There has been a lot of enterprise blockchain work in banking over the years. Much of the early work stayed at the PoC stage. Institutions were testing what could work, and where it would actually be useful.
Now, some of that work has reached production. Besu gives us quite a few examples to look at. For example, Citi processes close to $1 billion a day through Citi Token Services, which runs on Besu. DTCC, Clearstream and Euroclear have projects on it. SWIFT, Fnality and Project Agorá use it as well.
The official adoption numbers are interesting too. Among institutional platforms that disclose their technology and are in production or close to it, 29.4% use LF Decentralized Trust projects. Besu is one of the three most adopted platforms in that data. The data excludes all PoCs, by the way.
For me, this is a useful way to look at banking blockchain adoption. Follow the projects and see what they are doing. Some are live, some are getting there, and some still have a fair amount to work through.
But before we start, let’s brief Besu in a line. It is an open-source Ethereum client maintained under LF Decentralized Trust. Banks mostly use it to run permissioned networks where the validators are known institutions.
So why Besu, and what are these institutions using it for? Let’s see.
Why enterprises choose Besu
Besu can be used to connect to public Ethereum or to run a private network with a closed validator set. Most institutional deployments use the private network option.
There are a few reasons it keeps coming up in these projects.
One is the token standards. A token issued on Besu follows Ethereum standards, which developers and service providers already work with. That helps with hiring, audits, wallets, and the other things needed around the network.
It also gives institutions a way to work with public networks later. A bond or fund share issued on a permissioned ledger can be represented on a public one. There is work involved in doing that, of course. But they have a common standard to start with.
Then there is controlled participation.
Project Agorá runs Besu with QBFT consensus across permissioned validator nodes. A known group of validators votes to approve blocks. Banks and central banks know which institutions are writing to the ledger.
Open source is another reason. The Bank of Korea said it helps avoid vendor lock-in and allows a quick response to security issues. Citi could tune Besu’s consensus mechanism for its own volumes because it had access to the code.
I think that flexibility matters when an institution is building something it expects to operate for years.
And there is the ecosystem around it. When other institutions are already using the same foundation, there are more opportunities to agree on standard
s and connect. That becomes part of the decision too.
Which tokenized asset use cases run on Besu?
What CSD adoption tells us
A central securities depository, or CSD, keeps the official record of securities ownership and settles transfers.
Banks, custodians and issuers already connect to these institutions. So when a CSD adopts a ledger, it has implications for the participants around it.
DTC holds $114 trillion in assets under custody. It is launching tokenized entitlements for Russell 1000 stocks, ETFs, and US Treasuries. These are available at launch on its permissioned Besu network and Canton.
Clearstream custodies €22 trillion for 2,500 institutional clients. Its Besu network started during the ECB’s 2024 wholesale settlement trials. It needed a permissioned network to test delivery versus payment in central bank money. It then continued building on that network for D7 DLT. The first phase covers fixed income, tokenized money market funds and retail structured products. Equities are planned to come later.
Clearstream gave two reasons for choosing Besu. One was interoperability with other market infrastructures already using it. The other was having a route to public networks. As a CSDR-regulated entity, Clearstream needs a permissioned chain. Partners such as Kraken and Ondo can lock a token there and issue a representation on a public network using Ethereum standards.
Euroclear’s Project Pythagore is working on the NEU CP market. This market has roughly €300 billion in outstanding commercial paper, and some maturities are as short as a day. They introduced native digital issuance alongside their existing systems instead of building a parallel platform. The pilot will be completed in late 2026, whereas deployments will start in 2028.
There was an earlier DLT venue for this market, ID2S. It closed two years after its 2019 launch. It had no on-chain settlement or connection to Euroclear or Clearstream, and the pandemic reduced commercial paper demand around that time. Pythagore starts with the incumbent CSD, a G7 central bank, and wholesale CBDC settlement involved. So the setup is different.
The three CSDs have also published a joint interoperability framework. The remaining questions around connecting them are mainly legal and commercial. For a bank, the relevance is high. If it issues, distributes, or custodies securities in US or European markets, it will likely touch any one of these Besu platforms.
Issuers are looking at portability too
KfW, which is Germany’s 3rd largest bank, issued a €100 million digital bond on Polygon in June and plans to move it to RL1, which is a Besu-based cooperative network running on the SWIAT ledger. Their target is to launch €80 billion to €85 billion in bond issuance this year.
So, they are changing the ledger of record. It is more complex than creating a copy on another network. SWIAT has processed 50 transactions and more than €700 million in issuances. That includes the €300 million Siemens digital bond.
This is something we will keep an eye on. Institutions are starting to consider how an asset can move between infrastructures over its life, as well as where it is first issued.
Why collateral mobility is an early use case
Collateral mobility means getting an eligible asset to the counterparty that needs it at the time it is needed. There are several systems and institutions involved in doing this today. Custodians, triparty agents, and settlement windows all affect when collateral can be used.
Tokenization can help with some of that coordination.
The IMF reported a 30-day average daily volume of $303 billion for tokenized repo. Activity is still small compared with traditional repo and concentrated among affiliated entities. But it is an area where there is already activity to study.
DTCC’s Collateral AppChain is a Besu example.
Suppose a bank pledges a bond that is tokenized on Clearstream’s D7. The bond stays on D7, and the pledge is marked on the tokens. The AppChain maintains a record of encumbrances across custody systems. Pledges can then happen in minutes and outside normal market hours.
I find the asset staying on its existing ledger an important part of this design. It avoids introducing a bridge to lock the asset on one network and issue it again elsewhere. The AppChain also includes stablecoins, tokenized deposits and tokenized money market funds in its scope.
Deka Bank tested repo transactions during the ECB trials in November 2024. It completed two intraday repos with NatWest Markets and a five-day repo with LBBW. Each was for €1 million.
One transaction used a natively digital Siemens bond. The others used a tokenized Bund.
Deka’s SWIAT ledger, which is built on Besu, used a PoA consensus and handled tokenization and instructions to block collateral in conventional settlement systems. The Bundesbank’s Trigger Solution handled the cash leg.
There is a balance sheet question here as well. And speed alone will not give the balance sheet benefit. Dealers will benefit from netting matched trades. Immediate bilateral settlement can reduce netting and increase gross funding needs. Smart contracts can batch transactions and settle them periodically to recover some of that benefit.
So this needs to be part of the design. Faster settlement is useful, but the bank also needs to understand how much liquidity the process uses.
Tokenized money market funds and the ownership record
A tokenized money market fund is a money market fund whose shares are issued and recorded as tokens. Its economic profile stays the same. The collateral benefit comes from how the shares can be used. With a traditional fund, the holder may redeem shares, receive cash, and post the cash. The receiver might then reinvest that cash in a fund. So, a total of four steps.
Tokenized shares can be posted directly, provided the arrangement supports it. This can reduce steps and allow transfers outside traditional hours. The market is about $17.5 billion now, according to the IMF. The shares are held mainly as yield-bearing, cash-like collateral.
But there is a legal problem, and because of which around half of the largest TMMFs use a traditional book-entry ledger as their primary ownership record, as per ECB. None of them are fully on-chain, and shares legally change hands only after that ledger is reconciled with blockchain. This is because a distributed ledger has legal significance only if the register or record is maintained on-chain.
This is where a permissioned Besu ledger can help. The validators are known, regulated institutions. For a fund issuer or transfer agent, that can make it easier to justify keeping the legal register on-chain instead of using the chain to mirror an off-chain book. The legal arrangement still needs to support it, but the network gives them a basis to work with.
Two Besu platforms already plan for TMMF collateral. Clearstream’s D7 includes tokenized money market funds in its first phase, and DTCC’s Collateral AppChain lists them in scope.
A bank receiving these shares as margin still needs to ask the same question. Is the ledger the legal register for the fund, or is it a mirror of an off-chain register?
Because that answer will determine when settlement is final.
How far tokenized deposits have progressed
A tokenized deposit is a commercial bank deposit represented on a programmable ledger. It remains a claim on the issuing bank.
One example could be Citi Token Services, which has been live since April 2024 and processes close to $1 billion a day. Clients in the US, UK, Ireland, Singapore and Hong Kong can move liquidity 24/7. They do not need to open new accounts, hold tokens, or manage wallets. That’s a major benefit. The client can get access to the service through banking arrangements they already use. The service runs on CIDAP, Citi’s digital asset platform built on Besu. Citi has connected it to its 24/7 USD Clearing network too, allowing clients to pay any of 250 banks across 40 markets.
DCJPY in Japan and UBS Digital Cash are other Besu-based examples.
The Bank of Korea’s Project Hangang shows how spending conditions can be added. Around 80,000 people paid merchants with tokenized bank deposits. Banks settled with each other using wholesale CBDC on the same ledger. The deposits are plain ERC-20 tokens. Spending conditions sit separately in an ERC-1155 layer.
For example, electric vehicle charging subsidies can be restricted to approved suppliers and released as installation milestones are completed. In a civil servant expense pilot, a payment at an entertainment venue or on a weekend fails at the point of sale.
The conditions can change without changing the deposit token.
Single-bank platforms do have a reach limitation. Payments between customers of the same bank are easier to coordinate. Other banks add more steps. Swift’s Besu-based ledger can address this. It does not hold deposits. It records, sequences, and validates the payment commitments banks make to each other. The deposits remain with the banks, which settle later on through conventional channels. Its minimum viable product went live in July 2026 with 17 pilot banks. The first live transactions followed within a month.
What cross-border projects are testing
Correspondent banking involves a chain of banks. Each works to its own hours and cut-off times. A payment gets held up when one of the institutions is closed.
Project Agorá is testing how shared infrastructure can coordinate this. It brings together eight central banks, covering all G7 economies plus Korea, Mexico and Switzerland, and more than 40 financial institutions. Its design needed two things Besu provides. Each central bank and commercial bank had to hold a known validator role under QBFT, and smart contracts had to coordinate payments across several ledgers.
Agorá has two layers. Tokenized commercial bank deposits are on a shared unifying ledger. But tokenized central bank reserves are on separate ledgers, one for each currency area under each central bank’s control. A smart contract coordinates the payment. If any bank fails to confirm in time, the payment will be stopped, and no money moves. Only if everyone confirms both legs settle together. In July this year, Agorá did live trials with real money. Twenty-two institutions and five central banks completed 30 transactions worth roughly CHF 800,000 across six currencies. Average settlement time was around 80 seconds.
Finality tells us another part of wholesale payments. The Sterling Finality Payment System runs on Besu and went live in 2023 as a regulated, systemically important payment system. Participants hold balances in an omnibus account at the Bank of England. The ledger records their entitlements. A settlement finality designation granted in late 2024 makes transfers legally final on the ledger itself.
For FX, liquidity use needs more attention.
CLS gives members liquidity savings of about 99% through multilateral netting, according to Deutsche Bank Research. Most tokenized settlement networks do not yet net at that scale. So before moving FX volumes, a bank needs to compare the funding requirements with its existing position. The settlement process may be quicker, while the liquidity requirement is different.
Privacy is a separate requirement
Permissioning decides who can join a Besu network. It does not automatically hide transaction details from the other members. Banks need confidentiality around clients, counterparties, and transaction amounts. So a shared network needs a privacy layer as well.
Zeeve’s privacy layer is an open-source privacy framework for EVM-compatible chains, including Besu. It brings four layers of privacy. Application layer,which is the most common one, then asset layer, ledger layer, and, of course, network layer privacy.
It is modular, which allows a network to change or upgrade privacy engines without rebuilding the ledger. Banks can start with application layer privacy and then integrate other modules as the need arises. Or if they are already using application-layer privacy, they can integrate other modules too. The use of ZK Circuits, ZK tokens, and important modules will bring in the required privacy a bank needs.
Zeeve for Besu implementation from pilot to production
The work around the ledger is a large part of taking a Besu project into production. Privacy, integrations, and ongoing operations all need to be in place.
Zeeve is a certified Besu service provider and a member of LF Decentralized Trust. We have a growing number of Besu-certified engineers who can bring to the table the level of depth a bank requires in its use case design. Plus Zeeve deploys and operates permissioned Besu networks for banks and financial institutions on ISO 27001 and SOC 2 Type II compliant infrastructure.
Institutions can run nodes in their own cloud accounts through Bring Your Own Cloud, on-premise or in a hybrid model as well.
A scoping session with Zeeve’s team can start with the first use case, whether that is tokenized deposits, collateral mobility or tokenized fund shares. From there, the team can map the privacy, integration and operational requirements onto a Besu reference architecture.
Schedule a call with our Besu experts today.
