ATS-N Regulatory Brief: What 31 Filings Tell Us About the Overnight Tape

The past 60 days saw 31 ATS-N filings across the registered alternative trading system universe. Most are administrative updates under Rule 304(a)(2)(i)(B). Four filings tell a different story. They reveal the FINRA rule change that quietly enabled the overnight tape, OneChronos's full architectural disclosure, BNP Paribas's new counterparty dampening mechanic, and the fee economics that subscribers to Bruce Markets actually face.

// PART 01The Rule That Lit Up Overnight

FINRA's Trade Reporting Facilities now open at 4:00 AM ET. Before March 30, 2026, the earliest TRF open was 8:00 AM. That single change is the structural infrastructure on which the overnight ATS ecosystem depends.

BlueOcean ATS, Bruce Markets, and Moon ATS all require the ability to print trades during pre-market hours. The TRFs are the pipes. Without TRF access in the 4:00 to 8:00 AM window, those trades cannot be reported to the consolidated tape. FINRA Regulatory Notice 26-07, effective March 30, 2026, made that window available.

Notice 26-07 also adds a carve-out for "qualifying overnight transactions." Last-sale reports carrying the .W modifier that either result from an overnight batch process or involve ETF NAV-based trades can be reported by 8:15 AM ET on trade date. This matters for practical venue operations. Not every overnight trade prints at execution time. Batch processing workflows can accumulate trades across the overnight session and report them as a single package at open.

The May 8 cluster in the ATS-N amendment filings reflects this change directly. Goldman Sachs (SIGMA X), J.P. Morgan (JPB-X), and Instinet all filed unilateral amendments on the same date to update their ATS-N disclosures to reflect the new TRF operating hours. This is procedural compliance, not operational change, but the simultaneous filings confirm the breadth of impact across bank-affiliated venues.

Any IBIT, ETHA, or SLV trades priced off NAV may carry the .W modifier and report later. This affects how clean overnight VWAP is for ETF flows.

For analysts parsing overnight ATS tape data, the .W modifier has practical significance. An ETF that prices off closing NAV and reports at 8:15 AM appears on the tape as an overnight trade, but its price discovery happened hours earlier at the fund level. When evaluating timing differentials for overnight ETF flow, the VWAP calculation may include trades whose prices were determined outside the overnight session window entirely.

// PART 02Same Category, Different Universe

OneChronos's February 17 material amendment is the public statement of its architectural differentiation from every other U.S. ATS. It is essential reading before any comparison against continuous-order-book venues.

The filing discloses, under Rule 304(a)(2)(i)(A) as a material amendment, the full operating architecture of the OneChronos matching engine. The core elements:

Any analysis comparing OneChronos (MPID: CGXS) against BIDS, BLUE, or other continuous-matching ATSs must account for this architectural difference. They serve different use cases. A periodic call auction with Expressive Bidding is not a dark pool. Share count and notional alone miss the operational reality.

// PART 03How Cortex Dampens Toxicity

BNP Paribas filed a Form ATS-N Updating Amendment on May 14 for its Cortex ATS. The substance is counterparty interaction logic, not a new order type or matching paradigm. But the mechanics are worth understanding for any institutional desk routing to Cortex.

The filing discloses a tiered subscriber categorization system. Subscribers can opt out of interacting with specific counterparty types, numbered 1 through 5. Type 1 is opt-out-by-default. The standard architecture for toxicity management in a dark pool: separate the flow you want from the flow you do not.

The headline mechanic disclosed is dynamic recategorization. Orders from Category 4 or 5 Subscribers — the "Direct Low" tiers, the most likely to be classified as high-toxicity flow — are reclassified to Category 3 after the order persists in the order book for 5 milliseconds.

Five milliseconds is an eternity in electronic market structure. It is also long enough to distinguish a passive institutional display order from a directionally aggressive sweep. An order that sits in the book for 5 milliseconds without filling is, by revealed behavior, not a market-impact-seeking order. The recategorization mechanic rewards passivity with expanded counterparty access. Subscribers in the higher-toxicity tiers can effectively earn their way into broader matching by displaying patient behavior.

Additional changes in the May 14 filing:

• Mid Point Peg opt-out now has separate configuration from non-midpoint orders.
• Self-cross prevention is now ON by default for external broker-dealer Subscribers identified by MPID.

An algorithmic mechanism for softening venue toxicity classifications. Not as headline-grabbing as a new order type, but a real flow-handling change worth understanding for any institutional desk routing to Cortex.

// PART 04The Economics Inside Bruce

Bruce Markets filed a Correcting Amendment on April 6 updating Item 19 (Fees) of its ATS-N. This is an ATS-N/CA under Rule 304(a)(2)(i)(C) — a correction, not a material change. But the disclosure is the clearest public accounting of Bruce Markets' fee structure available, and it reveals a revenue model that goes beyond execution fees.

Transaction fees for stocks priced at or above $1.00: $0.00 to $0.0020 per executed share when taking liquidity. For stocks under $1.00, the fee is up to 30 basis points of notional. Fees are negotiated individually per Subscriber, with discount discretion applied based on early-adopter status, historical volume, and expected future volume. The schedule defines the upper bound; the realized rate is set bilaterally per counterparty.

Pass-through fees are disclosed without markup: CAT (Consolidated Audit Trail), SEC Section 31 transaction fees, and FINRA TAF (Trading Activity Fee). These are regulatory costs, not revenue.

The more consequential disclosure is the market data licensing schedule. Bruce Markets Holdings, the parent company, licenses market data separately from execution. The consumer base for that data is broader than execution counterparties. It includes information portals and financial media (Yahoo Finance, Google Finance, CNBC, Bloomberg, and similar), brokerage front-ends and retail-facing platforms that redistribute ATS data to large user populations, and analytics and research providers, alongside trading firms that consume the data for internal use. The tiered schedule reflects that range of use cases:

The higher tiers in the display ranges correspond to redistribution use cases — licensees whose downstream audience runs into the millions — while the lower tiers cover smaller internal-use cases. This is a structural feature of how registered ATS venues monetize their proprietary tape, and it parallels the market data licensing models at NYSE, Nasdaq, IEX, and other regulated venues. The disclosure simply makes the schedule explicit. For any Bruce-versus-BlueOcean comparison, the data licensing line is material to the venue revenue mix, but it should not be conflated with execution-side cost to a trading-firm subscriber, which is governed by the transaction-fee schedule above.

// PART 05Methodology & Sources

All ATS-N filings analyzed in this brief were retrieved from SEC EDGAR via full-text search at efts.sec.gov. The search covered all ATS-N, ATS-N/MA, ATS-N/UA, and ATS-N/CA filings with effective or filed dates between March 28, 2026 and May 27, 2026, yielding 31 distinct filings across the registered ATS universe.

FINRA notices referenced were retrieved directly from finra.org/rules-guidance/notices and verified against the underlying rule text in the FINRA rulebook at finra.org/rules-guidance/rulebooks.

Filing PDFs were fetched directly from SEC EDGAR and text-extracted via the pypdf library. Redline documents — where available as a separate exhibit — were prioritized as the primary source for change identification. All quoted provisions and operational mechanics were verified against the filing document text and underlying rule citations before publication.

// PART 06References