China's self-custodied stablecoins turn over 33.2 times a year, Chainalysis finds

A turnover rate of 33.2 means the average coin rests in a Chinese self-custodied wallet for about 11 days before it moves again ^[12]. At the global average of 9.3, the same coin would sit for roughly 39 days ^[13]. Chainalysis said the pattern was consistent with users treating stablecoins as working capital ^[5].

The flow also implies a small stock. If the 33.2 figure is calculated on the same $104.1 billion of transfers ^[3], the average balance behind a year of activity is about $3.1 billion ^[17]. Spread across 18.1 million transfers, the mean ticket is about $5,750 ^[14], and a few large payments can pull a mean a long way up. The published summaries of the October 5 report ^[1] do not say how turnover was computed, how many sending wallets there are in absolute terms, or how transfer sizes are distributed.

Two of the report's figures appear to describe the same money. Domestic peer-to-peer activity is 59.1% of a crypto economy that Chainalysis puts at no less than $176 billion ^[6]. Multiplied out, that share is about $104.0 billion ^[15], within rounding of the self-custodied transfer total. A period earlier the share was about 16.9% ^[16], so direct transfers went from roughly a sixth of China's measured crypto activity to most of it.

The timing runs against the regulators. Authorities reinforced the rules in February 2026, aiming at unauthorized yuan-pegged stablecoins and tokenized real-world assets ^[7]. In March, Chainalysis recorded a $4.9 billion spike in monthly domestic transfer volume ^[9]. Self-custody means the user holds the keys and no exchange sits in the middle ^[8]. Crypto Briefing wrote that "Restrictions aimed at platforms work best when activity runs through platforms." ^[10]

Two other explanations fit the velocity data about as well as working capital does. A small group of over-the-counter dealers recycling one float would also turn coins over fast, and a 43-fold rise in sending wallets ^[2] measures growth in the count, not how concentrated the money is. Coins bought with yuan and handed quickly to someone offshore would move fast too, and that would be capital leaving the country.

I think the working-capital reading is the best fit. According to Crypto Briefing's account, the report frames the change as a lasting move by users toward direct wallet-to-wallet transfers ^[11], and an 11-day holding period ^[12] is hard to square with hoarding. The counter-case is the dealer one, and it wins if most of the $104.1 billion ^[3] turns out to run through a small number of wallets. The thesis also fails if later data shows turnover sliding toward 9.3 ^[4] while the implied balance grows, since that would mean Chinese holders have started saving in dollars. On these numbers, Chinese users are not building dollar savings; they keep about $3.1 billion ^[17] on hand at any moment to move $104.1 billion a year ^[3].