East Asia: China banned crypto and it moved, Hong Kong licensed it and capital stayed
Chainalysis published its Eastern Asia chapter of the 2026 Geography of Cryptocurrency report on 5 October. It covers July 2025 to June 2026, a period in which the region's crypto economy contracted modestly overall. The interesting part is not the total but how differently five markets behaved under five different rulebooks.
The sizes. South Korea $449.1 billion, Japan $228.3 billion, Hong Kong $192.2 billion, China at least $176.3 billion, Taiwan $140.4 billion.
South Korea: retail, untaxed, and long AI. Korea grew 12.3%. No crypto tax applied during the period; a 22% tax on profits is scheduled for the start of 2027, but has been delayed before. By June 2026, AI-linked tokens were the largest thematic category of won trading volume — about 19.5 times the yen rate — with Worldcoin alone at $7.41 billion. Chainalysis ties it to the same appetite that made SK Hynix the centre of the Korean stock market.

Japan: institutions on paper, retail on-chain. Roughly one in four people who withdraw from Japanese exchanges later deposit into DeFi; decentralised exchanges hold nearly 35% of Japan's services activity, and DEX activity is up more than 200% since 2022 while centralised exchanges stayed flat. A top marginal rate of 55% applied through the period; reforms advanced in July 2026 came too late to show.
Hong Kong: the licensed perimeter works. Institutional platforms took 16% of service inflows, up from about 9% two years earlier and nearly three times any neighbour. Inbound business-to-business flows reached nearly $24 billion, and cumulative net B2B inflows reached $17.4 billion by mid-2026, while wallet-based flows ran net negative.
China: the ban moved the activity. With exchanges banned, Chainalysis estimates the economy mainly from peer-to-peer flows, which made up 59.1% of it — 3.5 times the prior share. Unique wallets sending stablecoins peer-to-peer grew 43-fold between Q1 2024 and Q2 2026. Self-custodied stablecoins turned over 33.2 times a year, more than three times the world average of 9.3. Chainalysis links the acceleration from March 2025 to the expansion of China's social credit system into finance, and says plainly that this is a working hypothesis.
The reading. Chainalysis's own conclusion is the clearest sentence in the report: China prohibited crypto and activity moved; Hong Kong, Japan and Korea built supervised access and pulled at least part of it inside a licensed perimeter. What the data cannot yet say — and the report admits it — is whether supervision captures retail and cross-border activity as well as it has captured institutional money.