The Clearing Gap, CME 24/7 Edition
This afternoon, when Chicago closes its Thursday session, CME Group’s regulated crypto futures and options will already have started trading around the clock. Bitcoin, Ether, XRP and Solana, continuously. A record $3 trillion in notional volume across the complex in 2025. Year-to-date 2026 average daily volume of 407,200 contracts, up 46% on the same period last year. The exchange question is answered. Institutional crypto derivatives are not going back to a five-day week.
Most coverage stops there and calls the market 24/7. Not actually true, and the way it is wrong is worth a few minutes of your time, because the distinction sitting under the headline is the most useful mental model you can carry into every clearing decision in tokenised markets over the next two years.
CME did not move one clock. It moved one and held the other still. Trading is now continuous. Clearing is not. The official wording from the 19 February announcement is precise: “all holiday or weekend trading from Friday evening through Sunday evening will have a trade date of the following business day, with clearing, settlement and regulatory reporting processed the following business day as well” (CME Group press release, “CME Group to Launch 24/7 Cryptocurrency Futures and Options Trading on May 29,” 19 February 2026, https://www.prnewswire.com/news-releases/cme-group-to-launch-247-cryptocurrency-futures-and-options-trading-on-may-29-302692346.html). Trading runs continuously on Globex. Clearing stays on Monday-to-Friday. That is two clocks, deliberately decoupled.
The reason the decoupling matters is not because CME has been timid. It is because the entire post-trade regulatory architecture rests on the assumption that a business day exists. Pull on any of the threads holding clearing in place and you find the same assumption underneath.
Start with CFTC Regulation 39.14
The rule that governs how a Derivatives Clearing Organization settles with its members. The text is unambiguous: “A derivatives clearing organization shall effect a settlement with each clearing member at least once each business day, and shall have the authority and operational capacity to effect a settlement with each clearing member, on an intraday basis, either routinely, when thresholds specified by the derivatives clearing organization are breached, or in times of extreme market volatility” (17 CFR § 39.14(b), Legal Information Institute, https://www.law.cornell.edu/cfr/text/17/39.14). Note what the rule does and does not require. It requires at least one settlement per business day. It permits intraday settlement on top. It assumes the business day as the anchor. There is no version of this rule that says “every twenty-four hours regardless of what day it is.” The cycle is the business cycle.
PFMI Principle 8
The international standard on settlement finality, sits on the same assumption. The principle requires that “an FMI should complete final settlement no later than the end of the value date” (CPMI-IOSCO, Principles for Financial Market Infrastructures, April 2012, https://www.iosco.org/library/pubdocs/pdf/ioscopd377-pfmi.pdf). Value date is the regulatory term for the business day to which a transaction is assigned for settlement purposes. The principle does encourage intraday or real-time finality “where necessary or preferable,” and CME Clearing’s own PFMI disclosure of December 2025 notes intraday mark-to-market capability for exchange-traded derivatives (CME Clearing Principles for Financial Market Infrastructures Disclosure, https://www.cmegroup.com/clearing/risk-management/files/cme-clearing-principles-for-financial-market-infrastructures-disclosure.pdf). But the standard against which a CCP is measured is end-of-value-date finality. The clock that matters for compliance is the business-day clock.
Now the payment rail
Fedwire Funds Service, the wholesale dollar settlement system that derivatives clearinghouses rely on, currently operates 22 hours per day on weekdays only. The National Settlement Service, used for clearinghouse settlement, runs weekdays only. The Federal Reserve announced on 9 October 2025 that both services will expand to include Sundays and weekday holidays, but the implementation is “no earlier than 2028” (Federal Reserve Board, “Federal Reserve Board announces expanded operating days of two large-value payments services,” 9 October 2025, https://www.federalreserve.gov/newsevents/pressreleases/other20251009a.htm). For now, and for the foreseeable next two years, the dollar rail that DCOs use for cash settlement does not run on Saturdays or Sundays. You cannot settle a margin call against Fedwire on a Saturday because Fedwire is closed. That is not a clearing infrastructure question. That is a payments infrastructure question, and it is upstream of every CCP.
Add the Basel and bank-side mechanics.
The capital benefit a bank receives for centrally cleared exposures runs through SA-CCR netting and the qualifying central counterparty risk weight. Both rest on the assumption that the CCP’s settlement and default-management processes operate within the regulatory framework, which in turn assumes the daily settlement cycle. Shift the clearing cycle away from that anchor and the capital treatment has to be rebuilt around whatever replaces it. No one has built that yet.
So when CME chose to extend the trading clock while holding the clearing clock still, it was not making a conservative choice. It was making the only choice the existing framework permits. To run continuous clearing, you would need: a payment rail that settles cash continuously, a CCP rulebook that defines finality without reference to a value date, a CFTC and international standards framework that does not anchor settlement to the business day, and a bank capital regime that recognises the new cycle as equivalent. Today, none of those exist. CME’s response is to prefund the gap: clearing members opt into a separate weekend settlement account and post collateral on Friday afternoon against anticipated weekend activity, so the exchange can monitor exposure during weekend hours even though final settlement waits for Monday’s business-day machinery to catch up.
This is where we think the main point is.
CME could decouple the clocks cleanly because trading and clearing sit in separate operational layers with separate governance, separate risk frameworks, and a clean interface between them. Extending the trading clock does not contaminate the clearing engine, because the clearing engine is downstream and architecturally separable. That separation is the legacy of a horizontal market structure, the same structure that traditional finance arrived at deliberately after working out, over decades, that exchange risk and clearing risk are not the same risk and should not sit inside the same operating entity. Branding the same, but legally separate.
A fully vertically integrated venue does not have that separation. When trading and clearing are run by the same legal entity, sharing infrastructure and governance, the two clocks are tied together by the structure itself. The vertically integrated venue facing the same constraints CME faced has three live options. It can run clearing continuously without the regulatory frameworks supporting it, in which case the capital benefits do not flow through to bank counterparties and the QCCP path is not available. It can hold the trading clock back to match the clearing cycle, in which case the 24/7 question is closed by structure rather than answered by design. Or it can absorb the mismatch internally without the architectural firewall that lets you contain it, which is what auto-deleveraging mechanisms in crypto venues are doing today: substituting socialised position reduction for the default-management framework that horizontal clearing relies on, because the venue has no other place to put the risk.
The fourth option, doing what CME just did, is not available to a fully vertically integrated venue. The decoupling CME demonstrated this week is a capability that horizontal architecture confers and vertical integration removes. It is more flexible and preferable.
That is the architectural fact under the 24/7 headline. The exchange question is what the headlines are covering. The clearing question, which is the one that determines whether the capital benefits flow through, whether the regulatory framework holds, and whether default management is structurally separable from trading activity, is the one this newsletter exists to follow.
This week the team has worked the same thread from different angles.
Neil Nguyen opened with the operational mechanics of what genuine continuous clearing would actually require. Matt Durkin walked the capital arithmetic, including the precision point that the capital benefits of clearing derivatives depend on the trade type and the underlying, with SA-CCR governing the counterparty exposure and SCO60 governing the Group 2 crypto underlying separately. Tomorrow, Phil Staddon takes the architecture question head-on, including the Gemini DCO licence in April as a vertical-integration precedent. Wolfgang Rückerl follows on Friday with the procurement dimension, the 18 to 24 month timeline from clearing counterparty selection to operational integration. If you want to go deeper on any of those, the posts are on LinkedIn this week.
On Saturday, Neil will have the first data point: what the opening weekend of decoupled trading and clearing actually looks like, not in theory. That data point will tell us how the prefunded weekend settlement framework holds up under live conditions. Watch it carefully. It is the first real test of architecture that has only existed on paper until now.
The big take away. Two clocks, not one. The clock you can move is the trading clock. The clock the regulatory framework rests on is the clearing clock. Whether a clearing venue can move one without the other is determined by whether the architecture separates them in the first place.
Horizontal clearing keeps that option open. Vertical integration closes it.
CTA
Following this story week by week:
Neil Nguyen https://www.linkedin.com/in/nghia-nguyen-neil/,
Phil Staddon https://www.linkedin.com/in/philip-staddon-393124273/,
Matt Durkin https://www.linkedin.com/in/matt-durkin/,
Wolfgang Rückerl https://www.linkedin.com/in/wolfgang-rueckerl/,
and Me on LinkedIn.
This first appeared in The Clearing Gap, my LinkedIn newsletter, on 28 May 2026. If you want to comment, discuss, or subscribe, that’s the place.