The Market’s Blindfold: Understanding Dark Pools, Hidden Venues, and Information Asymmetry
Everything so far in this series has assumed a shared stage: one order book, one price, one chart everyone can see. The stage is split. Some trades happen on public order books where every bid and offer is shown before it trades. Others happen in venues that show nothing until the trade is done, and a growing share of the market has moved there. The price on your screen is built from both, and your chart's volume may include only one of them.
This page is meant to be come back to. It starts with a map of the series, since this is its last part, and then sets out where trades actually happen: exchanges, dark pools, single-dealer platforms, wholesalers and OTC desks, with their crypto counterparts. It covers how trades in the dark reach the public record, how much of the US market trades that way, and what that means for the chart in front of you. Then it covers who sees what and when, and what all this does to price discovery. It ends with a reference table, a case study of two dark pools that didn't work as advertised, where the argument fails, and a checklist.
The Series So Far
| Part | Layer | Takeaway | Later posts that build on it |
|---|---|---|---|
| 1 · The Long Shadow of Markets | Time and place | The same value priced in two places disagrees, and arbitrage drags the prices into line as far as costs, capital and rules allow | Beyond Leverage, Turn the Candles Off, The Benchmark Must Bleed |
| 2 · The Invisible Balance | Outcomes | Outcomes concentrate, a crowded edge decays once enough traders act on it, and trading against other traders is zero-sum before costs | The Lab Book |
| 3 · The Depths of the Market | Liquidity | Resting orders pool where people look and price is drawn to them; markets drift back to where volume built up | Tool 01, Turn the Candles Off |
| 4 · The Script and the Rhythm | Structure | Accumulation and distribution leave a footprint in price and volume that can be read as it forms | Tool 01, Setup 01, Turn the Candles Off |
| 5 · The Shadow Engine of Markets | Derivatives | Derivatives shape the underlying through basis, funding, liquidations and hedging, and often lead it | The Lab Book |
| 6 · The Hidden Hand of the Market | Market makers | The spread pays for processing, inventory and adverse selection, and dealers skew or pull their quotes to manage that risk | |
| 7 · The Market's Blindfold | Venues and data | About half of US share volume trades away from the exchanges and reaches the tape only as prints, and your chart may not include it | Tool 01, Turn the Candles Off |
The Blindfold of Transparency
A chart presents itself as the whole market: every trade, every price, laid out for anyone to read. What it actually shows is a selection. On an exchange, the order book displays resting bids and offers before they trade, though some of them are bluffs and some size is hidden, as Part 3 covered. Off the exchange, nothing is displayed beforehand. The trade happens, and the public record learns of it afterwards, if your data includes that record at all.
So two different things can be missing from what you see. Intentions are missing everywhere: no venue shows you every order before it trades. And completed trades can be missing from your chart, depending on which venues your data comes from. The first is built into how markets work. The second is something you can check.
Where Trades Happen
| Venue | What it is | Shows orders before they trade? | How its trades reach the US tape | Crypto counterpart |
|---|---|---|---|---|
| Exchange | A public order book anyone can reach through a broker | Yes, displayed orders; hidden order types also rest there | Directly, as they happen | Exchange order books |
| Hidden orders on an exchange | Non-displayed or partly displayed (iceberg) orders inside a lit book | No | As exchange trades | Hidden and iceberg orders |
| Dark pool (ATS) | A broker-run system that matches orders without displaying them, usually at the midpoint of the best public bid and offer | No | Reported to a FINRA trade reporting facility (TRF) | Rare; OTC desks and request-for-quote systems do most of this work |
| Single-dealer platform | One firm's own liquidity, offered electronically to its clients | No | Reported to a TRF | Market makers' direct streams to institutions |
| Wholesaler | A firm that buys retail order flow from brokers and fills it from its own inventory | No | Reported to a TRF | App brokers that fill orders at a market maker's prices |
| OTC or block desk | A trade negotiated privately, usually large, by phone, chat or request for quote | No | Reported to a TRF | OTC desks, and block-trade facilities on derivatives venues |
Exchanges are what most people picture. Everything below them in the table is off-exchange, and in the US it all reaches the public record the same way: as a trade report after the fact. Hidden orders sit in between. They're inside an exchange's book but invisible until they trade, and by Rosenblatt's count they executed about 11% of US volume in December 2025.
Dark Pools: Liquidity in the Shadows
Dark pools began as a practical answer to a real problem. A pension fund that needs to buy millions of shares can't show that order on a public book without the market moving against it before it's filled. Other traders see the size, buy ahead of it, and sell back to the fund at higher prices. A dark pool lets large orders meet without being displayed. Orders match inside the pool, usually at the midpoint of the national best bid and offer, and the public sees the trade only once it's done.
That protects the fund, and it splits the market. Some of the supply and demand that sets prices is now hidden from everyone outside the pool, including other large traders who might have taken the other side. The pool's operator also sees every order resting in it, which is valuable information, and its subscribers have to trust the operator not to misuse it or let others misuse it. The case study below is about two operators who broke that trust.
Dark pools are a smaller part of off-exchange trading than their name suggests. In Cboe's count for 2025, 18.7% of off-exchange volume traded on ATSs and 81.3% through principal dealers, firms trading from their own inventory, and retail orders filled by wholesalers are a large part of that.
Internalisation and Hidden Venues
When a US retail trader presses "buy", the order usually doesn't go to an exchange. The broker routes it to a wholesaler, a market-making firm that fills it from its own inventory, and the wholesaler often pays the broker for the order. That payment is payment for order flow (PFOF), and it's what funds commission-free trading.
The wholesaler pays because retail orders are good business to fill. They come in small sizes and usually aren't based on information the market doesn't have, so a dealer that takes the other side rarely finds the price moving against it straight away, the adverse selection Part 6 described. The wholesaler typically fills at or slightly better than the best public quote and keeps the rest of the spread.
Whether this is good for the retail trader is argued both ways:
- For: wholesalers' fills often improve on the public quote, commissions have gone to zero, and execution is fast and certain.
- Against: the broker has a conflict of interest, choosing where to send orders partly on who pays most. And taking the least risky orders off the exchanges leaves the public books trading against more informed flow, which may make quoted spreads wider for everyone.
Regulators have come down in different places. The UK's regulator has treated payment for order flow as incompatible with its conflict-of-interest and best-execution rules since 2012, though it said in March 2026 it would review that position. The EU banned it in its 2024 review of market rules, with a transition for countries where it was established that ended on 30 June 2026. In the US the SEC proposed in 2022 that retail orders be exposed in auctions before a wholesaler could fill them, and withdrew the proposal on 12 June 2025. US brokers do have to publish quarterly Rule 606 reports showing where they route customer orders and what they're paid for them. That's the place to look if you want to know where your own orders go.
How Off-Exchange Trades Reach the Tape
In US equities, an off-exchange trade has to be reported to one of FINRA's facilities as soon as practicable, and within 10 seconds of execution. The report goes out on the consolidated tape alongside exchange trades, marked as a FINRA trade. So the trade does become public, with its price, size and time. What never becomes public is the order before it traded, and the venue is hidden too: the print says the trade was off-exchange without saying which dark pool or dealer did it. FINRA does publish each ATS's and each dealer's weekly volume in each stock, two weeks later for the most active stocks and four weeks later for the rest.
How much of the market trades this way has risen for over a decade:
By Cboe's market-volume data, off-exchange trading was about 34% of US share volume in 2009, still under 38% in 2019, 47.0% in 2024 and 50.6% in 2025, the first year it was more than half, as Cboe's own year in review also reports. Counted by value traded instead of shares, the 2025 figure is lower, about 44.7%, because a lot of off-exchange volume is in very cheap stocks. Either way it's close to half the market.
Crypto works differently. There's no consolidated tape and no FINRA: each exchange publishes its own trades, and OTC trades aren't reported publicly at all. When an OTC trade is settled on-chain, the transfer may be visible, but nothing marks it as a trade, its price or its counterparties.
What Your Chart Shows
This is the part that matters for reading a chart. Whether the off-exchange half of US volume is in your volume bars depends on your data feed. A consolidated feed includes the FINRA prints. A feed from one exchange or one exchange group includes only that venue's trades, which can be a small fraction of the total, and charting platforms often show such a feed until you pay for consolidated data. The price will look right either way, because every venue trades close to the same price. The volume won't.
In crypto the question is which venue's volume you're reading. A single venue shows only its own share, and as Part 1 covered, a move can start on one venue and reach others through arbitrage with much less volume behind it. Tool 01 treats a volume climax picked on partial data as provisional for this reason, and Turn the Candles Off covers how to compare volume bar to bar when the bars only show part of the market.
Information Asymmetry: The Unequal Gaze
Venues are one divide. Who sees what, and when, is another.
Speed Asymmetry
Firms that co-locate their servers beside an exchange's matching engine receive its data and reach its book microseconds before anyone further away. When a price moves on one venue, quotes on the others are stale for that long, and the fastest traders can trade against them before they update, the shortest gaps in Part 1's timescale. For everyone else, by the time a tick appears on the screen it's already history.
Data Asymmetry
A retail screen typically shows the best bid and offer and perhaps a few levels below. Professional desks pay for each exchange's full-depth feed, delivered directly and faster than the consolidated feed that sets the official best prices. Even the deepest feed is partial: hidden orders stay hidden from everyone except the venue holding them, and off-exchange orders are never shown at all.
Knowledge Asymmetry
Some participants see the order flow itself. A wholesaler sees a large share of US retail orders before they're filled. A dealer sees its clients' trades before it hedges them, as Part 6 showed. A dark pool operator sees every order resting in the pool. None of that is visible on a chart, and it's the information the largest firms trade on. Wyckoff's "composite operator", from Part 4, is a way of imagining that informed side as a single actor whose footprint can be read in price and volume.
The Illusion of Price Discovery
Price discovery is the process by which trading finds the price where buyers and sellers balance. With half the volume trading in the dark, it's fair to ask where that process still happens.
Mostly, it still happens on the exchanges. US rules stop off-exchange trades from trading through the best displayed prices, so dark pools and wholesalers take their prices from the public quotes: the midpoint, the best bid or offer, or a small improvement on it. By Rosenblatt's count for December 2025, about 37% of off-exchange volume traded exactly at the best bid or offer and 19% at the midpoint. The off-exchange trades borrow the lit market's price and add volume to it.
That makes the concern specific. The fewer orders that rest on public books, the fewer orders are setting the prices that everything else borrows. A displayed quote is the only order that tells the market where someone is willing to trade before they do, and as more volume moves away from displayed quotes, a smaller and more informed part of the market does all the discovering. Whether that has made prices worse is disputed. Exchanges argue it has, and wholesalers and the brokers they pay argue that retail traders get better prices than the public quotes would give them. What isn't disputed is that the price on your screen is set by a minority of the orders and the volume under it may be a minority of the trades.
Reference
| Term | Meaning |
|---|---|
| Lit venue | A venue that displays orders before they trade; in the US, the exchanges |
| Dark pool (ATS) | An alternative trading system that matches orders without displaying them |
| Single-dealer platform | One firm's own liquidity offered electronically to its clients |
| Wholesaler | A market maker that buys retail order flow and fills it from inventory |
| Internalisation | A broker or dealer filling an order against itself instead of sending it to an exchange |
| Payment for order flow | What a wholesaler pays a broker for its customers' orders |
| TRF | FINRA trade reporting facility, where US off-exchange trades are reported within 10 seconds |
| Consolidated tape | The combined US feed of exchange trades and FINRA prints |
| Rule 606 report | A US broker's quarterly disclosure of where it routes customer orders and what it's paid |
| NBBO | The national best bid and offer, the best displayed prices across US exchanges |
| Measure (US equities, by shares) | Value |
|---|---|
| Off-exchange share, 2009 | about 34% |
| Off-exchange share, 2024 | 47.0% |
| Off-exchange share, 2025 | 50.6% (44.7% by value traded) |
| ATS share of off-exchange volume, 2025 | 18.7% |
| Reporting deadline for off-exchange trades | 10 seconds |
| FINRA publication of weekly ATS and dealer volume | 2 weeks later (most active stocks), 4 weeks otherwise |
Case Study: Two Dark Pools on the Record
A dark pool asks its users to trust what they can't see. They can't watch how orders are matched, who else is trading, or what happens to their order information. On 31 January 2016 the SEC and the New York Attorney General settled cases against Barclays and Credit Suisse over two of the largest US dark pools, and the findings show what that trust covered.
Barclays ran a dark pool called LX. It told subscribers that a feature called Liquidity Profiling would continuously police the pool for predatory, aggressive trading, and that it ran surveillance reports every week. According to the SEC's order, it didn't police the pool with the tools it described, didn't run the weekly reports, and sometimes overrode Liquidity Profiling by moving subscribers from the most aggressive categories to the least aggressive. So subscribers who had chosen to avoid aggressive traders kept trading with them. Barclays also said it used direct exchange feeds to deter latency arbitrage, when in fact it used a mix of direct and slower feeds. Barclays admitted wrongdoing and paid $70 million, $35 million each to the SEC and the New York Attorney General.
Credit Suisse ran Crossfinder. It told subscribers that a system called Alpha Scoring sorted their order flow objectively and transparently. The SEC found the scoring had significant subjective elements and wasn't applied to every subscriber each month. Credit Suisse also failed to treat subscribers' order information confidentially, sending it out of the pool to other Credit Suisse systems. Its order router systematically favoured Crossfinder over other venues at certain stages, without telling subscribers. And Crossfinder accepted more than 117 million sub-penny orders, which the rules prohibit. Credit Suisse paid $84.3 million without admitting or denying the findings.
Nothing in the public record of trades would have shown any of this. The prints from LX and Crossfinder looked like any other off-exchange prints. It took the New York Attorney General's lawsuit against Barclays in June 2014, and SEC investigations of both banks, to find out what was happening inside. That's the cost of darkness that doesn't show up in a spread or a fill price: when you can't see how a venue works, you're relying on what its operator says.
Where It Fails
- Dark doesn't mean against you. Dark pools exist to protect large orders from being traded ahead of, and retail fills from wholesalers often improve on the public quote. Hidden doesn't mean harmful.
- Shares overstate the share. Counting by shares, off-exchange trading passed half in 2025; by value traded it was under 45%, because cheap stocks trade in huge share counts off-exchange.
- Volume isn't discovery. Off-exchange trades take their prices from the public quotes. A rising off-exchange share means more trades borrow the lit price, which isn't the same as the lit price being wrong.
- Crypto is different. The US numbers and rules here don't carry over. Crypto has no consolidated tape and no reporting facility, and its OTC trades aren't reported at all.
- Full transparency has costs too. On public blockchains, pending transactions sit in an open queue before they're executed, and bots trade ahead of and behind them. Some traders now route around that queue privately, which is going dark again.
Checklist
- Which venues does your chart's volume include: a consolidated feed, or one exchange?
- For a US stock, how much of the day's volume printed off-exchange?
- Is a large print a block from a dark pool or OTC desk, reported after the fact, rather than an order you could have seen?
- Where does your broker send your orders, and what is it paid for them? Read its Rule 606 report.
- In crypto, which venue's volume are you reading, and was it the venue that led the move?
- If you use a dark venue, what does its operator say it does, and could you check?
Since this was published: Tool 01 and Turn the Candles Off both point here for the volume a single venue's chart leaves out, and Applied Market Mechanics is the practical companion to this series.
Conclusion: The Market's Blindfold
Markets can't work with full transparency. A large order shown in advance would be traded against before it filled, and a dealer whose every quote and inventory change was public couldn't manage its risk. Some darkness is built in, and some of it protects the people trading.
The blindfold isn't worn evenly, though. The wholesaler sees the retail flow, the dealer sees its clients' trades, the pool operator sees every resting order, and the fastest firms see each price change first. The retail trader sees a quote that someone else's orders set, and volume bars that may show only part of the trading behind them. That asymmetry is the last layer this series set out to describe, after time, outcomes, liquidity, structure, derivatives and market makers.
None of it makes trading hopeless, but all of it is a reason to trade knowing what you can't see. Know which venues your data covers, know where your orders go, and treat what the chart shows as part of the market rather than all of it.
This was the last part of Understanding Market Mechanics, which set out to describe how markets work underneath the chart: who is on the other side of a trade, what moves price, and what you can and can't see from where you sit. Each part added a layer, and the layers together describe a market whose advantages are unevenly distributed by its structure. Seeing that clearly is useful in two ways. It makes you a more careful participant, and it makes plain what a fairer market would need: transparency that's real, conflicts that are disclosed, and rules that don't depend on trusting what nobody can check. Applied Market Mechanics picks up from here with the practical side: tools and setups, written so they can be repeated.
Series note: Part 7 of Understanding Market Mechanics. Previous: Part 6 · The Hidden Hand of the Market. The practical companion is Applied Market Mechanics.
Understanding Market Mechanics describes how markets work, published purely as education. Nothing in it is financial advice or a recommendation to buy or sell anything, and I'm not a financial adviser. Every case study is drawn from the public record and has already resolved, and nothing in a part should be read as a statement of what I hold.