Chapter 3 of the fund's October stability report puts realized volatility in a $2.3 billion market at approximately 1.5 times higher than on traditional venues, with about 80 percent of trades sized below one share. Those traditional venues absorb between 87 and 99 percent of the overnight price move within five minutes of the opening bell.
More than 50 percent of activity in tokenized US equities happens outside regular US market hours, and about 80 percent of trades are executed in sizes below one share, according to a chapter of the International Monetary Fund's October stability report released on October 8.
Those are the two things a token offers that a share on an exchange does not, and the chapter measures what they come with. Realized volatility in the tokenized versions is "approximately 1.5 times higher" than in traditional markets, in the chapter's wording, and it describes the tokenized shares as significantly less liquid than their traditional counterparts.
Chapter 3 of the Global Financial Stability Report, titled "Scaling Tokenization: New Efficiencies, New Vulnerabilities," draws those equity findings from a sample rather than the whole market. The fund's staff looked at tokenized versions of the S&P 500, the Nasdaq, Tesla, Google and NVIDIA, issued by Ondo and xStocks, across 11 trading venues over 365 trading days. That produced more than 2 million observations at five-minute intervals covering about $345 million in reported market value, using data from Allium, CoinDesk and Bloomberg.
Access outside market hours is also where the plumbing strains. Price feeds for tokenized equities have not uniformly followed the tokens into the weekend: Aave's collateral market for Coinbase-issued stock tokens runs continuously while the Chainlink feeds pricing that collateral publish only from Sunday evening to Friday evening Eastern time, holding over weekends and market holidays. A market that trades when its reference price does not update is carrying the gap somewhere.
On price discovery the chapter is more reassuring than the volatility figure alone suggests. Traditional equities absorbed between 87 and 99 percent of the overnight change in tokenized prices within the first five minutes of the US trading session, a range the authors give for their estimates rather than a single number. The fund reads that as both markets responding to similar information, which is a weaker claim than the tokenized market leading the primary one. What it establishes is that the overnight moves are not noise the opening bell discards.
The scale involved is small. The chapter puts the tokenized real-world asset market at about $65 billion excluding repos, with a figure note of $65.10 billion as of July 31, 2026, drawn from RWA.xyz with IMF staff calculations, and tokenized equities at an estimated $2.3 billion on the same date. On those two inputs, both from the same source and the same observation day, equities are about 3.5 percent of the tokenized asset market.
Repo is where the volume actually sits, and it is measured differently. The chapter reports a 30-day average daily volume of $303 billion in tokenized repos, and its own figure note is explicit that this reflects repo activity rather than outstanding tokenized asset value. The distinction matters because the two headline numbers are not comparable: $65 billion is a stock of assets, $303 billion is a day's turnover in a separate instrument. An accompanying IMF blog post published the same day described other tokenized asset trading as adding about $65 billion daily, which does not match the chapter's own definition of that figure.
The blog, written by Gonzalo Fernandez Dionis, Caio Ferreira, Mindaugas Leika and Athanasios Vamvakidis, sets out four conditions the fund considers necessary for tokenization to scale: legal certainty that a tokenized asset represents an enforceable right, regulatory clarity on how existing rules apply to new ledgers and market functions, interoperability between platforms rather than isolated liquidity pools, and settlement in safe and widely accepted forms of money.
The first of those conditions is the one the products themselves keep running into. When Securitize opened trading in twelve tokenized US stocks on October 8, its own disclosure said holders are not registered shareholders of the underlying issuer unless they convert, and named no issuer among the twelve that had adopted the sponsored tokenization needed to make conversion available. The instrument is a security entitlement rather than a share on a register, which is precisely the enforceable-rights question the fund raises.
Regulatory clarity is moving faster in the United States than the chapter's data window captures. The Securities and Exchange Commission issued an exemptive order in September creating a venue category that is freed from exchange registration on notice rather than approval, and OKXICE told the Commission on October 4 that it intends to trade 63 tokenized US stocks under it. The fund's sample ends before any of that trading begins, so the venue count and the liquidity it measured describe an earlier market.
The authors attach their own caveat, writing that the findings should be interpreted with caution because the market is at an early stage. The rest of the October report has not arrived yet. The fund's release schedule puts the full Global Financial Stability Report and its opening chapter, "Beneath the Resilience: Rising Risks, Unresolved Vulnerabilities," on October 13.
