Whither Central Clearing Counterparties? Counterparty Risk in Stablecoin-Based Securities Settlement

This paper argues that while distributed ledger technology (DLT) can dramatically reduce traditional post-trade counterparty risk through atomic settlement, it cannot eliminate a deeper problem that becomes central in a tokenized world: redemption counterparty risk. That is, even if a DLT system guarantees that ‘what’s on-chain is correct’, it cannot guarantee that on-chain tokens are in fact redeemable for the off-chain assets they claim to represent. As securities settlement migrates onto stablecoins and tokenized real-world assets (RWAs), this redemption risk becomes the critical bottleneck that prevents DLT from fully displacing central clearing counterparties (CCPs). 

The paper begins by contrasting traditional market infrastructure with DLT-based systems. Under the conventional central clearing model, trades go through three phases—execution, clearing, and settlement—mediated by exchanges, CCPs, custodians, and broker-dealers. CCPs novate trades, provide multilateral netting, demand margin, and mutualize default risk through a loss-allocation ‘waterfall’, making them central nodes for both risk management and liquidity. This framework is costly and complex, but it is legally and institutionally mature, and it is specifically designed to manage counterparty and systemic risk.

DLT-based systems promise a very different architecture. By pre-positioning tokenized securities and tokenized cash (often stablecoins) on a ledger and using smart contracts to effect atomic delivery-versus-payment, DLT can remove principal risk between trading parties, collapse reconciliation steps, and potentially allow near-instant or 24/7 settlement. Many policymakers, industry reports, and scholars have therefore speculated that DLT could make CCPs obsolete, since real-time atomic settlement seems to make a central guarantor of trades unnecessary.

The paper’s core claim is that this enthusiasm is overstated because it focuses only on on-chain performance risk and overlooks off-chain redemption risk. On-chain, a DLT system can reliably and automatically transfer tokens. But for most existing stablecoins and tokenized RWAs, each token is just a pointer to something in the real world: bank deposits, securities portfolios, real estate, commodities, or corporate shares. Whether token holders can actually redeem the token at par, access the underlying asset, or enforce their rights is not determined by code but by law, institutions, and off-chain behavior.

The paper develops the concept of redemption counterparty risk as the risk that the issuer of a tokenized asset—such as a stablecoin operator or an RWA tokenization agent—will be unwilling or unable to honor redemption promises. This risk mirrors traditional counterparty credit risk, but it arises even when the on-chain leg of the transaction is flawless. It is rooted in (i) uncertainty about whether reserves exist, are sufficient, and are segregated; (ii) delays or legal constraints in accessing those reserves; and (iii) the legal enforceability of claims to real-world assets.

A substantial part of the analysis is devoted to fiat-backed stablecoins. The paper explains that stablecoins like USDC and USDT depend on multiple layers of trust: the integrity and prudence of the issuer; custodians’ segregation and safeguarding of reserves; and auditors’ attestations, which are backward-looking and not real time. DLT cannot itself verify whether reserves are actually present or unencumbered, nor can it ensure that those reserves are excluded from the issuer’s bankruptcy estate. In many jurisdictions, it is unclear whether reserves would be ring-fenced or pooled with other assets in insolvency, exposing holders to delays and possible haircuts.

The paper further explains that even when laws attempt to protect stablecoin reserves—such as proposals to give stablecoin holders super-priority claims or exclude reserves from the bankruptcy estate—practical issues remain. Automatic stays, litigation over asset characterisation, and the mechanics of court-supervised distributions create time lags. In high-volume, DLT-based securities settlement, such delays could be systemically dangerous, because participants depend on intraday liquidity even when the issuer is ultimately solvent.

The analysis then extends redemption counterparty risk to tokenized RWAs. When a token is supposed to represent a bond, a share, a building, or a commodity, a DLT system cannot guarantee that the tokenholder’s claim is legally valid, senior, or enforceable. Without a robust legal link between token possession and real-world entitlement, tokenized RWAs may have weak or uncertain legal status. Platforms often attempt to patch this gap via trusts, Special Purpose Vehicles (SPVs), or contractual frameworks, but these solutions reintroduce centralized intermediaries and legal complexity, undercutting the vision of ‘trustless’ finance.

Having diagnosed the problem, the paper turns to legal and institutional solutions aimed at making DLT-based settlement robust without simply recreating today’s CCPs in new guise.

First, on the cash leg, the paper argues that central bank digital currencies (CBDCs) are the most direct way to eliminate redemption counterparty risk. A CBDC is a native digital liability of the central bank, not a claim on a private issuer backed by reserves. If cash exists natively on-chain under central bank governance, there is no need for a trust-based oracle to verify reserves and no run risk on a private stablecoin operator. CBDCs thus offer a powerful way to align DLT’s technical finality with legal and monetary finality.

Second, on the securities leg, the paper proposes that corporate and securities law explicitly allow natively digital equity—that is, tokens that are shares, rather than tokens that merely represent shares. If company law recognizes DLT-recorded tokens as the official shareholder register, then no external verification is needed to link on-chain state to real-world ownership. Some jurisdictions, like Delaware, have already taken early steps by allowing blockchain-based stock ledgers, and the paper sees this path as key to avoiding redemption risk for equity instruments and dramatically reducing settlement frictions.

Third, the paper calls for stronger regulation of stablecoin issuers, custodians, and tokenization agents. This includes mandatory use of independent, well-regulated custodians; stringent segregation and ring-fencing of reserves in bankruptcy-remote structures (e.g., SPVs or trusts); enhanced transparency and audit requirements; and resolution regimes that allow regulators to quickly take control of troubled issuers. Examples such as Hong Kong’s legal frameworks and the US GENIUS Act are cited as emerging models, although the paper emphasizes that even these regimes may not fully remove short-term liquidity risk and procedural delay.

Recognizing that legal protections and segregation alone cannot guarantee timely payout, the paper’s most innovative institutional suggestion is a mutualized insurance fund for DLT-based settlement. This would resemble a narrowly focused analogue to the Securities Investor Protection Corporation (SIPC): licensed stablecoin and RWA issuers would make risk-based, prefunded contributions to a central backstop that could provide rapid, capped liquidity when an issuer fails or redemptions are temporarily blocked. The fund would step in to pay tokenholders quickly, then recover from segregated reserves as courts release them. In this design, the fund is aimed at liquidity, not long-term credit support, and is structured to minimize moral hazard and ‘too-big-to-fail’ concerns.

The conclusion of the paper is deliberately ambivalent about the future of CCPs. DLT can indeed eliminate many traditional post-trade frictions and reduce reliance on central clearing for managing bilateral counterparty risk. But unless and until cash and securities exist natively on-chain—via CBDCs and legally recognized digital equity—and unless robust legal, institutional, and insurance structures are built around token issuers, functionally equivalent backstops will still be necessary. In other words, the locus and form of central clearing may change, but the social need for institutions that manage systemic settlement risk will not disappear. 

The full paper can be accessed here.

James Si Zeng is an Associate Professor at the University of Hong Kong