IMF finds over half of sampled tokenized US stock trading happens after hours

Editorial illustration: A classical financial building with closed shutters stands between a sunlit skyline and a moonlit skyline. Transparent tokens containing miniature financial buildings cluster along an undulating blue track on the夜

In brief

  • More than half of volume in five sampled products traded outside regular US market hours.
  • About 80% of tokenized equity transactions involved less than one share.
  • Realized volatility ran about 1.5 times that of traditional equivalents, partly due to differing trading windows.
  • Overnight pass-through to the next US open ranged from 87% to 99%; the IMF didn't establish causation.

What the IMF sampled

The study doesn't cover the whole market. The IMF examined five of the most actively traded tokenized US equity products, issued by Ondo and xStocks, which give exposure to the S&P 500, Nasdaq, Tesla, Alphabet and Nvidia. Together those products had a market value of about $345 million (roughly 15% of the $2.3 billion tokenized equity market as of July 31). The analysis covered 11 centralized and decentralized venues and more than 2 million five-minute observations over 365 trading days.

More than half of volume came outside the regular 9:30 a.m. to 4 p.m. Eastern session, including overnight and on weekends. The IMF said this points to demand from investors in other time zones and from people who want to react to news without waiting for US exchanges to reopen. Most transactions were smaller than one share, which the researchers read as a sign of a retail-oriented market (it's an inference from trade sizes, since they didn't identify individual investors directly).

Overnight prices and the next open

The next finding is the more interesting one. Using centralized-exchange data, IMF researchers compared overnight tokenized price moves with how the underlying securities did in the first five minutes of the next US session. The estimated pass-through ranged from 87% to 99%.

That's a strong link, not proof of cause.

The analysis didn't establish that on-chain trading drove the matching moves, and both markets may have been responding to the same developments.

Liquidity and volatility gaps

According to the IMF, traditional exchanges were much more liquid than tokenized platforms, and centralized crypto venues did better than decentralized exchanges. On decentralized platforms, relatively small trades could cause disproportionately large price moves because of limited market depth and market-making. Realized volatility was about 1.5 times that of traditional equivalents, and higher still on decentralized venues. Part of that gap comes down to trading windows. Tokenized products trade around the clock, while the traditional-market measure covered only regular US hours.

During regular sessions, tokenized prices generally tracked their underlying securities. Decentralized venues saw wider and more frequent deviations, especially when volume was low. The findings come as exchanges and financial institutions expand their efforts to bring US equities onto blockchain infrastructure.

Frequently asked questions

Does overnight tokenized stock trading move prices on Wall Street?

The IMF found a strong link but not causation. Overnight tokenized price moves showed an estimated 87% to 99% pass-through to how the underlying securities performed in the first five minutes of the next US session. The analysis didn't establish that on-chain trading caused those moves, and both markets may have been reacting to the same developments.

Why were tokenized stocks more volatile than traditional shares in the IMF study?

Realized volatility across the sampled tokenized equities was about 1.5 times that of their traditional equivalents, and higher on decentralized venues. Part of the gap reflects different trading windows: tokenized products trade continuously, while the traditional measure covered only regular US hours. On decentralized platforms, limited depth meant small trades could move prices disproportionately.

How large was the IMF's tokenized stock sample?

The IMF studied five actively traded tokenized US equity products issued by Ondo and xStocks across 11 venues and 365 trading days. The products were worth about $345 million, or 15% of the $2.3 billion tokenized equity market as of July 31.