COLUMN | How the industry re-aligns post-trade plumbing for digital assets is foundational for the future. It’s a great enabler if done well, or mud to get stuck in if not — resulting in friction. The Bankers’ Plumber Olaf Ransome reflects on FMI design-build projects that come close, and lessons learned.

Tokenisation has been a slow train coming. What we build now should be built to last: design matters. Right now we are designing both the rails and the railcars for our future financial market infrastructures (FMI).

Central to the design everything is cash, or the means of payment (P). It is the common denominator across everything we do in wholesale banking — DvP settlement of securities and PvP settlement of FX.

As we build for the future, our design choices mark out the boundaries of what is operationally possible for years to come.

As we look at stablecoins, tokenised deposits, central bank digital currencies (CBDCs) or synthetic CBDCs (sCBDC), I want to share a seminal lesson on FMI design from 2018, when Utility Settlement Coin, later Fnality, reached a pivotal research phase to create tokenised versions of major fiat currencies for instant, back-office financial settlements between global banks.

What we know to be true of the liquidity landscape

First, cash management is about ensuring we have the right amount of money, in the right currency, at the right time. It’s more an art than a science.

One particular enemy is the word “and” — as in, we need the same currency in many places. In an uncertain world, where we don’t know for sure if everything will settle, or when all the money owed to us will be paid, we end up needing a little extra: a float, to grease the wheels. A bit here, a bit there, and pretty soon we’re talking real money.

There is also no ready means of same-day cross-border or cross-currency funding that would let you swap what you have now for what you need now with PvP settlement. There is pretty much no same-day FX market.

Knowing this, regulators have a clear responsibility to ensure the integrity of financial markets, and to make sure we have enough intraday liquidity to settle trades and make payments. 

On intraday liquidity, the Basel Committee of Banking Supervision’s BCBS 248 from 2013, is the main regulatory stick. But BCBS is guidance, not law. It evolved after the 2008 Global Financial Crisis during which liquidity was a visible issue. Different countries have enacted local regulatory requirements at different speeds. The Bank of England’s intraday liquidity risk and regulation is arguably the most prescriptive. The European Central Bank added its highly prescriptive guidelines in 2024. Following the 2022 UK LDI pension fund crisis, guidance for the buy-side has also been sharpened.

These regulations are increasingly warming the pot of water in which the cash-management frog at financial institutions still needs to jump out every day, to hit the button marked “all trades settled and all payments made today.”

With an instinctive knowing that assets are going to be tokenised, we can guess that the means of payment needs to be tokenised too. Perhaps we could make our existing payment systems 24x7x365 and add enough bells and whistles that they could interoperate with tokenised assets. That’s as likely as an English tennis player winning Wimbledon any time soon.

How design choices get made

When the first railroad tracks were laid in the 19th century, railways were highly fragmented. Different operators made their own choices about track width — aka gauge — often case by case, based on engineer preference, local business strategy, or regional politics. By the 1870s there were over twenty different gauges in use in America, and incompatible lines often forced freight and passengers to be transferred between cars rather than moving through on one train. Put another way: fragmentation and a lack of common standards meant a constant need to off-ramp and then on-ramp.

One of my wise friends, a former senior Google exec, often uses the term “friction” when talking to me about why new applications and methods succeed. Successful founders and designers see where the friction is, and work to eliminate it.

Lessons from Switzerland

Here’s where the seminal lesson comes in. In late 2017, two pioneers in the digital asset space, UBS’s Hyder Jaffrey and Rhomaios Ram, asked me to work on the Utility Settlement Coin (USC) project, which later became Fnality. They were on a mission: they could see that assets were heading on-chain, and that the market needed a means of payment on-chain too.

As I was weighing their offer, I went to lunch with a former boss of mine, Fritz “Itzi” Klein.

Over lunch I talked about the USC project — assets on-chain and payment on-chain. I said it felt like the existing set-up in Switzerland, where both payments (P) and securities settlement (DvP) use a single, shared cash account with the central bank’s RTGS systems to settle trades and make payments. This was good for cash management, and a good model to borrow from.

In 1999, Itzi had asked me to lead Credit Suisse’s CLS project. I knew he’d had a hand in designing Switzerland’s RTGS payment system, SIC (Swiss Interbank Clearing), in the 1980s. He was also intimately involved in designing CLS and was its chairman as the project went live in 2002.

At lunch, Itzi shared a history lesson that was new to me.

Today’s SIS securities settlement system was designed at the end of the 1980s, just after SIC went live as the Swiss RTGS payment system. The draft design, version 1.0, borrowed from the international central securities depository (ICSD) Euroclear, and proposed a securities settlement system with a dedicated cash account. This was perfectly plausible, and in actual use at the time, but it was a design that fragmented cash.

Itzi had just helped design and implement the RTGS system, SIC. For the CSD design, he saw that cash was going to be fragmented. In good Swiss-German dialect, he described that design to me as “en Seich” — rubbish. He pushed back and suggested an alternative: the new CSD system should use the cash already in the SIC system.

The Swiss banks couldn’t find consensus, so the decision was escalated all the way to the CSD Chairman’s office, where, after some debate, the design was changed to what Itzi had proposed and what I’d call a “single pool of liquidity” — a SPooL — serving both P and DvP.

Itzi told me how he’d used a domino analogy as the clinching argument for the benefits of a single pool of liquidity. Imagine a series of dominoes representing all the payments related to securities, plus all the other payments in the market to be made on any one day. The height of each domino is proportional to the size of the payment. If all the payments are in one place, you can push from one side, and the dominoes fall over easily. If you separate the payments, you fragment liquidity, and payments are not made so easily.

Importantly, in the Swiss set-up liquidity is not fragmented; there is no separation of the cash for securities (DvP) from the cash for payments (P). There is one pool, with a bridge driving the ability for the means of payment to interoperate with securities processing. Equally importantly, in times of normal liquidity availability — i.e. pre-2008 — the net funding requirement in CHF across all payment activity, including securities-related movements, was just 1–2% of the gross amount. That little is needed to settle every payment, gross, in real time, and in central bank money. That is both effective and efficient.

In addition, SIC does not allow overdrafts; banks needing liquidity can access repo markets, with the Swiss National Bank (SNB) as a regular counterparty for intraday repo — “intra” meaning both legs settle on the same value date. No use of overdrafts translates to little or no impact on intraday liquidity buffers.

The structure gives market participants the ability to quickly exchange what they have for what they need. Once securities collateral is available, it can be quickly repo’d, with the resulting “cash” in the SIC account available for payments.

All this makes the case for a single pool of liquidity very persuasive.

The Swiss solution for interoperability (1992) vs. best practice

Let’s look at the key ingredients in this Swiss solution. There is earmarking, one form of programmability — a hot topic in the world of blockchain and new digital assets. In simple terms, this means reserving one thing while you go and check that another thing can happen too. This requires interoperability between the means of payment and a settlement system. And payment is made using a regulated liability and central bank money. This follows the best practice recommended in the PFMI — the Principles for Financial Market Infrastructures. These are guidelines published by the Committee on Payments and Market Infrastructures (CPMI), whose work is administered by the Bank for International Settlements (BIS).

Let’s unpack the central bank money part. In the PFMI, “Principle 9: Money settlements” states:

“An FMI should conduct its money settlements in central bank money where practical and available. If central bank money is not used, an FMI should minimise and strictly control the credit and liquidity risk arising from the use of commercial bank money.”

Today, the Swiss solution doesn’t serve PvP — only P and DvP.

Design choices for tomorrow’s FMI

Whilst “money makes the world go around,” in wholesale banking it is “liquidity which drives our ability to settle trades and make payments.” How we change and realign our plumbing — aka FMI — matters a great deal. Plumbing is foundational. Get the design right and it’s a great enabler. Get it wrong and we’re stuck with awkward ways of doing things. Friction.

I can see a case for stablecoins — a case today, based on the alternatives we have available now. Yes, we could work to make them interoperable, and the folks at Ubyx are hard at work on that. But there will be friction. Sub-optimal.

I also see a case for tokenised deposits (TDs) — a case today, if that means banks have always-on capability and can readily support programmability. Project Agora showed how we might have a two-tier system with TDs and CBDC. I can see how a TD-centric world could work if all TDs are CBDC-backed and ring-fenced. That’s a lot of friction to overcome. Sub-optimal.

My reductive Bankers’ Plumber view is that we should build on the Swiss experience. What we need is a SPooL — a single pool of liquidity — in each currency, that can be held by any regulated FI during the business day. Both parts are necessary. We need a payment asset that can do all things P: P for payments, P for DvP settlement of securities, and PvP settlement of FX. And we need a way to break out of the dependency on correspondent banks and intraday credit — firstly because correspondents are increasingly selective about who they serve, and secondly because dependency on intraday credit drives up costs.

Looking at what’s cooking in our industry, I think Fnality — which I’d describe as an sCBDC, a synthetic CBDC — comes closest. It’s based on discrete, national, regulated payment systems. To really work, though, we need a compatible solution in each currency, and wider access to accounts. It’s a long railroad ahead.

• Olaf Ransome will be speaking at the PostTrade 360° conference happening on 2-3 September 2026. For more information about the event and to register, click here