For educational purposes only; not investment advice. Investing may result in loss.
Direct answer
A dark pool is an informal name for a trading venue that does not display some or all trading interest to the public before execution. In U.S. equities, many dark pools are alternative trading systems, or ATSs, but the terms are not synonyms. An ATS can support displayed or nondisplayed trading and securities other than stocks, while off-exchange volume also includes broker-dealer internalization and other non-ATS activity.
Institutions may use nondisplayed liquidity to reduce information leakage and market impact, obtain midpoint or other price improvement, find a block counterparty, or control how an order interacts. The trade-off is uncertain fill probability, fragmented liquidity, routing and operator conflicts, adverse selection, latency, and less pre-trade information for public price discovery.
The venue label does not determine whether an execution was good. Evaluate the full order and benchmark: price, size, fill rate, speed, fees or rebates, information leakage, price movement while waiting, unexecuted quantity, and the cost of completing the order elsewhere.
Venue structure and regulation
Under the U.S. framework, an ATS generally meets the legal definition of an exchange but operates under an exemption from national-securities-exchange registration. It must be operated by a registered broker-dealer and comply with applicable Regulation ATS requirements. Filing Form ATS is a notice rather than blanket SEC approval of the venue or its executions.
An ATS that trades NMS stocks has public Form ATS-N disclosures about its broker-dealer operator, affiliates, subscribers, order types, matching and priority, segmentation, fees, market-data use, routing, counterparty interaction, and safeguards for confidential trading information. Read the current filing and amendments; a venue name, ownership, rules, or operating status can change.
Nondisplayed orders may be midpoint-pegged, limit-priced, immediate-or-cancel, conditional, minimum-quantity, or available only through a crossing session or negotiation protocol. Some venues segment participants or apply anti-gaming controls. These design choices affect eligibility, queue priority, information leakage, fill rate, and the counterparties with which an order may interact.
For listed U.S. stocks, executed ATS trades are reported to a FINRA trade-reporting facility and enter consolidated post-trade reporting. FINRA separately publishes security- and venue-level OTC trading statistics on a delayed basis. Do not confuse a post-trade print, delayed aggregate venue statistics, and the pre-trade order book: they reveal different information at different times.
A broker’s best-execution duty is broader than obtaining the best displayed price on one trade. FINRA Rule 5310 requires reasonable diligence under prevailing circumstances and regular and rigorous review where applicable. Relevant factors include price improvement and disimprovement, likelihood, speed and size of execution, transaction costs, customer needs, and internalization or payment arrangements.
Rule 605 execution-quality reports and Rule 606 routing disclosures can assist analysis, but coverage, order categories, timestamps, and reporting entity matter. They are not a complete score for every institutional algorithm or order. A broker, venue operator, affiliate, liquidity provider, or customer may have different incentives, so trace fees, rebates, payment for order flow, ownership, routing discretion, and use of order information.
Execution-quality examples
Suppose the national best bid and offer when a marketable buy order arrives is $49.98 bid / $50.02 ask. The quoted midpoint and spread are:
midpoint = ($49.98 + $50.02) / 2 = $50.00
quoted spread = $50.02 - $49.98 = $0.04 per share
If 10,000 shares execute at $50.01, price improvement relative to the displayed ask is:
price improvement = $50.02 - $50.01 = $0.01 per share
For this buyer, effective spread relative to the midpoint at order receipt is:
effective spread = 2 × ($50.01 - $50.00) = $0.02 per share
effective-to-quoted spread ratio = $0.02 / $0.04 = 50.0%
If the full order instead executes at the midpoint, its effective spread under this simplified calculation is $0.00. These measures address execution price, not fill probability, delay, fees, information leakage, or the cost of shares that did not execute.
Now suppose an institution wants to buy 100,000 shares. Immediate execution of the entire order at the initial displayed ask, assuming unrealistic full depth and no price movement, would cost:
immediate benchmark cost = 100,000 × $50.02 = $5,002,000
Instead, the institution first seeks a dark midpoint match. The venue fills 60,000 shares at $50.00, so:
dark fill rate = 60,000 / 100,000 = 60.0%
dark fill cost = 60,000 × $50.00 = $3,000,000
If the remaining 40,000 shares then execute at the unchanged ask of $50.02:
stable-market completion cost = 40,000 × $50.02 = $2,000,800
stable-market total cost = $3,000,000 + $2,000,800 = $5,000,800
stable-market average price = $5,000,800 / 100,000 = $50.008 per share
stable-market savings versus benchmark = $5,002,000 - $5,000,800 = $1,200
But assume the wait matters and the displayed ask rises to $50.06 before the remainder is completed:
rising-market completion cost = 40,000 × $50.06 = $2,002,400
rising-market total cost = $3,000,000 + $2,002,400 = $5,002,400
rising-market average price = $5,002,400 / 100,000 = $50.024 per share
cost above immediate benchmark = $5,002,400 - $5,002,000 = $400
Measured against the initial midpoint, the simplified arrival-price shortfall is:
arrival-midpoint shortfall = ($50.024 - $50.00) × 100,000 = $2,400
The dark fill saved spread on part of the order, yet delay made the overall strategy worse than the immediate-ask benchmark. This is an illustration, not proof that dark or displayed routing caused the subsequent price move. A real transaction-cost analysis should include market movement unrelated to the order, commissions, fees, taxes, spread, impact, delay, opportunity cost, cancellations, partial fills, and the chosen benchmark’s timestamp.
Review checklist
- Define the sample: ATS, dark ATS, all off-exchange trading, internalizer, block venue, or a specific order protocol.
- Confirm the security universe, trading session, date range, venue status, corporate actions, and symbol changes.
- Read the current Form ATS-N and amendments for ownership, affiliates, subscribers, order types, matching, priority, and segmentation.
- Identify midpoint, primary-market, NBBO, limit, auction, negotiation, or other pricing references and their timestamps.
- Check whether quotes were accessible, protected, locked, crossed, stale, odd-lot, or affected by data and routing latency.
- Separate submitted, eligible, conditional, firm, canceled, rejected, routed, and executed quantities.
- Measure fill rate, partial fills, time to fill, size improvement, and the cost of completing residual shares.
- Compare execution price with arrival midpoint, decision price, displayed quote, volume-weighted price, and other justified benchmarks.
- Calculate quoted spread, effective spread, price improvement or disimprovement, realized spread, and implementation shortfall consistently.
- Include commissions, venue fees or rebates, taxes, borrow cost, market impact, delay, and opportunity cost where relevant.
- Segment results by side, order type, size, urgency, liquidity, volatility, time of day, spread, and market regime.
- Test whether midpoint fills are followed by adverse price moves and whether selection effects explain the result.
- Examine minimum-quantity rules, conditional orders, indications of interest, pinging, counterparty filters, and anti-gaming controls.
- Trace router discretion, venue ownership, affiliates, principal trading, internalization, payment for order flow, and other incentives.
- Reconcile consolidated trade prints with FINRA’s delayed ATS and non-ATS statistics without treating either as a live order book.
- Use Rule 605 and Rule 606 data only after confirming coverage, definitions, order categories, and reporting periods.
- Compare venues using matched orders or controlled analysis; raw average price or fill size can reflect different order flow.
- Review outages, capacity, access, fair-access obligations where applicable, confidentiality controls, surveillance, and enforcement history.
- Do not infer hidden institutional accumulation or distribution solely from a venue print, volume ratio, or social-media indicator.
- Archive filings, data versions, clocks, filters, benchmarks, calculations, routing assumptions, exceptions, and sensitivity tests.
Common misconceptions
- Every off-exchange trade is a dark-pool trade. Off-exchange activity includes ATS and non-ATS trading, including broker-dealer internalization.
- Dark means illegal, unreported, or unregulated. Lawful ATS activity is regulated and completed listed-stock trades are reported, while pre-trade interest can remain nondisplayed.
- Dark-pool volume reveals bullish or bearish intent. Every execution has a buyer and seller; prints and aggregate volume generally do not reveal the full parent order, motivation, remaining quantity, or directional information.
- A midpoint fill proves best execution. Price is one factor; fill probability, speed, size, fees, information leakage, residual execution, and customer instructions also matter.
- Dark pools eliminate market impact. Nondisplay can reduce signaling, but information can leak and unfilled or later-executed shares can incur greater impact and opportunity cost.
Related topics
Sources
- U.S. Securities and Exchange Commission: current ATS list, Regulation ATS framework, and the distinction between notice filing and exchange registration.
- U.S. Securities and Exchange Commission: public Form ATS-N filings and operational disclosures for NMS stock ATSs.
- U.S. Securities and Exchange Commission: rules for NMS stock ATS transparency, confidential trading information, and Commission review.
- Financial Industry Regulatory Authority: delayed ATS and non-ATS transparency data derived from reported OTC trades.
- Financial Industry Regulatory Authority: best-execution factors and regular and rigorous review of routing and execution quality.
- U.S. Securities and Exchange Commission: amended Rule 605 execution-quality disclosures, including spread, price-improvement, size, and timing measures.