Suppose you need to sell a million shares. Finding a price is easy. Finding enough buyers before everyone notices what you are doing is harder. The first few sales can advertise the rest. A portfolio manager can be right about a stock and still lose money getting out of it. Liquidnet built a business around that awkward gap between an investment decision and its execution.
- The job: help institutions find large trades without broadcasting the whole order.
- The machinery: block matching, algorithms, analytics, and human execution specialists.
- The catch: discretion only helps when a suitable counterparty actually exists.
A market hiding inside the paperwork
Seth Merrin and Eric LeGoff founded Liquidnet in 1999. Merrin already knew the software on institutional trading desks: he had helped pioneer electronic order management. Those systems contained something valuable - the orders that fund managers wanted to execute but had not yet completed. The opportunity was sitting in their digital blotters.
Connect those blotters, and a seller at one institution might discover a buyer at another. The two could negotiate anonymously rather than feed an entire order through a public market in small pieces. Liquidnet launched in the United States in 2001 with 38 member firms. The network made existing institutional intentions easier to find.
That distinction still matters. Liquidnet serves asset managers and other institutional investors, including hedge funds. It is infrastructure for the people moving pooled capital, rather than a consumer brokerage app. In September 2026, its Meridian announcement described access for more than 1,100 members.
The bill you never see on an invoice
Trading costs have two faces. One is the explicit commission. The other is the price changing while an order is being executed. Saving a fraction of a percentage point sounds unexciting until the order contains several million dollars. Then a small improvement becomes real money for a fund and, ultimately, its investors.
A ten-basis-point price difference on a $10 million trade. This is an example, not a Liquidnet performance claim.
Liquidnet earns transaction fees. Its US H2O ATS filing describes a range of zero to six cents per share, including bundled services and commission-sharing arrangements. Clients can also elect percentage-based fees. Order type, access method, size, and commercial arrangements affect the bill. That range is a regulatory disclosure, not a universal quote for every product or country.
The practical attraction is avoiding unnecessary disturbance while finding size. The limitation follows immediately: a matching system cannot conjure a willing buyer. If urgency matters more than waiting for a block, a trader needs other execution routes.
Quiet does not mean invisible
Liquidnet's confidentiality promise has a difficult chapter. In 2014, the SEC found that employees outside its dark-pool operation had improperly accessed confidential subscriber trading information and used it in marketing. The conduct included an equity capital markets initiative running from 2009 to late 2011. Liquidnet settled for $2 million without admitting or denying the findings.
The failure was in information access and use. A venue selling discretion must police the boundaries inside its own organization, too. For customers evaluating such a network, who can see an order is a product question as much as a compliance question.
“Trust is earned one decision at a time.”
Mark Govoni, Liquidnet CEO, November 2025
Govoni's words describe the continuing work behind the proposition. More participants can improve matching, but more people and connections also make careful permissions more consequential. Growth and discretion have to be designed together.
The block specialist wants the rest of the order
Liquidnet now offers execution consulting, a global trading desk, analytics, and commission management alongside block trading. Its connectivity includes systems such as BlackRock's Aladdin, Bloomberg, Charles River, and FlexTrade. Existing workflows are a distribution channel: traders can reach the service through software already on their desks.
The competitive set includes Cboe's BIDS Trading, Virtu's POSIT Alert, and Instinet's BlockMatch. These services also pursue large orders and controlled information sharing. Liquidnet's pitch combines its institutional community with execution expertise and a broader set of tools. Anonymity alone is no longer a sufficient distinction.
Meridian, introduced in September 2026, packages algorithms, liquidity access, analytics, and execution support under one name. It supports orders of different sizes and benchmarks including volume-weighted average price. The shift is explicit: Liquidnet wants more institutional algorithmic flow alongside its established block franchise.
October's SuperBlock Targeted Invitations approaches the original problem from another direction. Traders choose recipients for exceptionally large trading interest; every invitation is backed by a live client order. Distribution includes Bloomberg IOIA, the Liquidnet app, and execution management systems. More discovery, with deliberate choices about who receives the signal.
What $575 million bought
TP ICAP completed its acquisition in March 2021. The announced terms were $525 million upfront, $50 million deferred, and up to $125 million contingent on performance. Those are acquisition terms, not venture funding. TP ICAP wanted Liquidnet's buy-side relationships alongside its own dealer connections.
The resulting business has widened. Listed derivatives arrived in 2023. In June 2026, Liquidnet Fixed Income became RealQ, combining its execution platform with Neptune Networks' dealer data. Treating today's company as an unchanged equities-and-bonds dark pool misses that reorganization.
TP ICAP reported £365 million of Liquidnet division revenue and £56 million of adjusted operating profit for 2025. Those divisional figures cover a broader business than the original matching network. They show how far the commercial model has travelled.
There is a useful lesson here for anyone building a marketplace: start where customers already record their intentions, then make an introduction worth accepting. Liquidnet's member-testing hub, Labs, extends that logic into product development. The hard part remains persuading institutions to reveal enough to trade - and giving them reasons to return.