On Holding tempers its 2026 outlook: what FP&A should separate

On Holding reported its second-quarter results on August 11, 2026 and replaced its prior full-year sales-growth floor with guidance for constant-currency growth in the low-20% range. Q2 net sales were CHF 850.3 million, up 13.5% as reported and 21.6% at constant currency. Reuters reported an analyst estimate of CHF 878.16 million, putting the quarter about CHF 27.9 million below consensus.

The outlook change is not a simple across-the-board cut. On lowered the sales-growth floor while giving a range whose reported-currency upper end exceeds the prior reported-currency minimum, raised its gross-profit-margin floor to at least 65.0%, and kept adjusted EBITDA margin guidance at 19.5% to 20.0%. For FP&A, the decision is therefore not to reduce every revenue line by the same percentage. The forecast needs separate bridges for guidance currency, DTC and wholesale, price and mix, and the margin effects that management has described but not quantified.

What changed in On’s 2026 outlook

On’s May 12 first-quarter outlook called for net sales growth of at least 23% year over year at constant currency. At the spot rates used that day, management said this implied reported net sales of at least CHF 3.51 billion. The company also expected gross profit margin of at least 64.5% and adjusted EBITDA margin of 19.5% to 20.0%.

The August outlook uses a less precise constant-currency formulation: growth in the low-20% range. At the spot rates used on August 11, On said this implied reported net sales of CHF 3.47 billion to CHF 3.56 billion. The gross-margin floor rose by 50 basis points to at least 65.0%, while the adjusted EBITDA margin range stayed unchanged.

Start the forecast reset with the guidance metric

Begin the operating model in constant currency, the measure used for full-year growth guidance. FP&A can then translate the channel and regional forecast at an approved exchange-rate set. Comparing CHF 3.51 billion from May with CHF 3.47 billion to CHF 3.56 billion from August without isolating FX would mix an operating change with two different spot-rate snapshots.

The Q2 result shows why that control matters. Reported sales growth was 13.5%, while constant-currency growth was 21.6%, an 8.1 percentage-point gap. That gap does not by itself identify which currency pairs, regions or transaction dates drove the difference. Keep local-currency revenue, translation rates and the reporting-currency overlay as separate model layers.

Separate DTC momentum from wholesale sell-in

DTC sales reached CHF 388.4 million in Q2, up 26.0% as reported and 34.3% at constant currency. Wholesale sales were CHF 461.9 million, up 4.8% and 12.7%, respectively. DTC share rose to 45.7% from 41.1%, but wholesale remained the larger channel. A blended growth assumption would hide both the faster DTC trajectory and the size of the wholesale base.

On said wholesale growth was moderated by disciplined sell-in in a promotional environment, particularly in the Americas, and expects DTC to strongly outperform wholesale in the second half. FP&A should not treat that statement as a quantified channel forecast. The wholesale model needs separate assumptions for partner orders, company-controlled sell-in, partner inventory and sell-through. The DTC model needs traffic, conversion, average selling price, returns, store openings and comparable-store performance. Only then can the team distinguish a management choice to protect pricing from a change in end demand.

Keep currency, geography and product mix on separate lines

Channel is only one dimension. Q2 constant-currency sales grew 13.0% in the Americas, 20.5% in EMEA and 54.7% in Asia-Pacific. By product group, constant-currency sales grew 18.9% for shoes, 56.2% for apparel and 102.2% for accessories. Apparel and accessories grew quickly from much smaller reported bases of CHF 54.2 million and CHF 14.5 million, compared with CHF 781.6 million for shoes.

Those views overlap. An Asia-Pacific apparel sale through an On store appears in the region, product and DTC cuts of the same revenue. FP&A should therefore use a controlled revenue cube or an equivalent reconciliation rather than add channel, region and product growth contributions together. The public filing does not provide a cross-tabulation of those dimensions, so any external allocation among them is UNVERIFIED.

Build the gross-margin bridge without double-counting

On’s Q2 management discussion and analysis says gross profit margin increased to 65.4% from 61.5%. Management attributed the increase mainly to freight efficiencies, higher DTC mix, premium brand positioning and favourable foreign exchange, which more than offset higher U.S. import duties. The filing does not assign basis points to any of those drivers.

A usable margin bridge should keep channel mix, price and markdown, product and regional mix, freight and sourcing efficiency, transaction-currency effects, and tariffs on separate lines. Full-price discipline should not be counted again inside DTC mix if the channel assumption already captures a different markdown rate. Foreign exchange on revenue should also remain distinct from currency effects within cost of sales. The same discipline applies to Vestas’s 2026 EBIT-margin outlook, where project execution, Service, warranty cost and working capital require separate forecast lines. Cisco’s component-cost and price-recovery bridge extends the control to inventory commitments, memory costs, hardware mix and customer pricing. Tapestry’s FY27 brand-and-margin bridge extends the control to brand, region, AUR, tariff and SG&A assumptions that overlap but should not be added as independent contributions. The dLocal Q2 payment-economics bridge applies the same non-double-counting rule to payment-flow mix, merchant pricing tiers, geography, processing spread and FX spread.

DICK’S segment guidance reset shows the same control at acquisition scale: Foot Locker, core DICK’S and corporate-and-other items need separate forecast lines before the consolidated outlook is rebased.

The higher gross-margin floor does not automatically raise the EBITDA forecast. DTC growth can improve gross margin while bringing store, ecommerce, marketing and fulfilment costs below gross profit. On kept adjusted EBITDA margin guidance unchanged, so FP&A should preserve the operating-expense and investment plan rather than treating the 50-basis-point gross-margin increase as available EBITDA.

What On has not quantified

• the endpoints of the low-20% constant-currency growth range;
• second-half DTC and wholesale growth rates;
• the share of slower wholesale growth caused by deliberate sell-in management versus partner or consumer demand;
• the basis-point contribution of channel mix, price discipline, freight, currency and tariffs to gross margin;
• the cross-tabulation of channel, geography and product performance; and
• the causal relationship, if any, between the Q2 consensus miss and the revised full-year wording.

The tariff treatment needs its own control. The outlook excludes benefits from anticipated tariff refunds in the second half. The MD&A also says On had received approximately CHF 27.9 million of IEEPA tariff refunds by August 11 and would recognise that amount in Q3. FP&A should keep the refund outside the recurring run rate and reconcile its guidance treatment before using it as a margin offset.

What FP&A should lock before the next forecast

1. Guidance register: record the source date, metric, currency basis, floor or range, and the exchange-rate set behind each reported-currency indication.
2. Channel forecast: model DTC and wholesale separately, with named owners for sell-in, sell-through, store, ecommerce, pricing and returns assumptions.
3. Dimension reconciliation: reconcile channel, geography and product views to one revenue total without adding overlapping contributions.
4. Gross-margin bridge: assign basis-point movements to price, mix, freight, FX, tariffs and refunds only when the evidence supports the allocation.
5. Scenario boundary: keep the latest expectation separate from management guidance, with clear triggers for a wholesale downside or DTC upside case.
6. Next verification: test Q3 wholesale sell-in, DTC share, regional growth, gross-margin drivers and tariff-refund recognition against the reset assumptions.

The defensible forecast reset is a driver-level reforecast, not a blended haircut. On’s disclosure supports lower confidence in the previous sales-growth floor and higher confidence in the gross-margin floor, but it does not supply the numerical bridge FP&A needs to connect those two conclusions.