The story begins in what Brandon Mulvihill cheerfully calls the “financial metropolis” of Stockton, California. The joke is his own. Stockton sits roughly an hour from Sacramento, and Wall Street does not often go looking there for its market-structure specialists. Mulvihill grew up in Northern California, played basketball, studied economics at UC Davis, and arrived in finance through the retail end of foreign exchange. It was an unglamorous entrance to an unusually useful education.
Today he runs Crossover Markets, the company behind CROSSx, an execution-only electronic communication network for institutions trading digital assets. The phrase is dense enough to frighten a dinner party. The idea beneath it is clean: let one business execute the trade while other businesses handle custody, credit, and brokerage. Each part has a job. Fewer conflicts get to hide in the bundle.
That view came from watching foreign exchange mature up close. Crypto may wear louder clothes, but to Mulvihill its trading mechanics look familiar. As he likes to put it, “Crypto crosses borders the way currencies do and the way equities do not.” His career has been a long study of what happens after an asset crosses those borders: who extends credit, where liquidity gathers, how quickly an order moves, and whether the venue must earn the next trade.
The apprenticeship inside the machine
Mulvihill joined FXCM in 2005 and started in retail sales. He became a team leader, then a regional manager, then ran the San Francisco office. He moved to New York around 2008 and to London in 2010. There, he began building an institutional operation on top of the scale FXCM had developed serving retail currency traders.
The operation became FXCM Pro. It distributed liquidity to hundreds of retail brokers around the world and included market data, introducing-broker, and prime-of-prime businesses. Mulvihill's company biography credits him with starting the wholesale business, launching the Prime of Prime offering, distributing the FastMatch ECN, and managing FXCM's acquisition of Tradestream, Citibank's margin FX business. In other words, he did not merely sell access to a market. He learned how the rooms connected.
In 2017 he moved to Jefferies to build out FX prime brokerage. Credit, he came to believe, was the lifeblood beneath trading. By the time he left in March 2022, the operation was clearing roughly 300,000 to 400,000 trades a day. Jefferies had also given him a front-row look at institutional expectations: a client might custody assets in one place, borrow through another relationship, and execute across several venues. Freedom to move was normal.
“I’m a market structure guy.”Brandon Mulvihill
A winter job
Mulvihill co-founded Crossover Markets in 2022 with two colleagues whose experience filled out the design. Anthony Mazzarese, now chief commercial officer, had worked with him at FXCM and Jefferies. Vladislav Rysin, the chief technology officer, had built trading systems at Credit Suisse, co-founded FastMatch, and later served as CTO of Euronext FX. Their shared wager was that institutional crypto would eventually demand the separation, speed, and precision familiar in mature markets.
Their timing had comic severity. Crypto winter arrived as they began. Prices dropped, confidence cracked, and the easy money went elsewhere. Mulvihill later described assembling the development team “at the onset of crypto winter” and launching CROSSx in the dead of it. The downturn did not cancel the structural problem. If anything, it made the problem harder to ignore.
holds assets
executes orders
provides credit and netting
Most crypto exchanges historically wrapped several functions together. The venue executed the order, held the assets, and often stood close to other risks. That bundle offers simplicity, particularly to smaller customers. For a large institution, however, it can concentrate exposure and trap liquidity inside separate pools.
CROSSx was designed to occupy the middle box and stay there. It matches buyers with sellers but does not custody assets or trade as principal. A client can buy bitcoin through CROSSx and later sell elsewhere. Mulvihill's blunt summary is also a useful business principle: “I have to win your business.” If the venue does not offer the right price, liquidity, and performance at that moment, the client can leave.
The scoreboard gets less theoretical
CROSSx debuted in 2023. Its first full calendar year, 2024, produced $13.5 billion in matched notional value across 3.14 million trades. The company said 99 percent of those trades were matched in single-digit microseconds. Those numbers gave the founders something more persuasive than a diagram: behavior under real load.
Company figures reported in the March 2026 Series B announcement. Not a measure of company revenue or assets held.
A $12 million Series A followed in June 2024, led by Illuminate Financial and DRW Venture Capital. In 2025 the company added infrastructure in the NY4 data-center hub and opened CROSSx access to US-domiciled institutions. CROSSx 2.0 introduced a new architectural framework, advanced order types, dark posting, and more sophisticated smart order routing.
Mulvihill described a small team shipping twice a week without taking the venue down. In October 2025, he said the company had only 13 people, even as the platform approached 100 institutions. Smallness here was partly a product of boundaries. Crossover was not building a custodian, accepting client deposits, or becoming a broker. It could put its attention on execution, then keep putting it there.
Tradeweb crosses the line
The largest institutional endorsement arrived in March 2026. Tradeweb led a $31 million Series B that valued Crossover at $200 million. DRW Venture Capital, Ripple, Virtu Financial, Wintermute Ventures, XTX Markets, and Illuminate Financial also participated. The deal paired capital with distribution: Tradeweb planned to give its global clients access to CROSSx spot-crypto liquidity through algorithmic order routing.
By then, CROSSx had matched more than $50 billion across 12 million trades and supported nearly 100 live participants. The new money was earmarked for technology, global operations, and deeper institutional integrations. For Tradeweb, the collaboration was an entry into institutional crypto. For Crossover, it connected a young venue to a network whose clients already understood electronic execution across other asset classes.
The connection also completed a neat loop in Mulvihill's career. Crossover's name reflects the original intention: carry expertise from traditional finance into crypto, then carry the useful innovations of 24/7 digital markets back into traditional assets. Tradeweb made that crossover literal.
The operator outside the order book
Mulvihill's public persona rarely wanders far from incentives. Even a light podcast question about inventing a metaverse turned into a meditation on alignment. He imagined a safer digital environment where people could surface conflicts between employees and employers, test difficult conversations, and ask awkward questions. At Crossover, he said, employee equity was one way to align interests.
The detour was revealing. His favorite subjects are systems in which participants want different things, and his instinct is to redesign the setting until useful behavior becomes easier. An ECN and an imaginary metaverse are distant cousins, perhaps, but both are rooms arranged around incentives.
Away from the matching engine, the biography gets warmer and less electronic. Mulvihill calls himself a sports guy and a family guy. He grew up playing basketball and still shoots with his son. On afternoon calls, particularly in summer, colleagues may hear the bounce of a ball. He remains loyal to the Sacramento Kings and considers their historical record a deeper source of pain than the New York Knicks. He named Blues Traveler as a favorite band and Braveheart as a favorite film. None of this predicts microsecond latency. It does suggest a tolerance for repetition, long seasons, and improbable comebacks.
“Clear separation of duties is fundamental to market structure.”Brandon Mulvihill, March 2026
What he is still building toward
Mulvihill's ambition extends beyond making a quicker matching engine. He wants CROSSx to become a primary venue in institutional spot crypto, supported by familiar custody and prime-brokerage relationships. He has argued that bank custody matters as much as regulation because large institutions want their digital assets held by names they already trust with other assets.
Tokenization is also on the map. Crossover has explored how execution and net settlement might connect with the Canton Network and other institutional blockchain projects. Yet Mulvihill's order of operations remains disciplined: establish the spot market first, prove the execution layer, then use it to support tokenized assets and adjacent markets. The future can wait its turn in the queue.
In May 2026, Crossover launched CROSSx Disclosed, connecting institutions with more than 30 over-the-counter liquidity providers through a disclosed bilateral model. It sits alongside the anonymous venue and broadens the ways institutions can arrange liquidity without turning Crossover into the custodian, broker, and principal all at once.
Four years after the difficult winter, Mulvihill is still arguing for restraint in an industry fond of maximalism. Do one job. Measure it in microseconds. Let clients keep their freedom. Make the venue compete again tomorrow. The most interesting feature of his crypto company may be everything it has chosen not to become.