Why US and EU Venue Taxonomies Diverge
Prerequisites
- trading-venues — the venue categories themselves (exchange/RM, ATS/MTF, OTF, SI) and how they map across jurisdictions
- trading-fundamentals — spreads, market makers, the NBBO, Reg NMS basics
The US has two venue categories (exchange and ATS). The EU has four (RM, MTF, OTF, SI). This isn’t an accident of bureaucratic style — it reflects fundamentally different starting problems, different political economies, and different legal traditions. This note explains how each jurisdiction arrived at its taxonomy and why convergence is unlikely.
The deepest driver
The US had a coordination problem — one dominant exchange with monopoly rents, needing competition within an existing structure. The EU had an integration problem — 15+ national monopolies, needing competition between structures while harmonizing rules across jurisdictions.
The US solved its problem with architecture (a linked national market system). The EU solved its problem with category creation (each new category addressed a specific market failure). Architecture needs few categories. Category creation, by definition, produces many.
Starting conditions: path dependence
US (c. 1975): monopoly with a dealer challenger
By the early 1970s the US market had a simple structure:
- NYSE (est. 1792) — the dominant auction market. NYSE Rule 394 (later Rule 390) required member firms to route orders to the exchange floor rather than internalize or execute away. This created a consolidation norm: one venue per security, effectively.
- NASDAQ (est. 1971) — a dealer market for a distinct universe of OTC securities. It competed with NYSE for listings, not for order flow in NYSE-listed stocks.
- Regional exchanges (Boston, Philadelphia, Pacific) — small, linked, supplementary.
The problem was not “too many competing venues” but one dominant venue with monopoly rents and fixed commissions. Congress’s response — the Securities Acts Amendments of 1975 (15 U.S.C. §78k-1) — did not create new venue categories. It mandated a National Market System (NMS): linked markets with best execution, order protection, transparency, and competition among market centers.
The crucial structural consequence: the US started with one regulator (SEC), one basic distinction (exchange vs. not-exchange), and a competition problem inside that structure. The remedy was plumbing (linkage, data consolidation), not taxonomy.
EU (c. 2000): 15 fragmented monopolies
The EU had no equivalent starting point. It had approximately 15–18 national stock exchanges (LSE, Paris Bourse, Deutsche Börse, Borsa Italiana, BME, SWX, OMX…), each operating as a domestic monopoly, each regulated by a different national competent authority (FSA, AMF, BaFin, Consob, CNMV…), each with different trading rules and settlement systems.
Several member states enforced a concentration rule under the Investment Services Directive of 1993 (ISD, 93/22/EEC, Art. 14(3)): retail orders had to be routed to the domestic regulated market. This was statute, not convention — it protected national exchanges and ensured domestic price discovery.
The EU problem was therefore the mirror image of the US problem:
| US | EU | |
|---|---|---|
| Structure | One dominant venue | 15+ domestic monopolies |
| Problem | Monopoly rents, no competition | No cross-border competition |
| Solution needed | Competition within | Competition between |
Any EU reform that wanted competition had to simultaneously dismantle domestic monopolies. You couldn’t just add plumbing (a consolidated tape, a trade-through rule) because there was no common tape, no common rulebook, no common concept of what an “exchange” even was. The EU had to build new categories from scratch — granular enough to accommodate pre-existing national idiosyncrasies without expropriating them.
The common-law vs civil-law overlay reinforces this: US regulation relies heavily on SEC rulemaking and no-action letters (flexible, principle-light). EU regulation uses Level 1 Regulation/Directive + Level 2 Delegated Acts + Level 3 ESMA Q&A (prescriptive, definition-heavy).
The US path: binary classification through pragmatism
1975 Amendments → NMS architecture
The Securities Exchange Act §11A directed the SEC to facilitate a national market system with “fair competition among brokers and dealers, among exchange markets, and between exchange markets and markets other than exchange markets.” Note the implied binary: exchange markets and other markets. This legislative framing embedded a two-category worldview from the start.
The SEC implemented gradually: the Intermarket Trading System (ITS, 1978) to link exchanges, the consolidated tape (CTA/CQ plans), and the abolition of fixed commissions (May Day, May 1, 1975). No new venue categories were needed.
Reg ATS (1998): deliberate non-categorization
By the mid-1990s, electronic crossing networks (ECNs) — Instinet (1969, but scaling in the ’90s), Island (1996), Archipelago (1997) — were matching orders electronically, functioning like exchanges but registered as broker-dealers. The SEC faced a regulatory gap: these systems were either illegal unregistered exchanges or broker-dealers operating under an ill-fitting regime.
The SEC’s solution, Regulation ATS (Release No. 34-40760, December 1998), was deliberately ambiguous. Rather than creating a new legal category, the SEC offered ECN operators a choice:
- Register as a national securities exchange under §6 of the Exchange Act
- Register as a broker-dealer and comply with the ATS rules
This choice architecture reflected the SEC’s explicit philosophy that “form should not determine function.” The ATS rules imposed scaled obligations based on trading volume (>5% → must provide access to all broker-dealers; >20% → must register as an exchange), but the fundamental category remained “broker-dealer operating an ATS.”
Why binary and not more granular? Three reasons:
- Legal parsimony. The Exchange Act defines “exchange” with a functional test (§3(a)(1)). The SEC used its exemptive authority — which required only two categories: things that need the exchange exemption, and things that don’t.
- Regulatory competition. Over-categorizing might drive activity offshore or into less-regulated structures. A flexible broker-dealer category was a pressure valve.
- Institutional preference. The SEC’s economists and the broker-dealer industry both preferred a system where the exchange category was hard to enter (high compliance cost, SRO obligations) but the ATS category was flexible. This preserved exchange prestige while allowing innovation in the ATS space.
Reg NMS (2005/2007): hardening the architecture
Regulation NMS (Release No. 34-51808) was the most significant US market structure reform since 1975:
- Order Protection Rule (Rule 611): prohibited trade-throughs of protected quotes — best displayed quotes at trading centers
- Access Rule (Rule 610): required fair access to quotations, capped access fees at $0.003 per share
- Sub-Penny Rule (Rule 612): prohibited sub-penny quoting
- Market Data Rules: reformed consolidated tape revenue allocation
The critical structural effect: Reg NMS legitimized the proliferation of exchanges while hardening the two-category system. Because it created a unified tape and trade-through protection, multiple exchanges could compete as equals. NASDAQ became an exchange (2006), Archipelago merged with NYSE (2006), BATS registered (2008), Direct Edge registered (2010). By 2010 the US had ~13 registered exchanges, all connected by the NMS architecture.
But the category structure did not multiply. The proliferation was within categories, not a multiplication of categories. Reg NMS both enabled venue proliferation and ensured it happened inside the exchange category — because the Order Protection Rule made being an exchange (with protected quotes) competitively valuable.
Dodd-Frank (2010): the one exception
The Dodd-Frank Act introduced Swap Execution Facilities (SEFs) under CFTC jurisdiction (7 U.S.C. §7b-3) — a genuine new venue category. But SEFs were for a different asset class (swaps) under a different regulator (CFTC). The equity market’s two-category structure was untouched.
SEFs could operate with an order book or an RFQ system with at least 3 competing quotes — a relatively flexible single category. The CFTC, like the SEC, operated under a binary framework: designated contract market (DCM, the exchange equivalent) or SEF (the “everything else that must follow rules” category).
The EU path: layered categorization through political accommodation
ISD (1993): the failed first attempt
The Investment Services Directive (93/22/EEC) tried to create a single EU market through passport rights for investment firms. But it explicitly allowed member states to impose concentration rules (Art. 14(3)) — a deliberate political compromise. National exchanges were left untouched.
The ISD period saw a growing gap between rules (execute on exchange) and practice (large broker-dealers increasingly internalizing flow). In the UK (which had abolished its concentration rule earlier), a liquid off-exchange market developed alongside the LSE. The EU was developing a two-tier reality.
MiFID I (2004/39/EC, implemented 2007): creating the MTF
MiFID I is where the EU’s multi-category architecture was born. Three foundational choices:
First, it abolished concentration rules (Art. 22). This opened EU equity markets to competition for the first time — the central political achievement. But having abolished concentration rules, the Commission needed a framework for the new competing venues.
Second, it created three venue categories:
- Regulated Market (RM) — successor to the traditional exchange. Multilateral, non-discretionary, operated by a market operator, strictest obligations (Title III). The incumbent category.
- Multilateral Trading Facility (MTF) — multilateral, non-discretionary, but lighter requirements. Could be run by investment firms, not just market operators. This was designed explicitly to accommodate ECN-type venues (Chi-X Europe, launched 2007; BATS Europe; Turquoise) competing with incumbent exchanges.
- Systematic Internaliser (SI) — an investment firm dealing on own account on an organized, frequent, systematic basis when executing client orders outside any multilateral venue. Designed to regulate large banks internalizing significant flow.
Why the MTF specifically? The Commission wanted to enable competition while maintaining investor protection. Simply allowing any investment firm to operate a crossing system would have been too permissive. The MTF created a middle tier: lighter regulation than an RM, but still multilateral, still non-discretionary, still transparent.
Third, the SI regime was articulated but weak. MiFID I’s SI definition (Art. 4(1)(7)) was qualitative — “organized, frequent, and systematic” — with no quantitative thresholds. Most investment banks argued they didn’t meet the definition and avoided transparency obligations while continuing to internalize. The SI regime was, in practice, a dead letter.
MiFID I achieved its primary goal: exchange monopolies were broken. Chi-X, BATS Europe, and Turquoise launched between 2007 and 2009. By 2011 the LSE’s share of FTSE 100 trading had fallen from ~100% to roughly 50–55%. But three new problems emerged:
- Dark trading exploded. Banks created internal crossing systems (broker crossing networks, BCNs) that argued they were bilateral, not multilateral — evading MTF classification and pre-trade transparency. BCNs reached 7–10% of European equity trading by 2012–2013.
- Bond and derivatives markets were untouched. MiFID I’s framework was designed for equities. Fixed income and derivatives continued to trade OTC. The 2008 crisis revealed the opacity and systemic risk.
- The SI regime was hollow. Banks traded significant bilateral volumes without transparency obligations.
MiFID II/MiFIR (2014/65/EU, implemented 2018): adding the OTF
MiFID II added the fourth category — the Organised Trading Facility (OTF) — and strengthened the SI regime. Each addition solved a specific failure:
The OTF: capturing voice brokers and hybrid systems. The EU faced a specific challenge after the G20 Pittsburgh commitment (2009) to move standardized derivatives onto electronic platforms. MTFs required non-discretionary execution, but fixed income and derivatives markets depend on dealer discretion — a voice broker arranging a corporate bond trade chooses which dealers to call, may work an order over hours, may provide capital commitment. An MTF couldn’t accommodate this.
The OTF (Art. 4(1)(23)) was defined as a multilateral system that is not an RM or MTF, where the operator may exercise discretion about how orders interact — subject to best execution and conflict of interest rules. The operator cannot trade against client orders on own account (with limited exceptions for illiquid sovereign bonds). The OTF captured:
- Broker crossing networks (which had to convert to MTFs or OTFs, ending the unregulated BCN category)
- Voice/hybrid platforms for bonds
- Derivatives trading platforms
Why not just expand the MTF definition? Because the non-discretionary requirement was load-bearing for equity market price formation. Removing it would have undermined the transparency rules that depended on non-discretionary matching. The EU chose additive categorization rather than definitional expansion — a new category rather than diluting an existing one.
The SI regime: quantitative thresholds. MiFID II replaced the qualitative test with quantitative thresholds (Commission Delegated Regulation (EU) 2017/565): an investment firm is an SI if, over the previous 6 months, it executed client orders in a financial instrument more frequently than 15 times per week on average, or with a size exceeding 0.5% of total EU turnover in that instrument. This captured the largest internalizers while excluding occasional bilateral trades.
Key divergence points
Dark trading
| US | EU | |
|---|---|---|
| Mechanism | Non-displayed ATS; no volume cap | MTF/RM with transparency waivers (RPW, LIS) |
| Philosophy | Disclosure-based — dark trading is fine, just report post-trade | Volume-limiting — dark trading must not exceed a fraction of total volume |
| Volume limit | None | Originally Double Volume Cap (4% per venue, 8% total); replaced by Single Volume Cap (7%) in 2024 |
The Double Volume Cap (MiFIR Art. 5) was immediately controversial and widely regarded as technically unworkable. ESMA’s implementation in 2018 led to mass waiver suspensions, but data quality problems undermined enforcement. The DVC was replaced by a simpler 7% cap under the EU Listing Act amendments (Regulation (EU) 2024/791).
OTC derivatives: SEFs vs OTFs
Both jurisdictions committed at the G20 Pittsburgh Summit (2009) to move standardized OTC derivatives onto electronic platforms:
| US (SEFs) | EU (OTFs) | |
|---|---|---|
| Scope | Derivatives only | Cross-asset (bonds, structured products, derivatives) |
| Regulator | CFTC (separate from SEC) | Same MiFID II framework as equities |
| Execution | Order book or RFQ with ≥3 quotes | Discretionary matching permitted |
The key divergence: the CFTC created a derivatives-specific new category, while the EU created a cross-asset general-purpose category. This reflects the EU’s holistic approach vs the US’s asset-class-siloed regulatory structure.
Retail internalization: PFOF vs the SI regime
| US | EU | |
|---|---|---|
| PFOF | Legal and widespread | Restricted under MiFID II Art. 24(9a); banned in most member states by 2026 |
| Internalizer category | None — wholesalers are broker-dealers | SI with pre-trade quote obligations (MiFIR Art. 14, 18) |
| Transparency | Post-trade only (Rule 605/606) | Pre-trade firm quotes for liquid instruments |
The EU created the SI category partly because European banks and exchanges resented the US-style dark internalization by large US banks (Goldman, Citi, Barclays) operating in European markets. The SI regime imposed transparency costs on these internalizers. German retail banking politics (the Sparkassen and cooperative banks had long used PFOF with regional exchanges) produced a temporary German exemption until 2026.
Political economy: who shaped these choices?
US
- NYSE’s political power dominated for decades. The two-category system preserved the exchange category as prestigious and high-cost, which entrenched exchange incumbents.
- Major broker-dealers (Goldman, Merrill) wanted the flexibility of the ATS category — operate dark pools and internalize flow without SRO obligations. Their lobbying against stricter ATS regulation was consistently successful.
- HFT firms (Citadel, Virtu, Jump) favored the Reg NMS architecture, which created uniform national linkage they could arbitrage across venues.
EU
- National exchanges (LSE Group, Deutsche Börse, Euronext) initially opposed MiFID I competition but adapted. Post-fragmentation, they consolidated and lobbied for the SI regime and the DVC — tools that imposed costs on new competitors (US banks’ dark pools).
- The European Commission and ESMA had an explicit single-market agenda. Every new MiFID category was partly about harmonization — ensuring the same activity was regulated identically across 27 member states. More categories meant more specific rules, less room for national variation.
- US banks operating in Europe lobbied hard against the SI regime and DVC, viewing them as barriers to their internalization businesses.
Current trajectory
US: reform stalled
SEC Chair Gensler’s 2022–2023 equity market structure reform package proposed four major changes:
- Order Competition Rule — competitive auctions before internalization (would have fundamentally altered the PFOF model)
- Reg Best Execution — heightened best execution standards
- Tick size reform — reduced ticks for tick-constrained stocks
- Enhanced ATS disclosure
By late 2024, all four were effectively dead — withdrawn, not finalized, or subject to legal challenges. The two-category system is essentially unchanged from Reg NMS (2005/2007).
EU: four categories modified but intact
The Listing Act (Regulation (EU) 2024/791, in force November 2024) made significant modifications:
- Consolidated tape mandated — a pan-European consolidated tape for equities and bonds, the EU’s most significant missing piece of market infrastructure
- Volume cap simplified — DVC replaced by single 7% cap
- SI thresholds adjusted
- Tick sizes harmonized across venues
The four-category structure remains intact. There is no proposal to collapse OTF into MTF or merge RM and MTF.
UK post-Brexit: a third path?
The UK’s Wholesale Markets Review (2022) and subsequent FCA consultations have discussed simplifying venue categories (currently mirroring MiFID II). The FCA’s Discussion Paper DP23/4 (2023) explored reforming the SI regime. If the UK adopts a simplified US-style binary taxonomy, it would create competitive dynamics between London and EU venues.
Summary
| Dimension | US | EU |
|---|---|---|
| Starting problem | One dominant exchange, monopoly rents | 15+ national monopolies, fragmentation |
| Primary tool | Architecture (NMS linkage) | Category creation + competition |
| Legislative style | Functional flexibility, exemptive authority | Exhaustive definition, harmonization |
| Categories | 2 (exchange, ATS) | 4 (RM, MTF, OTF, SI) |
| Dark trading | Permissive, disclosure-based | Volume-capped, waiver-based |
| Retail internalization | PFOF-permissive, no special category | SI regime with pre-trade transparency |
| Derivatives venues | SEFs (CFTC, one category) | OTFs (cross-asset, integrated into MiFID) |
| Current trajectory | Reform stalled, binary entrenched | Four categories modified but intact |
The US solved its 1970s problem with architecture and needed only two categories to implement it. The EU solved its 2000s problem with category creation and ended up with four categories as a result of solving four different problems sequentially across two directives.
Source summary
| Source | Coverage in this note |
|---|---|
| Securities Acts Amendments of 1975, Pub.L. 94-29 | NMS mandate, §11A architecture |
| Regulation ATS, Release No. 34-40760 (SEC, 1998) | Binary choice architecture |
| Regulation NMS, Release No. 34-51808 (SEC, 2005) | Order Protection Rule, venue proliferation |
| Dodd-Frank Act, Pub.L. 111-203 (2010), Title VII | SEF creation |
| ISD, 93/22/EEC (1993) | Concentration rules, passporting |
| MiFID I, Directive 2004/39/EC (2004) | MTF creation, concentration rule abolition |
| MiFID II, Directive 2014/65/EU (2014) | OTF creation, SI quantitative thresholds |
| MiFIR, Regulation (EU) No 600/2014 | Transparency waivers, DVC, SI pre-trade obligations |
| EU Listing Act, Regulation (EU) 2024/791 (2024) | Consolidated tape, volume cap reform |
| Ferrarini & Moloney (2012), Oxford Handbook of Financial Regulation | MiFID I fragmentation analysis |
| Gomber, Koch & Siering (2017), J. Business Economics | MiFID II context |
| Degryse, de Jong & van Kervel (2015), Review of Finance | EU dark trading post-MiFID I |
| O’Hara & Ye (2011), J. Financial Economics | US fragmentation and market quality |
| Angel, Harris & Spatt (2015), Quarterly J. Finance | US equity market structure |
| Harris, Trading and Exchanges (2003) | Foundational microstructure background |
See also
- trading-venues — the venue categories themselves: definitions, mapping table, and which assets trade where
- trading-fundamentals — spreads, market makers, Reg NMS, the NBBO
- trade-types — agency vs principal, block trades, algorithmic execution, retail PFOF
- settlement-and-clearing — CCP clearing for OTC derivatives post-2008