Gymshark v Castore: who is actually winning the activewear war?

The numbers

Two brands, two clocks

Every figure below belongs to one company or the other. Nothing on this page adds the two together, because no filing anywhere reports a Gymshark and Castore combined.

The takeaway

Five things the filings settle

Which brand is bigger, and by how much?

Gymshark, by a distance that grows the closer you look. Its filed turnover of £647m is 1.9 times Castore's £335m, and once the two are put on the same clock it is 2.9 times.

The direction of travel is the interesting part. Gymshark's turnover moved 6.6% over its year, implying £607m the year before, and its estimated profit before tax fell from £12m to £7m. That is a business converting a lot of revenue into a thin margin while it spends on something, with staff costs of £70m across 1,029 people and headcount up 16.8%.

Castore's £335m is a bigger number than it has ever filed, but it is bought at a cost. The estimated loss before tax of £27m improves marginally on the £29m before it, so the business is not yet turning its growth into profit. Its 776 people cost £36m over the period, a headcount up 50.1%.

On the balance sheet the two are nearer than the trading figures suggest. Both closed with £37m of cash, the single point where the pair are exactly level. Net assets favour Gymshark at £101m against £63m, and current assets favour Castore at £228m against £172m, a difference explained almost entirely in the debtors section below.

Why does Castore's 76% growth need a health warning?

Because six of those months are free. Castore's accounts cover 18 months to August 2025, Gymshark's cover 12 months to July 2025, so Castore's turnover figure is compared against a 12 month period that came before it.

Strip the extra half year out and the picture changes without any of it being wrong. The 75.8% movement implies about £191m in the previous 12 month period. Annualising the £335m gives roughly £223m, which is growth of about 17.2%. Still comfortably ahead of Gymshark's 6.6%, and a very different headline.

The same correction applies everywhere a flow figure appears. Staff costs of £36m become about £24m a year. Average cost per head of £47,000 becomes about £31,000 for 12 months of work, against Gymshark's £68,000. Interest payable of £13m becomes about £9m. Balance sheet figures need no adjustment at all, which is why cash, net assets, borrowings and headcount are quoted as filed throughout this page.

Where does every £100 of sales go?

Both brands spend almost exactly the same share of sales on people, around £10.80 in every £100. The difference is what happens after that: Gymshark stops just under the line, Castore goes over it.

The shape of the two columns is the whole argument. Wages are not what separates these businesses: both spend a shade under £11 of every £100 on their people. Financing does separate them, at £3.90 against £1.10, and Castore's is the more expensive debt relative to what it sells even though it borrows less in absolute terms.

Everything else, the residual that holds stock, stores, marketing and sponsorship, is £87.00 at Gymshark and £93.40 at Castore. The extra six pounds forty Castore spends there, on top of its heavier financing, is what turns a thin margin into a loss, and it goes on the things a challenger brand buys when it is trying to be seen.

Why does Castore hold £73m of debtors and Gymshark £2m?

Because they sell to different people. Castore's debtor book is 37 times the size of Gymshark's, which is what happens when a large share of revenue is invoiced to clubs, teams and retail partners rather than taken at a checkout.

It shows up in the shape of the current assets too. Castore holds £228m of current assets against Gymshark's £172m, despite selling around half as much on a comparable year, and £73m of that is money owed by other businesses. Gymshark's £172m is dominated by stock and its £37m of cash, with only £2m of debtors behind it.

Neither model is better on the face of the filings, and the accounts do not say which converts more reliably. What they do say is that Castore has more of its year sitting in someone else's payment cycle, and that this is a working capital position rather than a trading one. The £61m of estimated borrowings and £13m of interest payable are the price of holding it.

Sector codes reflect the same split. Castore's company is registered under retail sale of sporting equipment in specialised stores, and it holds six leasehold property records covering stores in Manchester, Bristol, Milton Keynes and Brierley Hill. None of those records carries a disclosed price. Gymshark's filing entity is registered as a holding company and has no property records in the dataset at all.

Who controls each company, and who holds security over it?

Founders, in both cases, but recorded in different ways. Companies House lists Ben Francis in the 25% to 50% ownership band at Gymshark's topco, and the Beahon brothers as people with significant influence at Castore's parent. Neither company has an owner recorded at the 75% direct level.

The December 2025 charge sits after both companies' period ends, so it is not reflected in any figure on this page. Control records show who Companies House recognises as controlling each company, in bands. They are not stake sizes, prices or valuations, and they are not used anywhere here to combine or net off the two companies' financials.

The ledger

Every figure, as filed

Two companies, two rows in the source data, shown side by side. Losses appear in brackets. Nothing in this table is a total across the pair.

Method

How this page was built

Figures come from Companies House filings for Gymshark Topco Limited (16577203) and J.Carter Sporting Club Limited (09670915), the consolidated filing entity for each brand. Both are consolidated accounts, so subsidiary figures already sit inside them and nothing has been added on top.

Profit before tax is an estimate derived from the filings and will not reconcile exactly to a statutory figure, because a holding company carries its own costs in its individual accounts. It is labelled as estimated wherever it appears. Borrowings are likewise estimated.

Castore's accounts cover 18 months and Gymshark's cover 12. Figures are shown as filed throughout. Where an annualised figure appears it divides Castore's flow figures by 1.5 and is labelled as a Spark Intel working, not a filed number. Ratios expressed as a share of turnover, including the per £100 diagram and the pre-tax margin, are unaffected by the difference in period length. Balance sheet figures and headcount are point in time and are not adjusted.

Every number belongs to one company or is an explicit relationship between the two, such as a ratio or a gap. No combined turnover, profit, headcount or interest figure appears anywhere on this page, because no filing reports one. Amounts under £0.1m are shown in thousands, amounts from £0.1m to £2m to one decimal place, and amounts of £2m and above to the nearest million. Directors' remuneration is recorded in the source at roughly £1m granularity and is treated as a flag rather than an amount, so it is not quoted here.

Control information reflects the persons with significant control register and shows control bands, not stake sizes or prices. It is not used to combine or net off the two companies' financials.

Questions

Gymshark and Castore, answered

More from Spark Intel