — On revealed as a cyclical fashion moment rather than a durable performance franchise — compounded by footwear/silhouette concentration. Low-probability but high-severity; it is the scenario the §14 bear case must falsify.

10. Valuation — Embedded Expectations

USD conversions at 1 CHF ≈ US$1.25 throughout. TTM figures to Q1-2026 unless noted. No price target; no BUY/SELL — those live only in Claude’s Take.

Resolve the capital structure first — it is the single largest source of error in valuing On. On has 296.87M Class A shares (full economics, 1 vote) and 341.24M Class B “voting rights shares” (founder-held, 10 votes on a capital-invested basis, but only one-tenth the economic claim — distributions accrue by par value, CHF 0.01 vs 0.10; 10 B convert to 1 A). The economic-equivalent share count is therefore ~331M, not the ~638M raw A+B total. Anyone multiplying price by ~638M gets a phantom ~$23.5B market cap and ~$22.8B EV — exactly double the truth. (This error propagates: on-disk peer reports quote ONON at ~5.4× EV/Sales / ~30× EV/EBITDA — both roughly double the correct figures. On is not the wildly expensive outlier those cross-reads imply.)

Normalizing the earnings (why GAAP P/E is a red herring). FY25 GAAP net income of CHF 203.7M fell year-on-year even as operating income near-doubled, because a CHF ~181M non-operating loss — almost entirely FX re-measurement — sat below the operating line, and the tax rate was a lumpy 0.7%. Neither is a real economic cost of the operating business. Normalizing: operating income CHF 377M, taxed at a mid-cycle 19% = CHF 305M; on 331M economic shares that is CHF 0.92 ≈ $1.15 of normalized EPS, for a normalized P/E of ~32× ($36.83 / $1.15). On has effectively zero net financial cost to add back (net cash roughly offsets lease interest). The GAAP 48× is an artifact; the operating business trades at ~32×.

Own-history context: the cheapest On has ever been. On AZI’s own-history valuation index (~4.5-year post-IPO range), On sits at the 2.9th percentile composite, 0.9th P/E, 1.8th P/S, 5.9th P/B — cheaper than at essentially any point since the IPO. It traded at ~7–13× EV/Sales in 2021–2024; today ~2.8× — a ~4–5× compression while the business improved (FY25 revenue +30%, operating income +78%, record gross margin). The collapse is a multiple event, not an earnings event — the Deckers template (fallen growth darling stripped of its momentum tag), not the Nike (earnings trough) or Lululemon (earnings decline) one. That distinction is the whole valuation debate.

Peer comparison — the fastest grower and best gross margin, no longer the richest multiple. ONON corrected; DECK/NKE/LULU live from public peer filings; adidas/Birkenstock/Amer Sports directional (ASSUMPTION).

*NKE EV/EBITDA on trough earnings; not meaningful.

On is the fastest-growing name in the cohort and carries the highest gross margin of any of them — yet on the corrected EV/Sales it trades at ~2.8×, barely above Deckers’ 2.4× and below Birkenstock and Amer Sports. The quantified PEG-like tension: On’s normalized P/E ~32× against ~30% growth = PEG ~1.1, versus DECK’s 14.7× against ~10% = PEG ~1.5. On current growth, On is arguably the better-priced growth in footwear — the multiple is not demanding relative to the compounding. The catch is entirely the durability of the “g”: if On’s 30% fades toward Deckers’ high-single-digit, 32× normalized is expensive in absolute terms and the PEG inverts. On is expensive only if it is HOKA; it is cheap if it is early-innings On.

Reverse-DCF / scenario — what the $10.8B EV embeds. Anchoring on TTM sales ~$3.9B and EV ~$10.8B, compounding five years to a terminal EBIT × exit multiple:

What the current price embeds (the key output). Solving for the growth that makes today’s EV a fair ~10% annual return (appropriate for a beta-1.46 equity) at a 15× terminal EBIT multiple and 18% terminal margin: the implied five-year revenue CAGR is only ~10–11% — roughly one-third of the current ~30% run-rate. In other words, at ~2.8× sales the market has already priced a deceleration to low-double-digits and margins reaching only the low end of management’s 18–20% adjusted-EBITDA ambition. A HOKA-style fade is substantially pre-paid.

• What the market is underwriting correctly (INTERPRETATION): hot-brand footwear growth does eventually decelerate; tariffs and FX are live margin threats; a ~7–13× sales multiple for a fashion-exposed brand was never sustainable. The re-rate to 2.8× is a rational reset of an IPO-era bubble multiple.
• What the market may be underwriting incorrectly: the embedded ~10–11% CAGR implies the fade is nearly immediate and steep, yet Q1-26 growth remained ~26% cc with the group’s highest gross margin, APAC compounding triple-digits, and apparel/China/DTC penetration still low. The price prices a fade, not the extinction of the growth premium — but the fade it prices is aggressive for a brand showing no fundamental deceleration yet. If On merely delivers the base case, the current EV is materially too low; the burden of proof has quietly shifted onto the bear.

Valuation verdict: the cheapest-ever ~2.8× sales multiple is pricing a fashion-fade, and — for a still-~30%-growing, 62.8%-gross-margin, net-cash brand with a de-noising FX catalyst arriving in FY26 — that fade is priced early and harshly. Whether it is too harsh is not a valuation question; it is the durability question the memo has been circling. (No price target; no recommendation.)

11. Variant Perception

Consensus is bifurcated — analysts bullish, the tape bearish. Sell-side is constructively positioned but nervous (J.P. Morgan reinstated Overweight, $51 PT, 2026-07-02; the footwear complex flagged as “regaining stride” in June-2026), yet every bull note hedges on Vietnam tariffs, the strengthening franc, and the “is this a fashion peak?” question. The stock’s action tells a more bearish story than the ratings: −42% from the January-2025 ATH, 2.9th percentile of its own valuation history. That gap between the ratings and the tape is itself the setup for a variant view.

Strongest bull case — a durable premium brand caught early in its penetration curve, de-rated to the cheapest multiple of its public life. FY25 net sales +30% (all organic) to CHF 3.0B on the best gross margin in the group (62.8%), operating margin 9.1%→12.5%, ~$1.3B net cash, zero funded debt. The runway is genuinely early: apparel +75% and accessories +135% cc off tiny bases; APAC +107% cc at only 17% of sales; China running its own 38-store network; DTC only ~42%. These are the unit economics and penetration profile of a brand two-thirds into its S-curve, not at the top. And the multiple collapsed ~4–5× while the business accelerated. On the corrected share count, On is the fastest-growing, highest-margin, net-cash name in footwear at a PEG (~1.1) below Deckers’ — a compounder on sale, the DECK-style “fallen darling repriced by multiple, not earnings,” with the FY26 functional-currency change removing the FX optics that manufactured the fear.

Strongest bear case — a hot brand at a fashion-cycle peak, structurally margin-squeezed, and not absolutely cheap. The template is HOKA/Under Armour/Skechers: premium running brands run hot, saturate, then fade — and Deckers’ own HOKA just decelerated +58%→+16%, a live warning. A footwear brand is Greenwald’s weakest, “rented” moat (near-zero switching costs), and On’s 30% could reverse as fast as it appeared. The tariff/FX vise threatens the very 62.8% margin the bull case rests on. Governance is misaligned — founders hold 56% of the vote on 14% of the economics, so minorities bear the fashion risk without control. And on absolutes, ~32× normalized earnings / ~15.6× EV/EBITDA is not a value stock — “cheapest-ever” is a statement about On’s own bubble history, not cheapness in any traditional sense. If growth fades to high-single-digits, the Bear reverse-DCF (10% CAGR) implies dead money to a small loss even from here.

The factor-positioning read (FactorsToday) — abandoned momentum name, not (yet) a value trap. On screens as a high-beta discretionary / growth-momentum proxy: beta 1.46, negative alpha, R² ~0.33 (loadings dominated by broad Market), max drawdown −49.9%. Risk-adjusted record is negative near-term (y1 −31%, m6 −37% annualized) with a sharp recent reversal (m3 +55% annualized ≈ +11–12% raw for the quarter — the bounce off the March-2026 low, not a 55% quarterly gain); longer-horizon y3 ~+4.8%/yr. Its factor twin is Deckers (0.86) — the same fallen-footwear-darling signature; the remainder is a high-beta-discretionary basket (Wayfair, Wynn, growth ETFs). The read: On is an abandoned momentum name being repriced, not (yet) an earnings-broken value trap — the decisive contrast with Lululemon (earnings declining) and Nike (earnings trough): On’s earnings are still accelerating while its multiple collapsed. That is the DECK pattern, and it means the market is pricing a fade the fundamentals do not yet show — the offside the bull must exploit and the bear must confirm. What the factor read confirms rather than resolves: beta 1.46 makes On a leveraged bet on a risk-on discretionary regime — in a drawdown it falls harder than the group, fade or no fade.

The 3–5 assumptions that matter most — and what falsifies each side:

Net variant perception (INTERPRETATION). Consensus — analysts bullish, tape bearish — has converged on a single implicit forecast: On will fade like HOKA, soon. The reverse-DCF shows that forecast is ~80% embedded at $36.83. The variant view is not that On is definitely a compounder (the moat is a rented brand; the bear case is honest), but that the price already pays for the fade while the fundamentals still show none, making the disproof of the bear (another year of 25–30% growth at a held margin) the cheaper, higher-probability catalyst than the proof of the bull. The single most important thing to watch is Assumption #1 — the first quarter On’s constant-currency growth breaks below the high-teens is the quarter this becomes a value trap; until then, it is an abandoned momentum name priced for a deceleration it has not yet delivered.

12. Fact vs. Interpretation Table

13. Open Questions

1. How far is the US (Americas) in its maturation? Americas grew +17.6% reported / +23.4% cc — decelerating but not flat. Is this an early HOKA-style plateau or a mid-cycle pause? The single most important leading indicator.
2. Can apparel actually reach ~10%+ of sales at attractive margins, or does it stall at single digits (as it has for many footwear brands attempting the pivot)?
3. What is the real China ceiling and volatility given macro softness and geopolitics — is APAC +100% cc durable for years or a pull-forward?
4. What does On do with the growing net-cash pile as growth decelerates? A buyback/dividend framework would be a re-rating catalyst; continued idle USD cash prolongs the FX drag.
5. Does the new co-CEO/CFO configuration hold, or does the loss of Hoffmann’s combined operating/financial grip show up in execution?
6. Where do Vietnam tariffs actually settle (0%, 10%, 20%, refunds?), and how much pricing headroom remains after $150–330 price points?
7. Is On a PFIC in any US-holder tax year given the large cash balance — a technical but real diligence item for US holders.

14. What Must Be True

Bull case — On is an early-innings compounder, not a fad. For the bull to be right, On must sustain ~20%+ constant-currency growth for 2–3 more years while holding gross margin ≥60%, with apparel and APAC/China scaling to diversify the ~93%-footwear concentration, and no HOKA-style US plateau. At that trajectory the ~2.8× sales / ~32× normalized multiple re-rates and the base/bull reverse-DCF scenarios (+17% to +38% EV CAGR) come into play.

• Falsification test: two consecutive quarters of DTC and constant-currency growth decelerating into the high-single-digits accompanied by an inventory build — the unambiguous signature of a fashion brand rolling over. That is the quarter the bull thesis dies.

Bear case — On is a hot brand at a fashion peak, structurally squeezed, not absolutely cheap. For the bear to be right, On’s growth must fade toward Deckers-like high-single-digits within 12–18 months (the HOKA arc), tariff/FX must cap the gross margin below 60%, and ~32× normalized earnings must de-rate as the growth premium evaporates — delivering the Bear reverse-DCF outcome (dead money to a small loss even from here).

• Falsification test: sustained >20% constant-currency growth for 2+ years with continued gross-margin expansion, driving a visible re-rating off the cheapest-ever valuation percentiles. If On keeps compounding at 25–30% with a held margin through 2027, the “imminent fade” bear is simply wrong.

The elegant feature of On today is that the two falsification tests are the same data point viewed from opposite sides — the trajectory of constant-currency growth and gross margin over the next 2–4 quarters. The price has positioned for the bear’s answer; the fundamentals have so far delivered the bull’s. Whoever is right will be visible in the print, not the narrative.

15. Source Appendix

See the Source Appendix below for the full source list. Primary sources: On Holding AG Form 20-F for FY2025 (filed 2026-03-03, auditor PwC); Form 6-K interim reports (2026-03-03 FY25 results; 2026-05-12 Q1-26 results); EDGAR Form 3/4/144 filings (Apr–Jun 2026); ROIC.ai (statements, ratios, enterprise value, earnings-call transcripts Q4-25 & Q1-26); AZI price history & valuation-index percentiles; FactorsToday factor model; and on-disk peer reports (DECK 2026-06-19, NKE 2026-06-11, LULU 2026-06-06).

APPENDIX A — Standard Diligence Questionnaire — On Holding AG (NYSE: ONON)

Supplemental to the research memo. Report date 2026-07-03. Figures in CHF unless noted; USD at ~1 CHF = $1.25.

General

What thoughtful questions have other investors asked about this company? The dominant investor question is a single one: is On the next Nike or the next Under Armour? — i.e., is ~30% growth on a 62.8% gross margin an early-innings global compounder or a fashion-cycle peak with HOKA-style deceleration ahead. Adjacent questions: (a) how much of the reported earnings weakness is FX optics versus real (answer: almost all optics — see the QoE trap); (b) what is the true valuation given the dual-class share math (answer: ~331M economic shares, ~$12.2B cap, ~2.8× sales — roughly half what raw-share-count screens show); © can On absorb Vietnam tariffs without breaking the gross margin (answer so far: yes, ~−100bps in the Q1-26 bridge); (d) what happens to the growing net-cash pile; (e) is the founder super-voting structure a governance risk.

Cyclicality & Earnings Nature

Cyclical high or low? Growth is decelerating off a torrid base (47% 5-yr CAGR → ~30% FY25 → guided high-20s% FY26) — a high growth-rate but not a cyclical earnings peak in the industrial sense; margins are still expanding. GAAP net income is artificially low (FX-depressed), so reported earnings understate the run-rate. External or internal drivers? Predominantly internal (brand, product, distribution, geographic expansion), with an external athleisure/premiumization tailwind and external FX/tariff headwinds. Revenue stability? Discretionary, non-contractual, fashion-sensitive; no subscriptions or switching costs — repeat purchase must be re-won each season. Market size/outlook? ~$150B+ global performance-footwear market growing mid-to-high-single-digit, inside a ~$400B+ sportswear market; growing, global, with On under-penetrated in apparel, APAC/China, and DTC.

Business Quality & Competitive Moat

Industry more or less competitive? More — low and falling entry barriers (On is itself the proof), deep-pocketed incumbents (Nike, adidas) re-arming in running, and a capital cycle flooding the highest-return sub-category. How profitable (ROIC/ROE)? ROIC ~18% and rising, comfortably above a ~9–10% WACC; ROE 64.7% is distorted by a modest equity base (de-emphasize). Industry profitability / barriers? Structurally mediocre: many competitors, contested wholesale shelf, zero customer switching costs, periodic destocking. Easily understood? Yes — a single-brand, asset-light premium-footwear house. Undermined by low-cost labor? Manufacturing is already outsourced to Vietnam/Indonesia; the value is in brand/design/IP, not cost. Do brands matter? Decisively — the 62.8% gross margin is the brand, quantified. Nature of competition? Brand heat, product innovation cadence, athlete/cultural marketing, distribution. Switching costs? Effectively zero — the #1 structural vulnerability.

Financial Condition & Balance Sheet

Assets not on the balance sheet? The brand itself (no goodwill/acquired-intangible carrying value — internally built). Off-balance-sheet liabilities? Minimal; operating commitments are largely capitalized as IFRS-16 leases (CHF 521.5M). Accounting conservatism? Reasonable and clean (PwC-audited IFRS); inventory allowances fell (no bloat); the main complexity is FX re-measurement noise, which is transparent. CapEx-hungry? Reported capex only 2.6% of sales, but note growth infrastructure (warehouses, stores, HQ) is leased, so true reinvestment (ROU additions CHF 291.2M in FY25) is heavier than the capex line implies. Net cash ~CHF 1.02B, no financial debt, undrawn CHF 700M revolver, current ratio 2.7×.

Capital Allocation & Management

FCF generation & use? FY25 FCF CHF 280.9M (WC-depressed; through-cycle conversion ~65–75% of adj-EBITDA). Philosophy: retain everything to fund organic growth. Recent acquisitions? None — all growth is organic (a virtue at ~18% incremental ROIC). Buybacks? None. Large issuance to insiders? SBC modest at 2.2% of sales; no aggressive insider issuance. Director/management compensation? Heavily equity-weighted (FY25 exec comp CHF 19.3M, of which CHF 12.7M share-based); LTI tied to relative TSR — but no returns-on-capital metric (a mild negative). Motivations? Founder-led (13.8% economics / 56.1% votes), wealth overwhelmingly the stock; co-founder Bernhard bought ~$2.2M open-market near the 2026 lows — a genuine conviction signal.

Valuation & Market Data

ADR / MLP / K-1? No K-1. On is a Swiss AG; the NYSE-listed Class A ordinary shares trade directly (not a sponsored ADR). Note a US-holder tax wrinkle: On flags a PFIC-status question in some years given its large cash balance — US holders should diligence this. Dividend policy? None — never paid, none intended. How profitable? Best gross margin in the cohort (62.8%); operating margin 12.5% and expanding; net income GAAP-distorted by FX. Net income vs. cash from operations diverging? Yes, in both directions and mostly due to FX (net income) and working capital (OCF) — anchor on operating income and normalized FCF, not GAAP net income.

Risks & Downside

What would cause the stock to decline? A constant-currency growth deceleration into the high-single-digits (fashion fade), a gross-margin break below 60% (tariff/FX), multiple compression, a China/consumer shock, or a governance misstep. Catastrophic-loss risk? Low — net cash, no debt, high margins; no path to zero. Total-loss chance? Negligible. The realistic severe downside is a 30–50% de-rating (toward/below the ~$32 52-week low) on a HOKA-style fade, not insolvency.

Recent News & Events

Business environment changed recently? Yes — a strong-franc FX headwind (2025), Vietnam tariffs (Aug-2025 +20%; Feb-2026 SCOTUS invalidation), and a growth step-down to a guided high-20s% cc. Acquisitions? None. Accounting-policy change? Yes and material to optics — functional currency changed CHF→USD effective 2026-01-01, which removes the FX re-measurement distortion from GAAP earnings prospectively. Other recent changes? CEO transition (Hoffmann → co-CEOs Allemann & Coppetti; Frank Sluis CFO from 2026-05-01; Scott Maguire President & COO); LightSpray Busan factory (~30× capacity, Feb-2026); Investor Day set for 2026-09-21/22; JPM reinstated Overweight ($51 PT, 2026-07-02).

APPENDIX B — Source Appendix — On Holding AG (NYSE: ONON)

Report date 2026-07-03. Primary sources first. All financial figures reconciled to the FY2025 Form 20-F unless noted.

Primary — Company Filings (SEC EDGAR, CIK 0001858985)

1. On Holding AG — Form 20-F for fiscal year 2025 (filed 2026-03-03; auditor PricewaterhouseCoopers AG). Primary source for: net sales by product/channel/geography, gross/operating margins, FX reconciliation, adjusted EBITDA bridge, balance sheet, share-capital note (Class A/Class B structure, par values, voting), founder voting power (56.1%) and economic interest (13.8%), risk factors, manufacturing/sourcing (Vietnam/Indonesia concentration), tariffs, PFIC disclosure, compensation. https://www.sec.gov/Archives/edgar/data/1858985/000185898526000008/onholdingag-20251231.htm
2. Form 6-K — Q4 & Full-Year 2025 results (filed 2026-03-03; Exhibit 99.1 press release, 99.2 Compensation Report, 99.3 Annual Report). FY25 results and FY26 outlook.
3. Form 6-K — Q1 2026 results (filed 2026-05-12; Exhibit 99.1 press release, 99.2 MD&A). Q1-26 metrics: +26.4% cc net sales, gross margin 64.2%, adjusted-EBITDA ~21%; reiterated FY26 guidance; leadership transition.
4. Forms 3 / 4 / 144 (Apr–Jun 2026). Insider activity: co-founder Olivier Bernhard open-market purchase ~$2.2M (Form 4, 2026-05-15); ex-co-CEO Marc Maurer Rule 144 sales (departed executive); COO Scott Maguire routine exercise/cover; director RSU grants/withholding; Schedule 13G/A (2026-05-06).
5. Prior-year Form 20-F filings (FY2021–FY2024) for multi-year trend reconciliation.

Primary — Transcripts

6. On Holding Q4-2025 earnings call (2026-03-03) and Q1-2026 earnings call (2026-05-12), via ROIC.ai transcript tools. Management commentary on guidance, tariffs, FX, DTC/wholesale, China/APAC, apparel, LightSpray, leadership. Treated as hypothesis, validated against filings.

Secondary — Quantitative Data Providers

7. ROIC.ai — income statement, balance sheet, cash flow (multi-year), profitability/valuation ratios, enterprise value, earnings-call transcripts. Cross-checked to the 20-F.
8. AZI (azitrading.com) — daily price/OHLCV history (IPO 2021-09-15 → 2026-07-02) and own-history valuation-index percentiles (composite 2.9th; P/E 0.9th; P/S 1.8th; P/B 5.9th). News feed (JPM Overweight $51, 2026-07-02; footwear sector note, Jun-2026).
9. FactorsToday (factorstoday.com) — factor loadings (beta ~1.46, Market-dominated, R² ~0.33), leaderboard (y1 −31.4%, m6 −37% ann, m3 +54.6% ann, y3 +4.8%/yr, maxDD −49.9%), stock-info (alpha, RS), related/factor-similar stocks (DECK 0.86 the key comp).

Secondary — Peer & Industry Cross-Reads

10. Deckers Outdoor (NYSE: DECK) — HOKA public filings — the primary competitor and cautionary comp (HOKA deceleration +58%→+16%).
11. Nike (NYSE: NKE) and Lululemon (NASDAQ: LULU) public filings — cohort framing and comp multiples. (Note: some third-party screens quote ONON multiples built on the raw ~638M share count — ~2× the correct figures; this analysis uses the corrected ~331M economic-share count.)

Analytical Frameworks

12. Greenwald & Kahn, Competition Demystified — moat taxonomy (brand/intangible; no scale/switching/network), share-stability and ROIC tests.
13. Marathon / Chancellor, Capital Returns — capital-cycle read on performance-running (late-boom, capital-influx phase).

Industry framing was built from primary filings and public peer-company disclosures.