From Bilateral Exposure to a CCP Hub
This chapter opens the central counterparty layer of the U.S. Treasury market and follows what happens after a cash or repo trade is submitted for clearing. The focus is not whether a CCP sounds safer than a dealer, but which legal and financial relationships change at matching, novation, margining, netting, settlement, and default. That distinction gives readers a way to judge the reform’s balance-sheet benefits without confusing a guarantee of performance with a guarantee of liquidity.
What Novation Changes
A central counterparty, or CCP, changes the identity of the party standing behind a cleared trade. In a bilateral transaction, the buyer and seller remain exposed to each other until the obligations are completed. In a centrally cleared transaction, the CCP receives the trade details, matches them, and novates the contract. Novation means that the original buyer and seller no longer rely solely on each other as counterparties. The CCP becomes the buyer to every seller and the seller to every buyer, while the original parties retain their obligations to the CCP through their clearing relationships.
The trade itself does not vanish. Its Treasury security, price, quantity, maturity, and settlement obligation remain. What changes is the risk connection around it. A buyer does not acquire a different Treasury security merely because the trade is cleared. Instead, the buyer exchanges direct exposure to the seller for exposure to the CCP’s rules, resources, margin process, and default-management system. The seller makes the same substitution in reverse. This is why central clearing can make a security less “informationally sensitive”: participants need less private information about the specific counterparty’s ability to perform, although they do need information about the CCP and the member through which they access it.
The change usually happens soon after execution, not only at final settlement. Before novation, a dealer, fund, bank, or principal trading firm has the ordinary exposure created by an unsettled trade. Once the CCP accepts and novates the matched transaction, it can apply one risk methodology across the position. It can collect margin, monitor concentrated exposures, and calculate the net cash and securities obligations that each participant must settle. The original trade remains economically meaningful, but its counterparty-risk management is centralized.
The lifecycle has an important asymmetry. The CCP guarantees performance through its structure and resources, but it does not promise that the market price will remain stable or that cash will always be easy to obtain. Margin protects the CCP against the risk that a member’s positions lose value between the last margin collection and closeout. Netting reduces the amount that must move at settlement. Default management coordinates what happens when a member fails. None of these functions makes Treasury prices immune to selling pressure or makes every participant a source of funding.
Netting is the balance-sheet engine of the model. Consider three firms with three bilateral purchases in the same settlement cycle: Firm A buys $90 million of securities from Firm B; Firm B buys $80 million from Firm C; and Firm C buys $100 million from Firm A. The gross obligations total $270 million. Each firm would have to support its own bilateral delivery obligations, even though the trades form a circle. If the trades are centrally cleared and the CCP can net the positions, the source’s example reduces the collateral needed to support them to $40 million: $20 million for Firm C and $10 million each for Firms A and B.
The example is stylized, but the mechanism is not. A bilateral network records several promises even when the promises offset across different counterparties. A CCP can observe those positions together and replace them with smaller net obligations. That can reduce the cash and securities that have to move, the collateral needed to support unsettled trades, and the operational opportunities for a delivery failure. It can also reduce the balance sheet needed to intermediate a given amount of activity, particularly when accounting or capital rules do not permit a dealer to net repos with different legal counterparties outside a CCP.
Repo makes this balance-sheet effect especially important. A dealer can borrow cash in one repo and lend cash in another, or finance and distribute Treasury positions across several clients. If the transactions are not with the same legal counterparty under an arrangement that permits netting, the notional value of the positions can expand the dealer’s balance sheet even when the dealer is economically close to matched. Once eligible positions are centrally cleared, the dealer can offset qualifying repo and reverse-repo obligations across the clearing network. The value is not that the dealer stops intermediation; it is that more of the intermediation can be measured and settled on a net basis.
The SEC summarizes estimates that broader central clearing could reduce dealers’ daily cash-market settlement obligations by as much as 60% in the periods around the March 2020 disruption and by as much as 70% during the disruption itself. It also cites an estimate that central clearing of dealer-to-client repos could reduce an example of dealer exposure by more than 80%, from $66.5 billion to $12.8 billion in 2015. These are estimates based on particular assumptions and market data, not guaranteed savings for every dealer or every trade. The actual benefit depends on which positions enter the same netting set, how accounts are structured, and whether participants continue to intermediate effectively after costs are taken into account.
The Efficiency Gain Comes with a New Dependency
The most consequential account question is what happens when a dealer clears a trade for a client. In the source baseline, FICC’s existing practice generally allowed a direct participant’s proprietary trades to be margined on a net basis with its customer trades, except for positions submitted through the sponsored-member program. A dealer’s purchase from a customer could therefore offset a dealer sale elsewhere in the dealer’s account without generating a separate customer margin requirement at the CCP. That arrangement reduced the apparent cost of clearing, but it also meant that the CCP had less visibility into the customer position and that the customer’s trade could be entangled with the dealer’s own risk.
The SEC amendments change that relationship. A Treasury CCA must have policies designed to calculate margin for positions submitted on behalf of indirect participants separately from margin for a direct participant’s proprietary positions. Customer margin must be collected separately and held in a separate account. The goal is not to prohibit netting. It is to put netting in the correct account. The dealer’s proprietary side can still be netted against the dealer’s qualifying positions with other market participants, while customer positions are treated as their own source of risk and collateral.
That separation serves two purposes. First, it lets the CCP see where the risk comes from. A direct participant that looks safe after combining its house trades with customer trades may have a very different risk profile when the customer positions are examined independently. Second, it protects the customer if the direct participant defaults. The CCP can manage the defaulting member’s proprietary portfolio without automatically treating customer margin and positions as resources for the member’s own obligations. The improvement in protection can come with a higher or more explicit collateral requirement for the client relationship, which is why the economics of client clearing matter as much as the legal rule.
The companion change to the broker-dealer customer-reserve formula addresses part of that cost. Under conditions, margin required and deposited at a Treasury CCA can be recorded as a debit in the customer-reserve calculation. In practical terms, a broker-dealer can use customer resources to meet CCP margin generated by those customers’ trades rather than financing the entire requirement with proprietary funds. The change is intended to lower the cost of offering clearing while continuing to protect customer assets. It does not require every broker-dealer to use the option, and it does not make the margin requirement disappear.
These account rules illustrate a general principle of central clearing: a netting benefit is only real if the system is allowed to see the positions that should be netted and keep separate the positions that should not be mixed. If a dealer’s customer trade is invisible or bundled with house risk, the CCP cannot manage the exposure cleanly. If all customer positions are treated as wholly gross with no workable funding model, the cost may discourage access. The architecture must therefore balance three objectives that can pull apart: reduce unnecessary gross obligations, protect customers from a member failure, and charge the party whose activity creates the risk.
Margin is the CCP’s first line of financial defense. It is not a fee for using the market; it is collateral against the potential loss that could arise before a defaulted position is closed. A CCP can require initial margin for potential future exposure and variation or other margin as prices move. The precise model is set in the clearing agency’s rules, but the economic logic is stable: a participant with a larger, more volatile, more concentrated, or more difficult-to-liquidate position generally consumes more risk capacity.
If a member defaults, the CCP does not simply ask the market to improvise. The defaulting member’s margin and other resources are used first, together with proceeds from liquidating the member’s portfolio and other available contributions. The CCP may then use its own corporate contribution and resources supplied by nondefaulting members through the loss-sharing structure. This sequence is commonly described as a default waterfall. It creates predictability about who bears losses and gives the CCP a coordinated process for closing out or transferring positions.
A Treasury CCP must also manage liquidity, not only credit losses. Imagine a defaulting member with a net long position in a Treasury security that has not yet settled. The CCP may need cash to complete the purchase for the nondefaulting counterparty before it can liquidate the securities. If many obligations have to be completed at once, the amount of cash needed can be much larger than the eventual market-value loss. The SEC describes FICC as subject to a cover-one liquidity standard and able to use resources such as the Capped Contingent Liquidity Facility, or CCLF.
The CCLF is a rules-based commitment rather than a permanently prefunded pool. In a declared event, members can be required to enter into repos with FICC up to member-specific limits, effectively supplying financing while the defaulted portfolio is closed out. The arrangement gives the CCP a predetermined way to fund settlement, but it also gives members a liquidity obligation that they must plan for. A clearing reform can therefore free balance-sheet capacity in ordinary intermediation while imposing a new contingent demand for cash in a severe default.
This is the point at which a simpler network becomes a more concentrated one. Bilateral markets distribute counterparty decisions across many relationships. A CCP replaces much of that web with a hub-and-spoke structure governed by common rules. The hub can see positions, collect margin, net obligations, and coordinate a default more systematically than scattered bilateral contracts can. But the market now depends more heavily on the hub’s models, systems, liquidity resources, governance, and recovery plans. In the source baseline, FICC is the sole provider of Treasury clearing services and has been designated systemically important. A technical or liquidity problem at that institution would therefore matter to a larger share of the market than a problem at one ordinary dealer.
The SEC argues that supervision, daily testing of resources, operational-resilience requirements, and recovery and wind-down planning mitigate this concentration. That is a risk-management framework, not a claim that concentration is harmless. The distinction matters. Central clearing can reduce contagion from a dealer’s failure by containing and coordinating the default, while increasing the importance of the CCP’s own failure paths. Its success depends on both sides of the substitution: the CCP must be more transparent, better collateralized, and more operationally resilient than the bilateral network it replaces in the relevant risk dimensions.
A CCP hub therefore changes the Treasury market in a precise way. It replaces many direct counterparty links with a centrally managed obligation, makes netting and margin more systematic, and creates a defined process for a member default. It does not replace the dealer’s execution function, guarantee abundant funding, or make every customer a direct member. The next design question is how participants reach this hub and whether they can trade with one another without having to use the same institution as both their execution counterparty and their clearing sponsor.