Might the Nostro era come to an end?
Guest insight from Olaf Ransome on UNITE Global FMI and a “single pool of liquidity” vision to reduce nostro reliance, cut buffers and enable real-time PvP/DvP.
The nostro era is coming to an end
Guest article by: Olaf Ransome Wholesale banking liquidity, settlement and FMI market structure.
I am a liqiudity hawk. Liquidity is what makes the wholesale banking world go round. Making sure a firm has money to settle trades and make payments each day is an essential daily operational task.
Typically, we call that cash management, sometimes Treasury. It is not a simple task. A bunch of ingredients contribute to making it hard: most financial institutions need money in lots of places, most have a reliance on their correspondent banks for intraday overdrafts and on top that the liquidity buffers are pretty expensive.
With that background, a news story that tells me “the nostro era is coming to an end” is certain to grab my attention.
This is the claim which a start-up, UNITE Global FMI, made recently about its Real-Time Super-Centralised Liquidity mechanism. Intrigued, I spent some time digging into this with Unite Ambassador Stan Cole , and CTO & Founder Tord Coucheron . Whilst UNITE is still at the concept stage, there is an extensive on-going dialogue with regulators and central banks.
I have written at length about liquidity challenges in wholesale banking and what I think nirvana looks like. Click here to see a summary view and then dive into whichever part interests you.
Here’s the “how it works”
UNITE is aiming to create a unified global infrastructure, an FMI dubbed “Global Correspondent Hub", using central bank money, which will support real-time settlement with settlement finality in all things P: P for Payment, P for DvP / Delivery vs. Payment and P as in PvP, payment vs. payment for FX. You might also call this “multi-asset”. And all without today’s reliance on correspondent banks.
We have existing FMIs which do things P, several who do DvP and the clever Swiss at SIX Group who can do P and DvP from the same account. And we have CLSBank which does PvP and nothing else. As excellent as each existing piece of FMI is, we are not efficient as an industry; the liquidity buffers are huge and the associated costs high, there is still a huge amount of settlement risk associated with the payments for FX trades not settling in CLS, CLS settlement itself relies in large part on intra-day credit, and added to all of that the existing FMIs are not flexible enough to serve the emerging world of new digital assets. That existing world of wholesale trade settlement & payments come with a really expensive amount of operational overhead.
So, UNITE has some promise and potentially offers what I consider to be nirvana, a SPooL, aka Single Pool of Liquidity in each currency. Liquidity is aggregated centrally on the UNITE platform and available in real-time. Liquidity in each currency is at all times kept as client funds at a designated central bank account with the currency’s respective central bank. Any UNITE user (bank) can therefore hold liquidity in any supported currency. It is that potential which makes it worth trying to understand it.
UNITE’s foundations are a common infrastructure supporting accounts in multiple currencies. Participation would be limited to banks. Each participant can do a single on-boarding and access all the currencies. Accounts would be pre-funded, so there is neither credit risk nor market risk. As an FMI, the structure would ring-fence the underlying assets, which means the operating would be bankruptcy remote from the underlying assets. This is an important feature as it means that the account holders in any one currency have a collective claim on the pool of central bank money in that currency.
In addition to settling FX trades on a PvP basis, UNITE’s network banks will also provide any FX needed when a user’s balance in a currency is not high enough to cover the payment. DvP and P settle instantly too. For banks on the UNITE FMI, settlement is continuous, gross and is not constraint by domestic RTGS hours. To banks not connected to UNITE, but in markets with efficient Instant Payment Systems (IPSs), payments are also delivered and settled in real time.
UNITE is Norwegian based and has obtained a local license. UNITE aims to access central bank money in any of the UNITE supported currencies indirectly, via its partner network banks for each currency, using an elaborate a specially designed legal framework. The construct needs a few ingredients to be successful:
1. The technical capability from the central bank to operate an extra, discrete account for an existing domestic bank. Should be possible; way back when we made that happen in Switzerland as part of the preparation for CLS settlement. The three settlement members were allowed an extra account so that funding from repo with the SNB to support timed payments to CLS did not go straight into the main account and get used for domestic payments.
2. The policy and political will to allow such a novel construct, to support a private sector initiative by setting up a framework which is enabling. Some central banks are convinced that “nobody does it better” than themselves.
3. Local banks willing to be the domestic partner. On first pass, this might be perceived as competing with their transaction banking business. In a new world, they would be trading liquidity rather than clipping say $1.50 for a payment.
Alongside access to a central bank account, another essential ingredient is that in a currency the settlement system has designation under the local settlement finality laws. In general, this requires a local company to have an FMI license of one or other flavour. In UNITE’s model, such “local company” is the network bank on UNITE for the given currency and market–as a rule, a systemically important bank (SIB). The network bank itself is not involved in the settlement process, nor does it have access to the UNITE client funds. This is key as it means UNITE client funds are bankruptcy-remote in case a network banks fails.
Another dynamic which plays a role are the “access rules”. Today, central banks will typically limit access to central bank accounts and the associated FMI to those institutions which they regulate, which in turn means the participants are local legal entities: UBS Switzerland does not have an account at the Bank of England (BoE) but UBS London Branch does, and it is regulated by the BoE. Because access comes with the privilege of holding central bank money, access is intertwined with monetary policy. If Switzerland’s ZKB could hold GBP at the BoE, this would reduce deposits held by commercial banks, which in turn restricts their lending and complicates the execution of monetary policy.
To date, there are very few cases where central banks allow “foreign” institutions direct access to a central bank. This is sometimes referred to as “remote access”. It is a very special privilege, which as far as I know was last used widely by CLSBank, and is allowed on a very limited basis by the Swiss National Bank to support domestic CHF repo needs. There is an existing start-up, Fnality , which also aims to be a SPooL. Fnality is going down the route of setting up a licensed entity in each currency, precisely because a licensed entity can obtain the local FMI license and with that the local settlement finality designation. If UNITE were able to deliver on its goal to access central banks in the G20 (less Russia, plus Switzerland) and beyond through its very specific legal setup, then this would be a huge step forward for wholesale market infrastructure.
Yes, we’d need to find a way to deal with non-local institutions in a currency so that they can accesss the system for the purpose of. Settling trades and making payments, without being able to have central bank money overnight. For me, this is super easy. Using GBP as an example, the regulators would set some rules. Foreign banks would be allowed accounts, but only for intraday usage. The rules would require them to fund to zero, or some small notional amount before close of business. The system operator would keep some standing instructions, for example “buy money market fund X”, or “transfer to locally regulated bank Y”.
For this to be implemented two things would be needed:
1. Any new FMI would need a construct that allowed the individual central banks to remain sovereignty over who can access their currency. Yes, a single on-boarding process is a good thing, and for sure nobody wants to go and ask every central bank if proposed member X is ok. That said, one could imagine that the Danes and the Canadians would like to be able to say “No American banks” if they were in the mood. Now, I am not saying this is good, or that they would, but I am saying, in a global, central system, the trade-off for flexibility is keeping currency sovereignty.
2. Regulators would need to get their heads around the idea that during the day, it does not matter how many say GBP Switzerland’s ZKB has, but they can regulate what it does to avoid overnight balances. And in any event, today financial institutions generally do not want to hold long cash balances, because these are unsecured receivables in the balance sheet, so they prefer to buy money market funds or do a reverse repo. Loads more on this in the linked piece highlighted in “Detail” above.
With an eye to the emerging need to support trade settlement and payments in new digital assets, another ingredient is interoperability; can a currency balance in a system interact with say an exchange where intraday FX is traded, or an exchange where tokenised assets are traded?
Building great financial market infrastructure (FMI) is neither quick nor easy. We need FMIs that support multi-currency for all things P, and we need that interoperability with new digital assets.
As we build for the future, we have the opportunity to build in a way which supports efficient trade settlement and payments. For that reason, new developments, like the one at UNITE, are worth keeping an eye on.
AUTHOR:
|
Olaf is a liquidity and financial services expert. He is the founder of 3C Advisory You can message Olaf directly here. |