In October 2025, a dislocation event in the digital asset marketplace saw many crypto exchanges fall down and liquidate clients at prices the exchanges themselves determined unilaterally. Because crypto exchanges marry credit and execution, institutions could not manage their risk as they do in other asset classes.
This unfortunate – albeit predictable – event has led to a marketplace that is rapidly maturing as institutions are migrating to the OTC market structure model that separates duties between credit, clearing and execution, explains Brandon Mulvihill, co-founder & CEO of Crossover Markets.
“This migration is more analogous to the global foreign exchange model,” he says. “As this migration continues to unfold, we are having more and more advanced conversations around spread compression, best execution and market impact. However, there is still much room to go before the digital asset marketplace is on par with either equities or FX.”
The reason why the traditional vertically integrated crypto exchange model is becoming less attractive to institutional participants can be attributed at least in part to the cost to trade. In all asset classes, the institutional market mandates separation of duties between custody, credit and execution because institutions have the size, sophistication and resources to manage those areas independently in exchange for compression in fees.
“Notably, institutions demand a compression in the costs of execution and costs of capital and in return these institutions trade significant amounts of volumes,” says Mulvihill. “Fee compression allows for a diversification of client types and a diversification of time horizons in trading strategies which in turn fuels a robust ecosystem.”
Retail client simplicity
Historically, the crypto market was built on the back of retail demand. The supply chain to meet retail clients is what is called a brokerage model or in crypto a vertically integrated model, which Mulvihill describes as a fancy way to describe a business model that offers simplicity to retail clients.
“Retail consumers can open an account, collateralise it and trade without the need to worry about having to manage custody, credit and various execution relationships independently. Retail clients get simplicity in exchange for paying high fees.”
“Institutions demand a compression in the costs of execution and costs of capital and in return these institutions trade significant amounts of volumes.”
Brandon Mulvihill
As crypto grew, these same crypto exchanges built an ‘institutional’ offering which was the same offering, a vertically integrated model. Like retail, the vertically integrated model is attractive to many mid-market and small-sized institutions who want simplicity in exchange for paying high fees, which is why some institutional crypto exchanges talk about having thousands of institutional users.
“The reality is that the crypto exchange business models are not built to scale tier-one institutional trading,” adds Mulvihill. “The market is maturing, opening a new layer of institutional trading from the largest institutions on Wall Street and those companies demand separation of duties and significant reductions in costs to trade.”
David Olsson, chief commercial officer, EDX Markets, notes that institutions increasingly expect transparent price discovery, competitive pricing, deep liquidity and reliable execution and that as institutional participation grows, consistent market quality will become an increasingly important differentiator for digital asset venues.
“Institutions are accustomed to execution, clearing, custody and other critical functions having distinct responsibilities and controls,” he says. “As the market matures, we expect this separation of responsibilities to become an increasingly important part of institutional market structure.”
FX has shown how institutions can operate efficiently in a global, highly electronic market with fragmented liquidity and extended trading hours. Crypto takes that further with true 24/7 markets, increasing the importance of resilient infrastructure and sophisticated liquidity, collateral and risk management.
“In addition, the need for primes and clearers to net risk for counterparties becomes obvious as the operational burden multiplies – something that FX has solved,” says Olsson, adding that institutions need to consider the overall quality and economics of execution rather than focusing too narrowly on the displayed spread.
When asked how smart order routing is evolving as liquidity becomes increasingly fragmented across exchanges, ECNs, OTC desks and liquidity providers, he explains that smart order routing increasingly needs to account for depth, execution certainty and the economics of accessing liquidity, not just the best displayed price.
“In addition, just as in FX, you might see SORs triangulate across different markets to create crosses. Providing a single point of access to sophisticated institutional market makers, transparent price discovery and a centralised clearing and settlement workflow reduces the need for institutions to maintain separate integrations across numerous providers, helping minimise operational complexity.”
Olsson believes that prime brokers will play an important role in helping institutions consolidate market access, credit and risk management, as opposed to maintaining separate relationships with every venue.
Market convergence expected
As for whether bespoke liquidity pools and relationship-based trading are likely to become more important than one-size-fits-all order books, Olsson observes that institutional crypto is unlikely to converge on a single execution model as participants have different requirements around size, liquidity, information leakage and workflow.“Transparent central markets will remain important for price discovery, while more tailored execution models can complement them.”
Looking ahead, he expects commodity digital assets to look a lot like FX market structure for global markets franchises in banks, while tokenised assets, retail or wealth-driven products will look more like securities infrastructure.
“Proven market principles such as separation of functions, credit-based trading, liquidity aggregation, central clearing and institutional risk controls are becoming more prevalent,” adds Olsson. “At the same time, 24/7 trading, stablecoin settlement and on-chain infrastructure can introduce efficiencies unique to digital assets, creating a market structure that draws from traditional finance while continuing to evolve.”
“Transparent central markets will remain important for price discovery, while more tailored execution models can complement them.”
David Olsson
In July, LMAX and Standard Chartered executed their first live digital asset prime brokerage trades, specifically applying bank-grade credit, governance and risk frameworks to digital assets.
According to Chris Knight, managing director LMAX Digital, the separation of execution, custody, credit and settlement for reducing operational and counterparty risk is fundamental.
“Separation of duties is one of the basics of institutional market structure,” he says. “Execution should sit in one place, custody in another and credit and settlement should be handled through frameworks clients already understand. That reduces concentration risk and gives institutions clearer control over assets, exposure and settlement.”
Knight observes that this is not about making the market more complicated but rather about making digital assets fit for institutional participation.
“The main lesson the digital asset industry has borrowed from electronic FX is that market structure matters. Electronic FX did not mature because of technology alone. It matured because the market built transparency, liquidity, credit intermediation and reliable post-trade processes around that technology. Digital assets are on the same journey. Institutions need the best parts of digital asset technology to connect with institutional workflows that already work.”
On the question of whether the ECN model is becoming a more appropriate structure for institutional spot crypto trading than the traditional CLOB exchange, Knight reckons the debate is less about ECN versus CLOB and more about how different liquidity models can work together.
“The main lesson the digital asset industry has borrowed from electronic FX is that market structure matters.”
Chris Knight
Enhanced liquidity valuable
Central limit order books provide transparent, efficient price discovery and anonymous all-to-all access to executable liquidity. Bespoke liquidity pools and relationship-based liquidity can also be valuable, particularly for larger or more specialised trading requirements.
“We expect many institutions will use a combination of both depending on their objectives,” he says. “The important point is that whichever model is used, clients should be able to understand how liquidity is accessed, how orders are treated and how execution quality is measured.”
Knight agrees that institutions should not just compare spreads when measuring execution quality and recommends looking at executable liquidity, fill rates, market impact, rejection rates, latency, slippage and post-trade certainty.
“A tight headline spread is not much use if the liquidity is not there in size or the operational workflow creates risk after the trade,” he adds. “Execution quality needs to be measured across the full lifecycle, not just at the point of price discovery.”
Smart order routing is evolving from a price-sweeping tool into a constrained optimiser as liquidity becomes increasingly fragmented across exchanges, ECNs, OTC desks and liquidity providers. That is the view of Francois Lamy, head of corporate strategy at B2C2, who says this is a much greater challenge.
“The first generation asked, ‘who is showing the best price’,” he says. “The current generation has to ask where the trading capital actually sits, what it costs to keep it there and how long it takes to move it, which venue has credit line capacity, what fee tier applies at this volume level and what the expected mark-out is on each source. Price is one input among many.”
The harder technical problem is that the router now blends objects of different types: a resting order on a CLOB is firm and anonymous; a streamed principal quote has a hold time and may reject; an RFQ is a discrete event that reveals intent. Optimizing across these types requires modelling expected all-in cost including rejection and leakage, not comparing quoted levels.
“Routing without a settlement answer just relocates the problem,” says Lamy. “If the router fragments your fills across six venues and six balances, you have optimized the fill and degraded the operational position. The routing logic and the capital management logic have to be in sync.”
He observes that the arrival of true prime brokers has been promised for years in digital assets and is starting to happen in earnest.
“Like in FX, they will play an increasingly critical role, extending credit, enabling cross-margining and netting and helping market participants trade across a fragmented market while minimising the capital drag of pre-funding and the operational burden of margin management,” says Lamy. “As execution spreads compress, financing is where more of the profit pool moves.”
Prime brokerage confusion
The distinction that matters is that much of what is marketed as prime brokerage in this market is custody with connectivity attached, rather than credit intermediation. Real prime brokerage requires balance sheet and legal certainty: enforceable close-out netting, proper documentation, disciplined haircuts on volatile and highly correlated collateral.
“That is a capital and legal business rather than just a technology business, which is why I would expect the credible providers to be banks and well-capitalised non-bank intermediaries rather than the venture funded platforms that attempted this in 2021 with a primary focus on technology,” adds Lamy.
Adam Sporn, head of prime brokerage and institutional sales at BitGo says the key principle is functional and legal segregation, not necessarily using a different provider for every service.
“Execution should be separated from the custody of client assets, with clear controls governing credit, collateral and settlement,” he says. “The objective is separation of risk without fragmentation of workflow.”
“If the router fragments your fills across six venues and six balances, you have optimized the fill and degraded the operational position.”
Francois Lamy
Sporn notes that one of the most important lessons from FX is that execution can remain distributed across dealers and venues while a prime broker centralises credit and post-trade management. Digital asset markets are also adopting multi-venue access, relationship-based pricing, RFQ protocols, smart order routing and transaction-cost analysis.
He distinguishes between ECNs and central limit order books, which are not directly comparable: an ECN describes a type of electronic venue, while a CLOB is a matching mechanism that an ECN may use.
“Institutional spot trading will likely include anonymous order books, disclosed liquidity streams and RFQ or RFS execution. The appropriate model depends on trade size, strategy, tolerance for market impact and the nature of the available liquidity.”
Institutions should evaluate total cost of execution using measures such as implementation shortfall, slippage, market impact, fees, fill rates, latency and post-trade mark-outs. They should also assess performance by venue, liquidity provider, trade size and market condition.
A tight displayed spread has limited value if the liquidity cannot support the intended trade, explains Sporn, adding that smart order routing is becoming an execution and risk-management function rather than simply a price-selection tool.
“A router must consider executable depth, fill probability, fees, latency, counterparty limits and where the institution’s assets or collateral are located,” he says. “The best displayed price is not always the best achievable outcome.”
“Execution should be separated from the custody of client assets, with clear controls governing credit, collateral and settlement.”
Adam Sporn
Sporn says prime brokers can serve as the connective layer between clients and fragmented sources of liquidity since a single prime brokerage relationship can help coordinate market access, credit, financing, collateral, settlement and reporting while providing a consolidated view of risk. Where appropriate credit arrangements are available, this can reduce pre-funding requirements and improve capital efficiency.
“Bespoke liquidity pools and relationship-based trading are likely to play a greater role, particularly for larger trades or in less liquid markets, where information leakage, market impact and certainty of execution matter,” he adds. “However, they should complement rather than replace public order books. Bespoke liquidity must still be measurable, governed and subject to appropriate oversight.”
