Reviewing the Regulated Settlement Network (RSN) - Part 3
Can the Legal Viability Report from Sullivan & Cromwell Deliver Practical Solutions? Does It Stand a Chance Against My Idea for a UAN (Unregulated Apple Network)?
I wrote about the Regulated Settlement Network (RSN) previously, and since SIFMA published three reports, this article covers the third one: the legal viability report.
“Aller Guten Dinge Sind Drei,” which translates literally to "All good things come in threes," reflects the idea that three is a number of completeness, luck, or success, often implying that the third attempt at something will succeed if the first two failed. It is often used optimistically, suggesting that good outcomes or successful attempts often align with the number three. Let’s see if Sullivan & Cromwell can deliver, as they are the authors of the legal analysis.
Although they call it something else:
“A high-level discussion based on our understanding of the structure [which] does not reflect all of the considerations that would be relevant in the design of an operational system.”
Lawyers often phrase things this way when they don’t want to be liable for the content. But documenting a “high-level discussion,” which is essentially what this document is, seems like one of the strongest ways to distance themselves from the conclusions.
The Legal Viability Report sets the stage with this description:
“The concept explored in the PoC was a regulated FMI that would utilize a private, permissioned, shared ledger system and tokenized central bank deposits, tokenized commercial bank deposits and tokenized securities (including U.S. Treasury securities and investment grade (“IG”) bonds issued by a U.S. corporate issuer) for settlement of payments and securities transactions between regulated financial institutions for their own benefit and on behalf of customers.”
So, a private, permissioned, shared ledger system that organizes assets and facilitates settlement between participants, including for the benefit of their clients.
The Technical Feasibility Report adds to this with its definition of the Shared Ledger Technology:
“Digital system for recording the transaction of assets in which the transactions and their details are recorded in multiple places at the same time.”
And for tokenization:
“Process of converting rights to an asset into a digital token on a blockchain, where each token represents ownership or a share of the underlying asset.”
Back to the Legal Report:
“The tokens used in the RSN would be issued by the various RSN Members, and different Members may issue different varieties of tokens or use their Partitions (as defined below) in different ways. There would be no one RSN token of any type (i.e., a ‘deposit token,’ ‘governance token,’ or ‘security token’) issued by the RSN FMI or anyone else. Accordingly, our analysis addresses only the contemplated use of the RSN Members’ tokens within the RSN and does not consider any other use of the same tokens or Partitions.”
And here’s the key point about the RSN Ledger:
“The system evaluated in the PoC would include an RSN FMI entity (the “RSN FMI”) that would operate a private, permissioned, shared ledger (the “RSN Ledger”) to effect settlement of transactions between RSN Members.
Each Member would maintain an individual ledger within a separate partition (its “Partition”) recording its balances, and each Member would have access to the information on the RSN Ledger pertaining to transactions in which it is involved.
Partitions would be updated based on the RSN Ledger, which could be done automatically, but the Members ultimately control their own Partitions.
Each RSN Member would have control over any transactions or other updates that result in updates to its ledger by virtue of its right to approve or disapprove each transaction and authorize or refuse to authorize any transaction.
The RSN Ledger would be the definitive record of the deposits and securities positions held and transferred through the RSN.”
I think it’s important to clarify that by the term “positions held and transferred through,” it’s missing the word “by” — i.e., “positions held [by ?] and transferred through.” It took me a while to get my head around this. The assets are held by each member (on their partition run on RSN software), but since RSN maintains a book of records of that, their ledger prevails. Because the system forces a member to make updates synchronized with the RSN ledger, it should not, in theory, make a difference.
That assumption can, of course, be made, but I think it’s highly unlikely anyone would agree to this once they fully understand what it actually means.
The RSN's Hybrid Approach to Asset Management
The RSN’s design presents a hybrid approach to asset holding and recordkeeping:
Members hold the assets on their partitions.
The RSN Ledger serves as the definitive record, ensuring synchronization and finality.
This approach attempts to combine the benefits of decentralization (member control over assets) with the advantages of a centralized system (a single source of truth and clear finality). However, the effectiveness and legal robustness of this hybrid model depend on addressing potential challenges related to external systems and the enforceability of the RSN Rulebook. And not only that—the bank would effectively give up control over updates to its ledger. For example, making a manual update on the system that creates a conflict with the RSN ledger (e.g., deleting a credit from a transaction processed via RSN) would not be possible. This promises some nirvana benefits but it is very unlikely to be agreeable in this form.
The RSN enforces synchronization with the RSN Ledger:
“Each RSN Member (including the Federal Reserve Bank) would update its Partition (which could be done automatically by the RSN FMI following any update to the RSN Ledger) and reconcile its books and records.”
This suggests the RSN system is designed to prevent conflicting updates and maintain consistency across all records. But since it’s unlikely that the RSN would cover all securities positions held by a bank, it’s highly doubtful that a bank would see this system as its book of records. It would probably replicate the RSN holdings in its existing core banking systems.
And if the idea is that all banks switch to the same core banking system to create efficiency, this will never happen because the cost and complexity of implementing such a change would be enormous. Even with hundreds of millions of dollars in investment, successfully delivering this outcome is far from certain.
The legal document supports my skepticism, so thank you Sullivan & Cromwell
It mentions the possibility of the RSN integrating with existing systems but doesn't elaborate on the feasibility or implications for core banking systems.
It focuses on specific use cases and suggests that the RSN might not be suitable for all types of assets and transactions.
It highlights that "the PoC focused on the features of the RSN concept rather than finding the optimal technology to support it," suggesting the current design might not be scalable or comprehensive enough to replace existing core banking systems.
A thorough cost-benefit analysis is crucial to assess the actual cost savings and efficiency gains the RSN can offer compared to existing systems. The sources do not provide such an analysis, leaving a gap in understanding the economic viability of the system. But in my experience, it’s so far-fetched from practical realities that it can never work. And no cost-benefit analysis would convince me otherwise to be fair, or to be more precise, a positive assessment would be based on implausible assumptions.
Critical Questions That Remain Unanswered
The feasibility of reconciling with external systems.
The enforceability and clarity of the RSN Rulebook.
The system's adaptability to diverse participant needs and complexities.
The concrete cost-benefit analysis justifying its implementation.
The legal analysis claims that the RSN could potentially work under UCC 8, and I’m not doubting that. But that’s not the relevant question.
The UAN Analogy
Oh what is that? It’s my new concept. The purpose: describing the arrangement they have in mind in simpler terms:
For whatever reason, my neighbour and I decided to set up a UAN (Unregulated Apple Network), which is like this. I consider myself a financial intermediary and open an account to record apple entitlements for my neighbor. We don’t really interact other than standing in the garden, which is segregated by a huge wall, so I can’t see what’s going on on the other side.
We start by throwing an apple over the wall, and some voice that I assume to be my neighbor says, “Yeah, got it.” I still owe my neighbor an apple that I borrowed the day before. But now he wants more apples, and I tell him he would need to give me the money to purchase them on his behalf. He says he already went to Tesco and bought a pack of 10 apples, and they keep them there for him. If I now record that in my security account, UCC 8 would apply, governing the relationship between me (the intermediary) and my neighbor (the entitlement holder). It focuses on how I, as the intermediary, must treat his entitlement to the assets I hold for him.
But this presupposes that I have control over these assets. It relies on book entries to indicate entitlement but can't govern actions that occur outside UAN. For example, if I go to Tesco and say my records show that my neighbor purchased 10 apples, and I grab a bag of apples without paying, claiming that the UAN records show I can take them, Tesco would likely think I’m about to commit theft. And the police would probably agree with Tesco.
The UAN analogy is just a whimsical analogy and probably has its own flaws. The point I want to emphasize is that achieving true "finality" in a system like the RSN requires a seamless and legally enforceable connection between internal records and the real-world systems controlling the underlying assets.
And since the concept doesn’t have this, it provides very doubtful benefits. In essence, while the legal document shows that RSN is not illegal, it doesn’t make a compelling case that it’s worthwhile, practical, or implementable.
So doing what is required to make RSN work would be so complex and costly that I bet my month’s supply of Tesco apples that I’m right.