Distributed ledger technologies (DLTs) have garnered growing interest in recent years and are making inroads into traditional finance. One purported benefit of DLTs is their ability to bring about “atomic” settlement. Indeed, several recent private sector projects (SDX, Fnality, HQLAx) aim to do just that. But what exactly is atomic settlement? In this post, we explain that atomic settlement, as it is often defined, combines two distinct properties: instant settlement and simultaneous settlement, which should be kept separate.
Settlement Today
Whenever two financial market participants agree to trade an asset, the act of transferring the ownership of the asset from the seller to the buyer, and the associated payment, is called the settlement of the trade. In traditional financial markets, trading and settlement are separate processes. For instance, markets commonly practice a two-day delay between the time a trade is agreed to, and the time the trade is settled. This delay is not driven by technological limitations. Some markets settle one day after a trade and, in some rare cases, even on the same day. Regardless, delays mean that trades do not always settle as planned. A settlement failure can occur either because the seller is not in possession of the security that must be delivered or because the seller chooses to strategically fail (see this Current Issues article by Fleming, Keane, and Garbade, and this Liberty Street Economics post by Fleming and Keane). Such outcomes give rise to a so-called replacement-cost risk.
Another potential issue with current settlement practice is that both legs of a transaction are not always settled at the same time. This creates the risk that one leg settles without the other, often called principal risk. A canonical example is the case of Herstatt Bank, a German bank active in foreign exchange (FX) markets that went bankrupt in 1974. German regulators closed the bank after it had received payment of Deutsche Marks from its counterparties but before it released the U.S. dollars that it owed, leaving the counterparties with a loss.
Financial markets have adopted solutions to mitigate these issues. The risk of settling one leg of a transaction before the other is addressed by settling on a “delivery versus payment” (or DvP) basis. This means that one leg of the transaction settles if and only if the other one does. In the context of FX transactions, the same idea is called “payment versus payment” (or PvP).
Settlement failure charges incentivize market participants to fulfill their trading obligations (see, for example, Garbade et al.). Furthermore, third-party intermediaries and platforms facilitate the orderly settlement of transactions (an example in the context of FX transactions is CLS). Margin requirements and other uses of collateral can also provide incentives for traders to fulfill their promises (Martin and Mills discuss the incentive role of collateral). Traders can also build long-term relationships and a reputation for credibility.
What Is the Future of Settlement?
DLT-based projects aspire to improve settlement practice along two dimensions. First, they aim to eliminate any time gap between trading and settlement, so that settlement happens immediately once a trade has been agreed upon. We refer to this as “instant settlement.” One potential benefit of instant settlement is that it would eliminate settlement fails, since traders would only be able to engage in a trade if they are able to settle it immediately. Another benefit could be that central counterparties are no longer needed.
Second, DLT projects also aim to provide DvP (or PvP) functionality. We refer to this property as “simultaneous” settlement. In the context of DLTs, simultaneous settlement extends the idea of DvP as it potentially allows for the settlement of all legs of multiple linked transactions between multiple parties simultaneously. Importantly, the settlement of any given leg of the transaction is conditional on the settlement of all the other legs. However, the settlement could occur sometime after the trade has been agreed, as in current arrangements. Simultaneous settlement doesn’t guarantee that settlement takes place, but it assures that there is no imbalance between the counterparties; if one counterparty doesn’t settle its side of the trade, the other will not either.
The term “atomic settlement” is sometimes used to refer to settlement that is both simultaneous and instant (see, for instance, this blog post and this article, among others). Conflating both properties into atomic settlement can muddle the discourse on the future of settlement. We believe that it is more useful to define “atomic” settlement as being equivalent to “simultaneous” settlement, as in this DTCC press release. Indeed, not only are “instant” and “simultaneous” two logically distinct properties, but while simultaneous settlement is probably always desirable, instant settlement may not be.
Why Is Instant Settlement Not Always Good?
Instant and simultaneous settlement resembles the process of paying physical cash for goods, in which case the trade (of goods) and settlement happen together and in real time. This has the benefit of completely eliminating settlement risk. However, instant settlement also comes with some unintended consequences. First, it can significantly restrict the set of permissible trades. For a trade to be instantly settled, all legs of the transaction must be “settle-able” at the moment the trade is executed, which makes netting of settlement obligations impossible. Whether a trade involves securities or cash, this rules out a significant portion of today’s financial activity in which traders enter trades with the expectation of obtaining the security or the cash necessary for settlement later. In other words, only trades in which cash and securities are pre-positioned can be executed.
In addition to the greater liquidity burden, instant settlement can fundamentally alter the information environment in which traders operate. The decoupling between trading and settlement enables traders to negotiate trades without revealing any information about their past activities. With instant settlement, traders can only sell securities they already hold, which reveals information about past trades. This can result in various issues, including a hold-up problem, as identified in this paper.
To Sum Up
New technologies have the potential to modify the way financial transactions are settled. DLT platforms may allow for an expanded set of DvP settlement, for instant settlement, or for both at once. While instant settlement may be desirable for some trades in some markets, such settlement may not be desirable in all cases as it may come with drawbacks.
Regardless of what the market decides regarding the benefits of instant settlement, this post argues that it seems desirable to keep separate the concepts of instant and simultaneous settlement, rather than combine them into one definition. In our view, it is more useful to define atomic settlement as being equivalent to simultaneous settlement.
Michael Junho Lee is a financial research economist in Money and Payments Studies in the Federal Reserve Bank of New York’s Research and Statistics Group.
Antoine Martin is the financial research advisor for Financial Stability Policy Research in the Federal Reserve Bank of New York’s Research and Statistics Group.
Benjamin Müller is a policy analyst at the Swiss National Bank.
How to cite this post:
Michael Lee, Antoine Martin, and Benjamin Müller, “What Is Atomic Settlement?,” Federal Reserve Bank of New York Liberty Street Economics, November 7, 2022, https://libertystreeteconomics.newyorkfed.org/2022/11/what-is-atomic-settlement/.
Disclaimer
The views expressed in this post are those of the author(s) and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the author(s).
RSS Feed
Follow Liberty Street Economics
