$2.6T

FINRA's weekly ATS transparency data ranks venues by shares. That ranking is consistent and auditable, and it produces a competitive picture disconnected from the notional value transferred. We pulled 19 weeks of issue-level data across 27 venues, matched it against a 27-venue FINRA notional dataset, and mapped which names drive the tape, and where.

The result challenges a simple narrative. It's not that different ATSs route fundamentally different stocks. SPY, NVDA, MSFT, AMZN: every venue routes them all. The difference is how much of each name concentrates at each venue. UBSA (UBS ATS) routes $86.7 billion in SPY T1 notional over 19 weeks. IATS (IBKR ATS) routes $9.7 billion in the same name over the same period. That 9x gap in one name, across the same tier, is the structural story.

Below the mega-cap layer, the structural signal becomes a tier signal. 13 of 20 venues run 50–62% T1 by share count, a band tight enough to define an industry norm; a wider 45–62% band captures 16. BLUE (BOATS (Blue Ocean ATS)) at 12.5% T1 and IATS at 35.7% are the extreme outliers. Those outliers are analyzed separately. The tier-split analysis covers 20 venues on a share basis; the notional analysis covers 27 venues with reported FINRA notional data (20 profiled in the venue deep-dive cards).

Venue Reference

All venues are identified by MPID (Market Participant Identifier: the FINRA ATS identifier, distinct from ISO MIC codes) throughout this analysis. The full registered name and operator for each venue is listed below.

01The Dataset: Full Universe by Tier

The analysis draws from FINRA's ATS Weekly Transparency Data: weekly symbol and firm summaries published by FINRA. Each row is a symbol-level weekly summary, one record per MPID per symbol per week, reporting Total_Shares and Total_Notional transacted in that seven-day window. The window below covers 19 complete trading weeks, Jan 05 through May 11, 2026.

How notional is calculated: FINRA receives trade reports from each ATS and publishes the weekly aggregate notional directly as the Total_Notional field. This is not estimated: it is the sum of (price × shares) for all transactions reported by that MPID in that week for that symbol. No VWAP estimation or price approximation is applied. The figures are reported in US dollars.

NMS Tier 1 covers securities in the S&P 500, Russell 1000, or with a national market system designation meeting FINRA's liquidity threshold. Tier 2 captures everything else, including large foreign listings (ADRs), leveraged ETFs, and small-cap domestics. The tier is determined by the security's benchmark classification, not its price or ATS volume.

The 10,464 T2 symbols represent a long tail: the top 100 T2 names by notional account for the majority of T2 notional, while the remainder distribute across micro-cap and low-liquidity names. Tier 2 is not a monolithic category: it contains ultra-premium ADRs like ASML ($1,398/share) alongside single-digit-dollar names such as ONDS ($10.35/share at XSTM) and IATS's CRWV book (avg $97/share).

02The Mega-Cap Names Driving T1 Notional

Across 27 ATSs and 16,837 unique symbols in the Jan 05–May 11, 2026 window, $12.8 trillion in Tier 1 notional was reported to FINRA. The top 10 symbols by notional, SPY, QQQ, NVDA, MU, TSLA, MSFT, AMZN, META, SNDK, AVGO, account for $2646 billion, or 20.7% of all T1 flow.

Nearly seventeen thousand symbols compete for T1 share, and the top 10 hold20.7% of notional. SPY at $801 billion represents 6.3% of total T1 notional on its own, a significant anchor, and by far the largest single name in the ATS universe.

Semiconductors appear four times in the top 10 (NVDA, MU, SNDK, AVGO). ETFs anchor positions 1 and 2 via SPY and QQQ. The implied average price per share across these names ranges from $190 (NVDA) to $725 (SNDK), all high-dollar assets generating outsized notional per share.

03Where Each Name Trades: The Name × Venue Routing Map

The table below maps the top 8 T1 names against a cross-section of 11 venues: the largest venues by T1 notional, plus BLUE and JPBX included for contrast across venue types. Darker cells indicate higher notional concentration. Each cell shows $B over 19 weeks. Row totals reflect the 11-venue sum; the full 27-venue universe is larger, and includes venues such as CGXS and KCGM that are not shown in this view.

Several patterns emerge from the matrix. PURE (PureStream) and MSTX (MS Trajectory Cross) concentrate in equity names: NVDA, MSFT, AMZN, AAPL each show $10–15B at PURE and $6–9B at MSTX. INCR (IntelligentCross) leads SPY routing at $101.4B, with UBSA (UBS ATS) second at $86.7B. Both dwarf IATS (IBKR ATS), which routes just $9.7B in the same name, a 9x gap between the top two venues and the smallest shown. IATS leads in TSLA at $48.7B, more than 10x BIDS (BIDS ATS) at $4.5B, consistent with IBKR's retail-adjacent client base concentrating in high-volatility names.

SGMT (Sigma X2) is a top-three SPY venue at $43.6B among the 11 shown. Goldman's Sigma X2 showing high SPY concentration is consistent with block-trading and program-execution flows where S&P 500 hedges move through the ATS.BIDS (BIDS ATS) concentrates in NVDA ($18.2B) and SPY ($17.6B), consistent with a buyside-focused block matching profile.BLUE (BOATS (Blue Ocean ATS)) shows an unusual T1 signature: NVDA ($12.3B), MU ($15.7B), and QQQ ($12.1B) dominate its T1 book, while SPY ($6.4B) is comparatively muted. This semi-heavy, ETF-light T1 mix contrasts with the broader market where SPY is the dominant routing target.INCR (IntelligentCross) tilts toward SPY and QQQ at $101.4B and $35.2B respectively, with NVDA ($33.6B) and MSFT ($32.1B) rounding out its T1 profile.

03bVenue Deep Dive: 20 MPIDs, Top 5 T1 & T2 by Notional

The cards below profile 20 MPIDs from FINRA's weekly transparency data, sorted by total notional (T1 + T2 combined). Each card shows the venue's T1 and T2 notional bars (scaled to INCR as the maximum), symbol counts, share volume, and the top 5 names by notional in each tier. Avg price per share is computed as Total_Notional / Total_Shares from FINRA-reported data, not estimated.

03cSector Composition by Venue

The stacked bars below show how each venue's combined top-10 notional (five T1 + five T2 names) breaks down by sector. A venue's color mix reflects both its client mandate and its NMS tier architecture: ETF-heavy venues tend toward broad-market and leveraged products; single-stock venues tilt into Semis and Technology. BLUE's inverted T1/T2 ratio appears here as an outsized ETF·Lev weight from SOXL, TQQQ, AGQ, and SQQQ dominating its T2 book. LQNT's book is the most diversified outside the ETF and Semis clusters, with names like BDX, COR, and ABBV in its top 5 T1, reflecting its negotiation-style matching of block-sized institutional orders.

Note: percentages reflect composition of the top 10 symbols only (top 5 T1 + top 5 T2 by notional). The full venue book contains thousands of additional symbols; this view captures the concentrated flow that drives the majority of venue-level notional.

03dMid-Market Name Routing Concentration

Below the mega-cap layer, routing concentration becomes more idiosyncratic. The 20 names below are the largest T2 symbols by notional in the window, each clearing over $13 billion in combined FINRA-reported ATS notional across 19 weeks. They are not homogeneous: a 3x leveraged ETF (SOXL) routes very differently from a foreign ADR (BABA), a semiconductor ADR (TSM), or a high-growth technology name (CRWV).

Each cell is the actual FINRA-reported notional for that name at that venue, shown as a percentage of the name's total across these 20 venues. Every cell is real data: a blank cell means the venue reported no notional in that symbol during the window, not that the figure was unavailable. The 20-V% column on the right shows what share of each name's full FINRA-reported notional (all 27 venues) is captured by the 20 venues charted here.

Full coverage surfaces flow that a narrower, name-recognition-only view would have hidden. CRWV at INCR, for example, clears $5.62 billion, a routing relationship invisible in a view limited to headline names only.

04The Structural T1/T2 Split

The share-based tier split analysis covers 20 venues from FINRA issue-level data. 13 of 20 venues run 50–62% T1 by share count, a band defining the structural industry norm; a wider 45–62% band captures 16. BLUE (BOATS (Blue Ocean ATS)) at 12.5% T1 and IATS (IBKR ATS) at 35.7% are the extreme outliers.

The 9-percentage-point spread between INCR (IntelligentCross) at 51.8% and PURE (PureStream) at 60.9% encompasses venues operating very different client books. The convergence is not a regulatory artifact: no rule requires a 55% T1 ratio. It reflects the underlying composition of NMS-eligible trading: the liquid securities in Tier 1 attract more flow at roughly this ratio.

BLUE's 12.5% T1 concentration is the most extreme case. Of its 9.43 billion total shares, only 1.18 billion are T1. IATS at 35.7% is structurally T2-heavy but within a standard deviation of the peer group in a way BLUE is not. When both are mapped to the name-level routing matrix, they tell different stories: IATS concentrates in expensive T1 names (including leading TSLA at $48.7B); BLUE's T2 book contains a micro-cap sub-population at near-zero per-share prices.

05T2 Surprises: Ultra-Premium Foreign Names

Tier 2 is not uniformly a low-price universe. FINRA's NMS tier classification places non-Russell 1000 securities, including large foreign-listed ADRs, in Tier 2 by market capitalization benchmark. The result is a T2 cohort containing some of the highest-price securities in the dataset.

Taiwan Semiconductor (TSM) at $96.1 billion T2 notional over 19 weeks is the most striking case. That places TSM within striking distance of IWM ($121.2 billion), the third-largest ETF in the T1 ranking, but entirely in Tier 2. At $358/share implied average, TSM generates significant notional per trade.

ASML at $1,398/share and MercadoLibre (MELI) at $1,850/share are the highest per-share T2 names in the dataset. Both are classified T2 because they are non-US primary listings (ADRs), placing them outside the Russell 1000 threshold. Their T2 notional is small in share count but meaningful in dollar terms.

SOXL, the Direxion 3x Semiconductor Bull ETF, appearing at $48.1 billion T2 notional reflects a specific classification mechanic: leveraged and inverse ETFs are frequently excluded from Tier 1 benchmarks regardless of their notional size.

06T2 Price Spectrum: $12.11 to $38.03 Across Venues

Looking at the venue-level T2 average price (from the share-based 20-venue analysis), the per-share T2 composition spans from $12.11 at BLUE (BOATS (Blue Ocean ATS)) to $38.03 at LQNT (Liquidnet Negotiation ATS), a 3.1x spread across a single tier.

LQNT (Liquidnet Negotiation ATS) at $38.03/share T2 sits at the top. Its Tier 2 book consists of specialized international names, SE, ICLR, BAP, ASND, that FINRA classifies as T2. At $38.03 average, LQNT T2 carries significant per-share notional despite a small symbol count.

IATS (IBKR ATS) at $25.59/share T2 reflects expensive T2 names including CRWV, TQQQ, and NBIS. SGMT, JPMX, and LQNA cluster in the upper-middle range, routing moderately priced T2 names alongside strong T1 books.

BLUE (BOATS (Blue Ocean ATS)) at $12.11/share T2 sits at the bottom. Its anomaly is not the T2 price average, it is the volume of T2: 7.0 times more T2 shares than T1. The T2 average is pulled up by a leveraged ETF cluster (SOXL, TQQQ, SQQQ) and pulled down by a micro-cap cluster trading at sub-$0.50.

07BLUE (BOATS (Blue Ocean ATS)) in Context

Placed in the 20-venue table, BLUE ranks #15 by notional, consistent with its #8 share rank. The notional penalty from T2 dominance is partially offset by the leveraged ETF sub-population, which carries real prices. The micro-cap sub-population contributes negligibly to notional despite accounting for roughly 60% of T2 shares.

MU and LBGJ figures in the stat above are drawn from our companion analysis, One MPID, Two Markets: The Structural Split Inside BlueOcean ATS, which examines BLUE's per-share fee economics in more detail.

BLUE's 7.0:1 T2/T1 ratio is not simply a quirk of having a large T2 base. Every other venue in the table runs near parity on shares per tier, the median T2/T1 ratio is approximately 0.8x. No other venue approaches BLUE's ratio.

The economic consequences compound through the share-based fee structure. PFOF paid at $0.002/share generates 0.05 basis points of notional on MU at roughly $449 and 455 basis points on LBGJ at $0.044. ATS fees at $0.003/share represent 6.8% of the traded notional in a $0.044 security. These are the per-trade economics of the micro-cap sub-population at BLUE, running through market structure designed for a $10–$100 stock universe.

Among the 20 venues, BLUE is the only one where the T2 share count is driven primarily by a micro-cap engine. The other high-T2 venues, IATS at 64.3% T2 and KCGM at 53.0% T2, achieve their T2 weight via moderately priced names, not sub-$0.50 flow.

08The Share-to-Notional Rank Inversion

Re-sorting the 20-venue share data by actual FINRA-reported notional (rather than share count) reorders the competitive table. The largest divergences reveal which venues gain or lose competitive standing when the metric shifts from activity count to economic weight.

IATS (IBKR ATS) rises from #13 to #11 (+2 positions). IATS holds 35.7% T1, the second-largest T2 concentration in the universe, but its T2 book is not cheap. A T2 average of $25.59 combined with a T1 average of $165.59/share reflects a book routing expensive names across both tiers. This is consistent with what the name-level matrix shows: IATS leads TSLA T1 at $48.7B.

JPMX (JPM-X) and CGXS (OneChronos) each rise +2 positions. Both route expensive T1 names, SPY, MU, NVDA, META, at implied T1 averages above $100/share. The notional rank captures what share count misses: per-share richness in the book composition.

BLUE (BOATS (Blue Ocean ATS)) is the largest notional loser: #8 by shares, #15 by notional (-7 positions). Its 87.5% T2 share weight, dominated by leveraged ETFs and sub-$0.50 micro-cap flow, means high activity count does not translate to dollar weight. This is the fee and PFOF distortion case analyzed in Section 07.

09Takeaway

Three observations from the combined analysis:

1. The mega-cap names are universal; concentration is the variable.Every ATS routes SPY, NVDA, MSFT, and AMZN. The distinction is how much. UBSA routes $86.7B in SPY. IATS routes $9.7B. BIDS routes $18.2B in NVDA. SGMT routes $17.6B in the same name. The competitive signal is not name selection: it is notional allocation per name, which FINRA's share-based ranking cannot surface.

2. T2 contains premium names that the tier label obscures. TSM at $96.1 billion T2 notional, ASML at $1,398/share, and MELI at $1,850/share are classified T2 because their primary listing is non-US. The economic weight of these names rivals T1 mega-caps in dollar terms, but FINRA's tier classification groups them with genuinely low-priced T2 securities. Any competitive analysis treating T2 as a monolithic low-price pool misreads the data.

3. BLUE's structural position remains unique in this universe.No other venue runs a 7.0:1 T2/T1 share ratio. No other venue combines a leveraged ETF sub-population at $20–$57 with a micro-cap sub-population at $0.04–$0.35 within the same T2 book. The PFOF and fee distortion concentrated at BLUE does not generalize to the rest of the 20-venue table.

10Subscriber Data Download

Take the data with you. The full dataset behind this piece is available as an 8-sheet Excel workbook covering the overview, top T1 and T2 names, the single-names ex-ETF ranking, venue tier split, the T2 venue heatmap, the venue directory, and methodology. Same source data used to write this piece.

Download requires an active Sapinover session (any tier including free trial). Anonymous requests are redirected to the subscribe page.