Why Prime Brokerage Is Facing a Structural Rewrite

In April 2025, DTCC announced a platform for tokenized, real-time collateral management. In December, it said it was preparing a service for tokenized assets held at DTC. Core market utilities are moving records, transfer rules, and settlement logic onto coded rails.

Prime brokerage grew by bundling services that were hard to link. The broker funded positions, held assets, sourced securities, moved margin, settled trades, and gave the fund one risk view. That bundle turned broken market plumbing into a high-margin client tie.

Tokenization does not remove the prime broker. It changes which parts of the bundle stay scarce. Balance-sheet credit and default support stay valuable. Record matching, reporting, and slow asset movement do not.

The prime broker earns from linked tasks. It lends cash, lends securities, charges swap spreads, holds assets, and manages margin. It also nets trades so less cash and collateral must move.

Its core edge is control of the collateral map. The broker knows which assets are free, pledged, borrowed, reused, or stuck in another account. That view lets it price credit, set haircuts, and choose how much balance sheet to use.

Coded settlement changes that map. A tokenized asset can carry title data, transfer limits, and use rules on one rail. A margin engine can check the asset, lock it, and move it when terms are met. Shared records can cut breaks among the fund, broker, custodian, and clearing venue.

This compresses the work layer. Fewer checks are needed. Collateral can move more often. Idle assets can support funding without waiting for batch windows.

The credit layer remains. A token cannot create a dealer balance sheet. It cannot absorb a client loss, find a hard-to-borrow security, or settle a dispute after default. It can move collateral fast, but it cannot decide who bears a shortfall.

Prime brokerage therefore splits into two stacks. The first is software for records, margin rules, settlement, and asset moves. The second is risk capacity for funding, netting, liquidity, and default control. The first faces fee cuts. The second can gain pricing power.

The legacy model has strong legal and financial control. Prime brokers work within known custody rules, master agreements, clearing links, and default law. They can lend without full pre-funding. They can also manage failed trades and market gaps.

That model is slow because each firm keeps its own books. Data must be matched across brokers, agents, custodians, and clearing systems. Risk and collateral still pass through closed ledgers.

Autonomous systems use shared state. Code can apply margin rules at any hour. It can settle assets against cash and show the same data to approved parties. This improves speed, clarity, and capital use.

The trade-off is control risk. Smart contracts can run a bad rule at machine speed. Price feeds can fail. Networks can split. Legal title may still rest on records outside the chain.

These rails may also need more cash up front. Legacy brokers can bridge timing gaps with their balance sheets. A strict atomic system may reject a trade unless both sides are ready at once. That cuts settlement risk but can raise cash needs at execution.

The stronger design is hybrid. Regulated firms keep credit, custody power, and default rights. Coded rails take over movement and records. The fight is over who controls the bridge.

Capital moves first to systems that cut friction without weakening legal claims. That favors tokenized collateral tied to regulated custodians, clearing firms, and bank money. It does not favor isolated tokens with weak redemption or unclear title.

Prime brokers that open their collateral systems can defend the client. They can use shared rails to cut costs while keeping funding and netting inside the bank. Firms that keep old batch systems will charge for delay that no longer serves a risk need.

Custodians and market utilities can gain leverage. The party that sets token rules can decide which assets count, which networks link, and which firms can move value. That is control at the rail level.

Software firms that only match closed records face pressure. Their margin came from mismatch. As shared records improve, value shifts to access rules, risk terms, and settlement standards.

Capital will still pay for scarce balance sheet. It will pay less for duplicate books and trapped collateral. Revenue therefore moves toward credit, liquidity, and control of the rule layer.

Prime brokerage is not being replaced by a protocol. It is being split into tasks that need a bank and tasks that only needed better software. Funding, netting, and default support remain protected by capital and law. Record matching, collateral moves, and reporting become cheaper shared services.

The new prime broker will not win by owning every system. It will win by controlling credit while linking to the fastest trusted rail. The old bundle is breaking. The risk franchise remains, but the plumbing margin is being stripped out.

Sources: DTCC, Bank for International Settlements, U.S. Securities and Exchange Commission, Financial Stability Board