CNCK — Q1 2026 Financial Report Analysis

Revenue up 36% on paper, but the business underneath barely grew and the operating loss widened

Coincheck Group runs Coincheck, one of Japan's largest retail crypto trading apps, and since March 2026 also owns 3iQ, a Canadian crypto fund manager, and Aplo, an institutional crypto broker. It reports in Japanese yen under IFRS (the international accounting rules) and files with the SEC as a foreign company, using a fiscal year that ends March 31. The quarter covered here, April–June 2026, is what the company calls the first quarter of fiscal 2027. Dollar figures below use the company's own conversion rate of ¥162.61 per $1.

The headline says total revenue rose 36% to ¥114.3 billion ($703 million). That number is misleading for this kind of business. Coincheck books the full value of the crypto it sells as revenue, and the cost of buying that crypto as an expense, so almost all of the ¥114.3 billion flows straight back out. What Coincheck actually keeps is the company's own Adjusted Revenue measure (revenue after subtracting what the crypto cost, a non-IFRS figure). That rose 19% to ¥2,920 million ($18 million), and most of that gain came from the newly acquired fund manager. Spending rose faster, so the operating loss (the loss from running the business, before interest, currency moves and tax) widened from ¥870 million to ¥1,373 million. The net loss shrank only because of currency gains, a smaller accounting charge on warrants and a tax benefit.

At a glance

• ¥2,920 million Adjusted Revenue (+19%) — strip out the ¥404 million of fees from 3iQ, which the company didn't own a year ago, and the rest grew about 3%.
• ¥1,373 million operating loss (vs ¥870 million) — overheads rose ¥722 million (+20%) as the company added staff and the costs of two acquired businesses.
• 12,210 million yen of Marketplace trading in September 2026, down from 20,709 million in June. Trading in July–September fell about 30% from April–June, so next quarter's core trading income is likely to come in lower.

The numbers

A note on the margin row: because revenue includes the full price of crypto sold, the -1.2% operating margin looks small. Measured against Adjusted Revenue (the money Coincheck actually keeps), the operating loss equals about 47% of revenue, worse than about 36% a year earlier.

Where the money comes from

Coincheck earns most of its money on its Marketplace, where it acts as the dealer: it sells crypto to customers and buys it back from them, charging a price spread (the gap between its buying and selling price). That spread averaged 3.26% of each trade this quarter, slightly below 3.31% a year ago. Its separate Exchange, where customers trade with each other, is generally free, so its volume (down 53% year on year to about ¥498 billion) matters less for revenue.

The core trading business shrank. Marketplace volume slipped 4% to ¥59.1 billion and the spread narrowed, so the money kept from trading fell 9%. Volume follows crypto prices and how much they swing, and the filing says customer assets fell 37% "due primarily to the decline in the market price of certain crypto assets, including Bitcoin, ETH, and XRP." Fewer people were trading: 69,859 customers made at least one transaction in June, down 17% from a year earlier. Staking (earning rewards by lending crypto to help run a blockchain network) was the one organic bright spot, up ¥176 million after customer payouts.

Aplo, the institutional broker bought in October 2025, added ¥8,255 million of gross revenue but only about ¥12 million after its cost of sales. It makes the top line look bigger without adding real income yet.

What the headline numbers hide

• The 36% revenue jump is mostly bookkeeping. Revenue from customers rose only 7% (¥48,033 million vs ¥44,860 million). Revenue from "cover counterparties" rose 47% to ¥56,743 million. These are the trades Coincheck makes with other firms to offset its customers' positions. The filing explains that when one crypto is settled with another, IFRS counts the settlement as a sale, which "results in higher reported gross revenue relative to the underlying customer transaction volume." Aplo's ¥8,255 million of pass-through revenue adds to this. Nearly all of it is cancelled out by cost of sales, which rose 37%.
• Without acquisitions, revenue was roughly flat. Take out 3iQ's ¥404 million of fees and Adjusted Revenue grew about 3% (¥2,516 million vs ¥2,445 million). Compared with the previous quarter (January–March 2026), Adjusted Revenue was flat at ¥2,920 million vs ¥2,907 million. But 3iQ contributed only one month then (¥139–140 million) and a full quarter now, so the rest of the business fell about 9%.
• Costs grew faster than the reported 20%. A year ago, ¥255 million of staking rewards paid to customers sat inside overhead costs. This year the equivalent ¥277 million is counted in cost of sales. On a like-for-like basis, overhead rose from about ¥3,316 million to ¥4,293 million, roughly 29%. Management attributes it to "increased headcount and business expansion, including the addition of employees and operations from the 3iQ and Aplo acquisitions."
• The smaller net loss came from items outside the business. Below the operating line, results swung about ¥546 million in Coincheck's favour. The main pieces were a ¥267 million foreign-exchange gain against a ¥131 million FX loss a year ago, and a ¥50 million warrant charge against ¥223 million. Warrants are rights to buy shares, left over from the 2024 SPAC merger, and their accounting value moves with the share price. Tax went from a ¥126 million expense to a ¥33 million benefit, a swing of ¥159 million, because newly acquired loss-making subsidiaries were added to the group's tax position. Together these improvements of about ¥700 million more than offset the ¥503 million worse operating result. Neither is a repeatable source of earnings.
• Most of the loss-per-share improvement came from issuing more shares. The average share count rose 30% to 169.8 million. With the old share count, this quarter's loss would have been about ¥8.99 per share rather than ¥6.92. So more than half of the ¥3.60 improvement comes from spreading the loss across more shares.
• Cash flow was negative and the cash balance rose only because of new shares. Operating activities used ¥1,143 million of cash, against ¥845 million generated a year earlier, roughly matching the ¥1,176 million net loss. Cash rose from ¥9,458 million to ¥16,081 million ($99 million) because KDDI, a Japanese telecom company, bought a 14.9% stake for ¥9,999 million net of costs (about $65 million). Part of the money went to repay ¥2,008 million borrowed from Monex, a related party. Equity rose from ¥21.7 billion to ¥31.1 billion.
• The account growth changed meaning in June. Verified accounts (fully opened, identity-checked accounts) rose by about 8,000–11,000 a month in April and May, then by 77,711 in June. From June the count includes accounts opened through "crypto-as-a-service" partners: other firms that offer Coincheck's trading under their own brand. The company says these accounts have "much lower transaction fees per trade" and that the partners take a share of the fees. The 12% account growth is therefore not comparable with earlier periods.
• 3iQ's assets under management fell. Assets under management (AUM), the client money 3iQ manages and charges fees on, fell from ¥128.8 billion at March 31 to ¥105.5 billion at June 30, down 18% in three months. Because its fees are a percentage of AUM, the new fee stream depends on crypto prices just as the trading business does.

Takeaway: Leave out the gross-revenue accounting and the acquisitions, and Coincheck's own trading income fell 9% while its costs rose about 29%. The smaller net loss came from currency gains, warrant accounting and tax, not from the business, and the cash buffer rose only because KDDI bought new shares.

What to watch next

Management gave no financial guidance. The filing warns that results from its new strategy (growth through fund management, institutional broking and white-label partnerships rather than retail trading alone) "may not be reflected in our financial results for the next 12 months or longer."

The company's monthly figures for July–September 2026 point to a weaker next quarter for the core business. Marketplace volume was ¥15.5 billion, ¥13.8 billion and ¥12.2 billion, about ¥41.5 billion in total. That is about 30% below April–June and about 56% below the same months of 2025 (¥94.7 billion). Unless the spread widens, Marketplace trading income should fall again. Customer assets recovered to ¥876.9 billion by the end of September as crypto prices rose. That helps staking income, and 3iQ's fees if its AUM followed. Verified accounts jumped to 2,868,544 by September, almost all through low-fee partner channels. The question for the next report is whether those accounts trade enough to matter.

Two other things to watch. Japan's parliament passed a law on July 15, 2026 that moves crypto under the Financial Instruments and Exchange Act (the law that covers securities). It is expected to take effect in 2027, and the company says it could raise minimum capital requirements as early as that year. The other is the KDDI partnership: the company has said little so far about what it will contribute in revenue.

Our view: costs are now set for a bigger, multi-business company while its main income source, retail trading in Japan, is shrinking. Without a strong rise in crypto trading, losses are likely to continue for at least the next quarter or two. The KDDI money gives it time, but the costs need new revenue to justify them.