NSCC 24x5 Clearing Goes Live

01 · What NSCC 24x5 Actually Unlocks

The headline is “24x5 clearing.” The substance is three things.

Same-cycle novation

NSCC is the central counterparty for nearly all US equities trades through the Continuous Net Settlement (CNS) system. Before this launch, overnight ATS trades executed at 11 PM ET sat in a queue and were novated into the next business day's clearing cycle. The risk window between trade and clearing was effectively a full business day.

Same-cycle novation removes that gap. An 11 PM ET trade clears through the same operational window as an 11 AM ET trade: same legal substitution of NSCC as counterparty, same multilateral netting against other positions, same intraday finality.

Settlement risk windows compressed

Counterparty risk on overnight trades is now structurally indistinguishable from daytime risk. The margin-call window, the default-procedure timeline, the loss waterfall, all collapse to the standard daytime processes that institutional desks and their compliance teams already accept.

Operational debt retired

Most institutional middle-and-back-office systems were built around US daytime clearing cycles. Overnight trades required exception-handling code paths, separate reconciliation, and bespoke risk overlays. With NSCC clearing 24x5, the standard processing flow handles everything. The cost of supporting overnight execution drops materially.

02 · Where We Were: The Pre-NSCC Baseline

On June 23, 2026 the three overnight US equity ATSes that publish data we can measure (BlueOcean, Bruce, Moon) hit an all-time session record: $15.66B in notional, 547M shares, 2.87M trades. We isolated the ETF subset and ran a head-to-head against the 16 institutional dark pools that FINRA publishes under Rule 4552.

The head-to-head used the most recent contiguous four-week window where both NMS Tier 1 and Tier 2 data were published (April 6 through May 3, 2026). Over that 28-day window, the three overnight ATSes processed $32.6B in ETF notional. The 16 daytime institutional dark pools processed $53.4B. Combined institutional ATS ETF flow was $86.0B. The overnight share was 37.9%.

Splitting by NMS tier sharpened the picture. In Tier 2 ETFs specifically (leveraged, inverse, single-stock leveraged, and thematic vehicles) the overnight venues moved $18.6B against the FINRA-16 daytime venues' $20.3B. Tier 2 overnight share: 47.8%. Parity with the entire institutional dark-pool complex for an entire product category, on the OLD clearing constraint.

The detail goes further at the ticker level. Of the top 25 ETFs by overnight notional, 16 had a majority of their ATS flow overnight. Five exceeded 80%: USO (oil) 97%, SLV (silver) 90%, QQQM 87%, IBIT (Bitcoin) 83%, GLD (gold) 83%. Even the broad-index complex showed it: VOO 71%, QQQ 52%. The full table and the 12-sheet subscriber dataset are in the research piece.

The point worth pausing on: those numbers were achieved with the old clearing constraint in force. Institutional desks that held off on overnight routing because their compliance and risk teams required intraday novation were excluded from the dataset. That exclusion is now gone.

03 · Phase 2 - DTCC + Canton + Kinexys

24/5 trading does not just need 24/5 clearing. It needs 24/5 collateral.

NSCC 24x5 (Phase 1) handles the trade side. A market that genuinely operates continuously needs the funding and collateral side to operate continuously as well. That is where DTCC's Phase 2 work lives, and it is being built on top of two institutional DLT networks that already have real institutional adoption.

DTCC Collateral AppChain

DTCC's tokenization infrastructure transforms securities entitlements into tokens stored on a Digital Omnibus Account that operates 24/7. The implementation uses Chainlink's Cross-Chain Interoperability Protocol (CCIP) and ISO 20022 data standards for automated margining and collateral optimization. The point of architecture is to enable collateral to move between asset classes and across chains without the multi-day reconciliation that has historically choked the post-trade stack. We covered this in our prior architecture explainer (Tokenization & 24/7 Settlement); the relevant detail for NSCC 24x5 is that DTCC's Phase 2 work is not a greenfield build, it is a production-grade rail that already exists.

Canton Network

Canton is Digital Asset's privacy-enabled, institutional DLT network using a proof-of-stakeholder consensus mechanism that lets transacting parties verify state without exposing it to the broader network. The participant list reads as a roll-call of institutional plumbing: Goldman Sachs DAP, BNY Mellon, BNP Paribas, BlackRock, State Street, Deutsche Borse, Cboe, Paxos, MarketAxess, Tradeweb, Liquidnet, EquiLend, HQLAX, Standard Chartered, S&P Global, and DTCC. Japan is actively testing Japanese Government Bonds as digital collateral on Canton - proving that sovereign-grade assets can move 24/7 across borders without compromising the privacy expectations of institutional counterparties. The architecture enables atomic delivery-versus-payment with sub-network privacy preserved, which is the property that matters for tokenized settlement at institutional scale.

Kinexys (formerly J.P. Morgan Onyx)

Kinexys is JPMorgan's institutional tokenization platform, rebranded from Onyx in November 2024. The platform processes over $5 billion daily in institutional flow. JPM Coin has expanded onto Base, an Ethereum Layer 2, achieving near-instant 24/7 settlement for J.P. Morgan's institutional clients. The three pillars: Kinexys Digital Payments (24/7 USD and FX settlement), Kinexys Digital Assets (tokenized cash, deposits, intraday repo, collateral mobility), and Kinexys Liink (institutional data and identity exchange). The bank that once defined the messaging-based settlement system is now operating its own settlement-native rail at production scale. Kinexys is the bank-side counterweight to Canton: federated network on one side, J.P. Morgan rails for tokenized USD funding and collateral on the other.

The numbers say production, not pilot

The framing matters. DTCC's Collateral AppChain, Canton, and Kinexys are not experiments. Per our prior architecture explainer, tokenized Treasuries grew from $1.2B at the end of Q1 2025 to $11.2B by Q1 2026 (BUIDL, USYC, USDY, BENJI as the core issuers). Tokenized equities crossed $1.8B in Q1 2026 and are projected to triple by mid-year. The total RWA market sits at approximately $26B in early 2026. Kinexys moves $5B+ daily. Canton has sovereign-grade pilots running. This is production-scale infrastructure looking for its dominant use case, not a use case looking for infrastructure.

The regulatory permission slip is already signed

Phase 2 also rides on a legislative substrate that landed in 2025. The GENIUS Act of 2025 established the federal framework for payment stablecoins, removing the ambiguity that had kept large US banks from issuing tokenized deposits at scale. The CLARITY Act, advancing through 2025 into mid-2026, draws the jurisdictional line between SEC and CFTC over digital assets and creates the legal cover for tokenized securities to trade and settle on permissioned chains. For institutional counsel, the legal questions that used to block tokenization adoption have answers. The full timeline of both acts is laid out in our Tokenization & 24/7 Settlement architecture piece.

What Phase 2 delivers when it lands

DTCC's Project Ion and ComposerX work, combined with Canton's federated DLT and Kinexys's tokenized funding rails, gives institutional desks a 24/5 operating envelope that includes:

• Tokenized collateral substitution at any hour. A desk holding a long ETF position overnight can post or substitute collateral without waiting for US business-hours triparty windows. Kinexys Digital Assets is already running intraday repo and tokenized collateral mobility in pilot form.
• Tokenized USD funding 24/5. JPMorgan's tokenized deposits on Kinexys settle in seconds, not the next business day. Funding an overnight position becomes a real-time operation across timezones, not a next-day fund.
• Atomic-capable settlement of tokenized inventory. Canton enables delivery-versus-payment with privacy preserved at the participant level. As US equity and ETF inventory tokenizes on Canton (the work has begun with money-market funds and treasury collateral), settlement of overnight trades can move toward T+0 in the cases where it is operationally desirable.
• Cross-network interoperability. DTCC sits at the intersection. The promise of Phase 2 is not that DTCC, Canton, and Kinexys each operate in isolation, but that they interoperate: collateral from a Canton participant can margin a position cleared at NSCC and funded through Kinexys, in a single continuous cycle.

The combination matters because overnight trading without 24/5 collateral and funding is operationally constrained. Institutional desks can clear trades through NSCC at 11 PM ET (Phase 1), but the margin calls and funding flows still wait for US business hours unless Phase 2 lands. Phase 2 is the missing half of a genuinely continuous market.

Timing

Phase 2 will not arrive as a single launch the way Phase 1 did. The infrastructure is being built in pilot waves: tokenized treasury collateral first, intraday repo second, broader inventory tokenization third. We expect production scale on the first two within 12 to 18 months. The overall thesis - that 24/5 US equity trading rides on a tokenized settlement and collateral substrate - lands by 2027 to early 2028 in our base case.

04 · Wave 1 - EMS Switches Flip On

Next 30 to 60 days. The most visible immediate effect.

Most large prime-broker desks (Goldman Sachs, Morgan Stanley, JP Morgan, Bank of America, Citi) have built overnight routing capability into their execution management systems over the last two years. The capability has been there. It was turned off because compliance held the position that overnight execution carried clearing-cycle counterparty risk inconsistent with the desk's risk framework.

That risk framework just got rewritten by the underlying market infrastructure (Phase 1). We expect a measurable step-change in overnight ATS notional within 30 to 60 days, concentrated initially in broad-index ETFs (SPY, QQQ, IWM, VOO) where institutional flow is the marginal participant. Tier 2 names will see slower growth in this wave because they were already mostly overnight.

This is the wave that shows up first in our data. The next quarterly republication of the ETF Overnight Record methodology will measure it.

05 · Wave 2 - Lit Markets Catch Up

Q3 to Q4 2026. The competitive shake-out.

The other gate on extended-hours US equities trading was the SIP (Securities Information Processor). The CTA/CQ and UTP SIPs filed amendments in late 2025 to extend their hours; the target effective date is December 2026. NYSE, Nasdaq, CBOE, and the 24X Exchange have all been waiting for SIP extension AND clearing extension. Both are now in motion.

When lit overnight trading launches, some Tier 1 ETF flow that has been on overnight ATSes (mainly because that was the only venue option) will migrate back to lit venues for tighter spreads and more transparent price formation. The overnight ATS share of Tier 1 will likely give back ground.

The Tier 2 share holds, and probably grows. Leveraged, inverse, and single-stock leveraged ETFs do not price well on thin lit overnight order books; cross-network mechanics are structurally superior for those products. The three incumbent ATSes have a multi-year head start on the operational and routing relationships that make Tier 2 overnight execution work.

The competitive question for the three overnight ATSes becomes: do they consolidate share as the incumbents with the institutional plumbing already wired, or do they face displacement from new entrants with exchange-grade lit backing? Our base case is consolidation in Tier 2 with managed share loss in Tier 1.

06 · The Tier 1 versus Tier 2 Split Sharpens

One of the cleanest findings in the pre-NSCC research was the structural product split between venue types. Daytime institutional dark pools (the FINRA-16) primarily handle NMS Tier 1 equities and ETFs: SPY, QQQ, IWM, VOO, SMH, SOXX, GLD, SLV. Leveraged, inverse, and single-stock leveraged ETFs (Tier 2) are almost entirely absent from those venues.

That split is product-type-driven, not behavior-driven. Block-crossing networks built for mega-cap NMS Tier 1 stocks structurally cannot price Tier 2 inventory the same way. NSCC 24x5 clearing does not change that. It changes the willingness of institutional desks to use whatever venue is best for a product class on a 24-hour cycle.

Practical implication: the overnight ATSes will likely become more concentrated in Tier 2 flow over the next 12 months, even as their Tier 1 share gives back ground to lit overnight markets. The three venues consolidate into the Tier 2 leverage and single-stock specialists. That is a real franchise, not a transitional position.

07 · What We Will Track

We will republish the same methodology against a post-NSCC window in approximately 90 days (target: late September 2026, using FINRA T1+T2 data from the August publication batch). The deltas to watch:

1. Overall overnight share of ATS ETF flow. The 37.9% baseline. We expect this to climb. The question is whether it climbs through Tier 1 picking up institutional flow (Wave 1) before lit overnight launches take some of it back (Wave 2), and where the equilibrium settles.

2. Tier 2 share. The 47.8% baseline. We expect this to climb past 50% and stay there. If it does, the news headline writes itself: the majority of institutional Tier 2 ETF ATS flow is now happening overnight.

3. Ticker-level concentration. The five ETFs at 80%+ overnight share (USO, SLV, QQQM, IBIT, GLD). Watch whether commodity and Bitcoin ETFs cement into 90-95% overnight-dominant, or whether lit 24x5 brings them back to a 70-80% range.

4. Issuer concentration. Direxion's 47% share of ETF notional on the June 23 record night. Does Direxion's dominance grow as institutional flow joins their 3x semis bull/bear pair? Or does institutional flow disperse across a broader issuer list?

5. Phase 2 settlement signals. Public announcements from DTCC, Canton, and Kinexys on production milestones. Specifically: tokenized treasury collateral live across multiple custodians, intraday repo at production scale on Kinexys, atomic-settled tokenized ETF inventory pilots involving Canton participants. Each milestone tightens the case that 24/5 trading has the underlying funding and collateral infrastructure to support it at scale rather than at pilot scale.

6. Wholesaler 24/5 commentary. Public commentary from Citadel, Virtu, Susquehanna, G1X, Hudson River, and Jane Street on overnight market-making footprint. The wholesalers that already operate overnight gain a step-function in flow as Wave 1 lands. The wholesalers that have been daytime-heavy will need to announce overnight expansion within 60 to 90 days or accept a relative position erosion in ETF inventory.

08 · Implication for Flow Desks

For ETF flow desks at large institutions, the practical workflow changes are real and measurable in the next quarter:

Routing logic. Smart-order routers need an overnight model that is not bolted on as an exception path. The Random Forest finding from our research (88.5% of overnight notional comes from each ETF's habitual overnight activity) was an artifact of the OLD clearing constraint, where new participants had no path in. With the constraint gone, the predictive feature set will reset. Habit will be re-formed over the next 90 days.

Benchmarking. VWAP and TWAP calculations that exclude extended-hours prints are about to lose meaning for institutional flow. The benchmarking question for any ETF that trades meaningfully overnight is whether to compute VWAP across the full 24-hour window or whether to maintain a session-segmented model. We think the answer is session-segmented for now, full 24-hour by mid-2027.

Liquidity provider relationships. Wholesalers and market-makers that have been overnight-active (Virtu, Citadel, Susquehanna, G1X, Hudson River, Jane Street) get a step-function bump in flow. Wholesalers that have been daytime-only see their relative position erode. Expect public commentary from at least one of the daytime-heavy firms within 60 days about extending their overnight footprint.

The infrastructure layer has been built. The volume layer follows.

## References

1. DTCC. NSCC Now Live With Clearing Hours Extended: Expanded trading hours usher in a new era of global access and efficiency for U.S. equities markets. June 29, 2026.
2. Sapinover Intelligence. ETF Overnight Record · June 23, 2026: ML Deep Dive. Sapinover Research. June 24, 2026.
3. Sapinover Intelligence. Tokenization & 24/7 Settlement: An Architecture Explainer. Sapinover. May 2026. Provides the full architecture context for DTCC's Collateral AppChain, Canton Network proof-of-stakeholder consensus + Japanese JGB pilot, Kinexys $5B+ daily flow on Base, and the GENIUS Act 2025 + CLARITY Act legislative timeline. The numbers and timetable used in Section 03 of this piece are drawn from this companion explainer.
4. DTCC. Project Ion: Settlement on a permissioned distributed ledger. Phase 1 pilot completed 2024; ongoing development of T+0 capable settlement infrastructure.
5. Digital Asset. Canton Network. Privacy-enabled institutional DLT network. Participants include Goldman Sachs, BNY Mellon, BNP Paribas, BlackRock, State Street, Deutsche Borse, Cboe, Paxos, MarketAxess, Tradeweb, Liquidnet, EquiLend, HQLAX, Standard Chartered, S&P Global, and DTCC.
6. JPMorgan. Kinexys (formerly Onyx). Institutional tokenization platform launched November 2024. Includes Kinexys Digital Payments (24/7 USD and FX settlement; over $2 billion in daily volume), Kinexys Digital Assets (tokenized deposits, intraday repo, collateral mobility), and Kinexys Liink (data and identity exchange).
7. FINRA. Rule 4552 ATS Weekly Volume Data. Tier 1 and Tier 2 NMS publishing schedule.
8. CTA/CQ Plan and UTP Plan amendments to extend SIP operating hours, filed Q4 2025; target effective date December 2026.