Crypto Exchange Liquidity: What to Keep In-House and When to Partner

The difficult liquidity decisions begin where routine matching stops: buyers outnumber sellers, a quieter pair receives a large order, or the other side of a trade arrives several hours too late. The matching engine may be working perfectly. What is missing is someone willing to take the trade at an acceptable price.

That is why crypto exchange liquidity is not a single choice between building internally and outsourcing. Running the market, committing trading inventory and securing an external execution are different activities. The useful question is which gaps the exchange should finance itself—and which a partner can fill without creating unacceptable costs or dependencies.

Matching orders is not the same as supplying inventory

Internal customer-order matching pairs compatible buy and sell orders already available on the exchange. It does not create a missing seller or buyer. Operating the matching system therefore does not, by itself, mean committing exchange-owned trading inventory.

Matching is not free: the exchange still pays for infrastructure, balance controls, reconciliation and operational support. But those costs are different from the capital and trading risk involved when its own desk becomes the counterparty.

Consider an explicitly hypothetical snapshot. Customers want to buy 10 BTC and sell eight BTC. Assume the orders are compatible at a hypothetical price of $100,000 per BTC, and ignore fees. The matching engine can pair eight BTC of buying with eight BTC of selling. Two BTC of demand—$200,000 at that price—remain unfilled.

Faster matching does not close that gap. Another customer could arrive, the price could move enough to attract a seller, or someone could supply the missing bitcoin. The exchange must decide whether to wait, commit its own inventory or obtain external liquidity.

The snapshot also shows why aggregate volume is an incomplete guide. Balanced daily flows can conceal an intraday funding requirement. Buying pressure in BTC does not neatly offset selling pressure in ETH without another trade. And two BTC requested at once can present a different execution problem from the same amount spread across an hour.

Define the gap by asset, size, timing and acceptable price—not simply by the percentage of orders matched internally. That is the starting point for choosing a crypto exchange liquidity provider or deciding whether an internal desk is justified.

For recurring gaps, decide who should quote

If an exchange needs dependable bids and offers throughout a trading session, the central decision is who will supply those orders and bear the associated inventory risk.

An internal desk buys control with capital and operating capacity

In-house market making means the exchange commits proprietary inventory or funding to place bids and offers. It can set its own pricing logic, position limits and replenishment policy. That control can be valuable when liquidity provision is central to the exchange’s service and recurring flow gives the desk a reasonable basis for managing inventory.

In the hypothetical example, an internal desk selling the missing two BTC needs access to $200,000 of bitcoin at the assumed price. It receives cash in return, but its asset mix changes. If it intends to keep supplying bitcoin, it must replenish or otherwise manage the reduced position. Its wider capital requirement depends on how long replenishment takes and how much additional one-sided demand it is prepared to absorb.

The economic return is therefore not simply the spread collected. It must cover financing, hedging or rebalancing, adverse price movements, systems and staffing. Keep the desk’s inventory, limits and profit and loss distinct from the matching operation: a higher fill rate can be purchased at a trading loss.

An external maker can supply the same book

External order-book market making is different from sending customer orders to another venue. An external firm places bids and offers on the exchange’s own book, where the matching engine can match them against customer orders. The exchange retains its venue and matching function while another firm supplies the quoting activity.

If that firm funds its own inventory, the exchange need not hold the same proprietary quoting stock. If the arrangement instead involves an exchange-funded inventory loan or credit facility, some funding and counterparty exposure remains with the exchange. The agreement—not the label “external”—determines how much capital responsibility moves.

This approach can suit a recurring need for on-book depth without building a proprietary trading desk. But quoted presence is not a promise to fill every size in every condition. The relevant bargain concerns executable size, price and availability, alongside fees or incentives and the conditions under which the maker may reduce or withdraw its quotes.

For execution gaps, distinguish routing from OTC

Standing quotes are not the only way to complete a trade. An exchange may need an execution outside its own order book, particularly when a request exceeds the available depth. Two routes deserve separate treatment.

Routing accesses liquidity elsewhere

With external routing, the exchange sends an order, or an offsetting trade, to another execution venue or source. That can supplement the local book, but it does not automatically add resting orders to it.

The distinction matters operationally. A price visible elsewhere is not necessarily the price available when the routed order arrives. Partial fills, account funding and the movement of assets back into usable balances all affect the result. If the exchange commits to a customer price before securing the external execution, it owns the intervening price and fill risk unless the arrangement explicitly transfers it.

OTC addresses a defined transaction

Crypto OTC liquidity is bilateral execution for an agreed transaction rather than continuous quoting on the exchange’s public book. An exchange can request a price for a specified asset, side and size, then agree the execution and settlement terms with the counterparty.

For a larger or irregular request, a firm quote for the full amount can make the proposed outcome easier to assess than a series of uncertain fills. That certainty applies only to the quote’s stated size, validity and conditions. OTC is not inherently cheaper than order-book execution, and an attractive quote does not remove the need to examine settlement timing and counterparty exposure.

It is also not a substitute for ongoing quoting unless the service has been expressly designed that way. An OTC transaction may fill the two-BTC gap in the example without improving the next customer’s view of the exchange’s order book. It can nevertheless complement that book—or replenish an internal desk after the desk has supplied customer demand.

Compare the cost of a usable fill

These arrangements should be compared against the same job: delivering a specified amount, within an acceptable price range and time. Comparing an internal desk’s gross spread with an external provider’s fee misses much of the economics.

For the internal desk, include the capital committed, inventory outcomes, rebalancing and operating costs. For external execution, include the quoted spread, explicit fees, any market impact, funding costs and settlement or transfer costs where applicable. An external maker may also require commercial incentives or funded balances. Neither internal nor external liquidity is automatically cheaper.

Return to the $200,000 gap. Buying the two BTC externally may remove the need to hold those coins in advance, but it creates a payment obligation under the agreed terms. Even if customer cash covers the purchase economically, that cash may not yet be in the account, currency or settlement rail needed to pay the counterparty. Prefunding can tie up capital; credit can defer payment, but brings its own terms and limits.

Use two views of the result. The customer cares about the effective price, filled amount and certainty. The exchange cares about the contribution left after trading, funding and operating costs, along with the risk retained. A sourcing arrangement must work on both views.

Keep ownership of the service, even when liquidity is shared

Hybrid liquidity sourcing need not be complicated. An exchange might continue matching natural customer flow, use an external maker for selected pairs, and arrange OTC execution for larger requests. Another might operate a limited internal desk and route trades externally when inventory reaches a defined threshold. These are combinations of functions, not mutually exclusive business models.

The useful boundary is explicit: which source handles which assets and sizes, when the exchange commits to the customer, and how much unhedged or unsettled exposure it will accept. Adding a partner without those boundaries can turn an apparent liquidity improvement into an uncontrolled balance-sheet commitment.

Funding and settlement responsibilities must be agreed separately from execution access. Who supplies each asset? Where must it be held? When does payment become due, and when is the received asset available for use? Those answers determine whether a completed trade actually restores the exchange’s ability to serve the next customer.

The exchange also retains operational work. Integration must reliably capture quotes, orders, fills and failures. Monitoring must distinguish a connected system from an executable source of liquidity. Reconciliation must confirm that trades and balances agree, while limits must prevent a delayed settlement or unexpected partial fill from expanding exposure unnoticed.

Continuity deserves the same attention under every model. An internal desk can reach an inventory limit just as an external partner can stop quoting. A fallback should be a tested, approved route—not an assumption that another source will always be available. Depending on the service, the appropriate response may be a secondary source, a smaller supported trade size or a temporary pause rather than an execution the exchange cannot safely fund.

Conclusion

Keep the distinction between matching and supplying liquidity clear. Retain ownership of customer commitments, risk limits and settlement control; decide separately who should provide inventory and execution. The strongest arrangement is the one that fills identifiable gaps at an acceptable total cost—not the one labelled most “internal” or most “outsourced.”

For requesting more information about how we can help reach out to us. We're here to help and answer any questions you may have.

Contact us!