More than half of trading in the tokenized equities examined by the International Monetary Fund occurred outside regular U.S. equity-market hours. Approximately 80% of transactions were smaller than one conventional share. Those are substantial signals of what users want from a tokenized stock: a way to trade when U.S. exchanges are closed and a way to buy a fraction of an expensive share.
The same study finds that these markets are considerably less liquid and roughly 1.5 times as volatile as the equivalent traditional stocks. That matters for anyone assuming that round-the-clock access necessarily improves execution quality.
The results appear in Chapter 3 of the IMF's October 2026 Global Financial Stability Report, Scaling Tokenization: New Efficiencies, New Vulnerabilities, publicly discussed ahead of the full report release. Public discussion intensified in October 11 reporting. The IMF's empirical observations deserve close reading because they answer a different question from recent exchange launches: how do tokenized stocks actually trade after they exist？
What the IMF measured
The IMF compared five liquid tokenized U.S. equity-linked instruments with corresponding conventional assets, using products from Ondo Finance and xStocks. The exercise covered 11 centralized and decentralized trading venues, 365 trading days, and more than two million five-minute observations of prices and trading volumes. The study's approximate $345 million covered sample was about 15% of the $2.3 billion tokenized-equity market identified as of July 31, 2026.
These qualifications are essential. The findings do not describe every tokenized stock, every platform, or today's entire market. They are an early empirical sample from a small and changing asset class.
The 80% statistic measures the share of trades executed in quantities below one share. It does not say that 80% of investment capital is fractional. Likewise, the more-than-50% off-hours number describes the study's transaction activity distribution, not a guarantee of sufficient liquidity at midnight.
Tokenization appears to satisfy two neglected trading demands
Fractional orders are an access feature. A user who wants $50 exposure to a $300 share can take a small position without buying a whole share. This does not guarantee low cost or appropriate shareholder rights, but it changes minimum-ticket constraints.
The off-hours data are more revealing. U.S. exchanges operate during a fixed session, while news arrives globally and continuously. Onchain venues can accommodate users in different time zones and investors reacting to earnings or geopolitical developments that occur outside those hours. The IMF's evidence suggests some of this usage is genuine rather than just the theoretical appeal of “24/7.”
The IMF also studied whether overnight token prices contain information useful to conventional markets. It regressed the conventional asset's return in the first five minutes of the U.S. session on the overnight tokenized-equity return. Its chart reports pass-through coefficients above approximately 0.87 across the sampled assets. That supports early price-discovery relevance. It does not demonstrate that every token price movement causes a stock to move by the same amount, and coefficients should not be confused with a percentage of individual trades predicted correctly.
Price discovery should not be confused with execution quality
Both tokenized and conventional prices can respond to the same overnight news. Informative price discovery does not establish causation or executable depth.
Consider two hypothetical traders holding a token linked to a U.S. stock. One wants to observe the implied direction of a post-earnings move. The other wants to sell a $50,000 position immediately. The first may find an informative overnight price even if only small orders are trading. The second needs executable depth. A mid-price can track the underlying share while the actual bid for a sizeable token order remains far away. It is possible for price discovery to improve before market quality catches up.
A stronger follow-up dataset would publish effective spreads, realized slippage by order size, venue-specific depth, and the time needed to restore alignment with the underlying share after the U.S. market opens. These measures would make the research more directly useful to investors than turnover alone.
The cost: less depth and more price variation
Tokenization places a share-like claim on a different trading infrastructure; it does not conjure market makers or arbitrage capital. If tokenized liquidity is fragmented across wallets, exchanges and chains, a small trade may cause a disproportionately large price change.
The IMF's empirical comparison found tokenized markets roughly 1.5 times as volatile, with decentralized venues generally showing the weakest liquidity. Part of the comparison reflects different trading hours: conventional-share volatility measures focus on exchange sessions, whereas a token that trades all day is exposed to overnight information continuously. The study uses careful normalizations, but that difference still matters to interpretation.
An investor's choice therefore involves at least three separate questions: Can the token be traded at the desired hour？ Can it be traded near a reliable fair value in the desired size？ Does the claim provide the legal and economic rights the investor expects？ A product can answer the first question well and the other two poorly.
The legal claim behind the ticker matters as much as the quote
A token displaying a familiar equity ticker may represent registered ownership, an intermediary-held beneficial interest, a contractual claim against an issuer, or synthetic price exposure. The quotation screen does not resolve which. Dividend treatment, corporate actions, settlement finality and redemption can differ materially among products with similar price charts.
This matters for the IMF comparison because a holder's liquidity depends partly on whether an authorized market maker can create or redeem the relevant claim efficiently. If issuance is slow, redemptions are restricted, or underlying markets are closed, deviations can persist even when onchain transfers are instantaneous. Investors should read the issuer and custody terms alongside the liquidity statistics before assuming the token is equivalent to the stock.
The financial-stability question extends beyond equity trading
The IMF identifies broader constraints: legal certainty over assets and holders' rights; consistent regulatory treatment; interoperability between networks; and reliable settlement money. It also warns that always-on automated margining and collateral reuse could amplify forced selling when liquidity is already thin.
This is particularly relevant as tokenized securities move from experimental investments into collateral, lending and treasury workflows. A token's legal title and settlement finality become important exactly when an intermediary defaults or an investor wants to redeem it.
The IMF estimates tokenized equities at about $2.3 billion in its July 31 snapshot, while its wider tokenized-asset perimeter totals about $65 billion. These numbers refer to a dated methodology, not necessarily all onchain representations of cash or every synthetic equity product. They should not be combined with broader market-cap estimates without checking definitions.
What would make the next study more compelling？
A larger sample across jurisdictions and legal structures would help. So would repeated comparisons that isolate identical U.S. market-hour windows from overnight periods, and measures of execution quality at realistic order sizes rather than only headline volume. The next proof point for institutions is whether better custody, legally enforceable shareholder rights and interoperable settlement narrow the actual liquidity gap.
The evidence now supports two conclusions at once: tokenized stocks meet real user needs, and they have not yet achieved the liquidity characteristics of the markets they mirror. A good product design has to address both findings, not use one to dismiss the other.
Sources: IMF, Global Financial Stability Report, October 2026; CoinDesk, October 11 summary of the empirical findings. Sample-size details and model coefficients describe the cited historical empirical exercise, not continuous monitoring or a universal property of tokenized shares.
