6 Costly Mistakes to Avoid When Building Tokenized Banking Products

6 Costly Mistakes in Tokenized Bank

In 2019, JPMorgan introduced JPM Coin to allow its institutional clients to move money using blockchain. At the time, there wasn’t much to refer to. The bank had to develop its own infrastructure and figure out how tokenized deposits could work with its existing banking systems.

Citi, HSBC, BNY, Deutsche Bank, and other large banks later followed with their own tokenized products and services. Some built their platforms from scratch, while others used existing blockchain frameworks. 

Today, JPMorgan’s Kinexys processes more than $7B USD daily and has processed over $4T USD in total. Citi Token Services handles close to $1B USD per day, while HSBC processed around $28B in payments in the first quarter of 2026. 

But getting here took years of development and considerable investment. 

The largest banks have spent hundreds of millions of dollars developing these capabilities. Much of that money went into understanding the technology, regulatory requirements, and whether tokenized deposits would actually be useful for corporate clients.

Now, if you’re a regional or mid-sized bank looking to offer similar services, do you really need to spend the same amount of money and time?

Well, no, you don’t. The banks that started early had to find answers to questions nobody had answered before. Today, much of that information is available, and so is the infrastructure needed to build similar services.

In fact, Wells Fargo, which announced its tokenized deposit initiative in August 2026, chose Cosmos technology instead of developing an entirely new blockchain platform.

So, before banks start building their own tokenized deposit services, it makes sense to understand what the early movers have learned and which mistakes can now be avoided.

What Have the Largest Banks Built?

A tokenized deposit is just a bank deposit represented as a token on a blockchain or another distributed ledger.

It remains a liability of the issuing bank, can earn interest, and may qualify for deposit insurance depending on the applicable legal requirements.

But if deposits are already digital, why do we need to tokenize them?

Well, because tokenization allows banks to add capabilities that conventional banking systems may not easily support. These include moving funds 24/7, executing payments based on predefined conditions, and managing liquidity across accounts without depending on normal banking hours.

And the largest banks have already demonstrated how these services can work.

JPMorgan started in 2019 with a private Ethereum-based ledger. Its initial objective was to reduce unsecured intraday exposure when institutional clients moved funds. Over the years, Kinexys expanded to other blockchain environments, including Coinbase’s Base and Canton.

Citi launched Citi Token Services in April 2024 using a Hyperledger Besu-based platform. Here, customers don’t need separate blockchain wallets or new bank accounts. They continue using their existing accounts while Citi handles tokenization in the background.

HSBC launched its Tokenized Deposit Service in 2025. But by the first quarter of 2026, it was live in five markets and six currencies, with about US$28 billion in payments that quarter, according to the HSBC Global Payments Trends report.

In August 2026, Wells Fargo also announced their service. At first, they will do USD → GBP payments for a few corporate clients, with a bigger rollout planned through 2027. They picked the Cosmos stack instead of building their own blockchain from scratch.

Although these banks took different approaches, their systems have many things in common. They use permissioned ledgers connected to their existing banking infrastructure. Only authorized participants can use the service, while the core banking system continues to maintain the official deposit records.

This means banks don’t necessarily need to replace their existing infrastructure to offer tokenized deposits.

And that’s where our first lesson comes in.

1. Don’t Build the Entire Blockchain Infrastructure From Scratch

    When JPMorgan started developing JPM Coin in February 2019 for its institutional clients, building its own infrastructure made sense. There weren’t many proven platforms for tokenized deposits, so the bank had to test different technologies and develop its own solutions.

    But things have changed.

    Today, banks can use existing blockchain frameworks and tokenization infrastructure rather than developing everything themselves.

    And there’s a good reason to consider this. A blockchain is only one part of a tokenized banking system. Banks also need identity management, privacy, compliance controls, transaction monitoring, and integration with their core banking systems.

    So, building an entire platform internally is the most expensive approach from a business angle. If we go by the BCG’s playbook, it ranks Kinexys in the “full build” category, which has the maximum revenue capture but also the highest cost. These kind of platforms also takes the most risk and have the slowest time to revenue, if you see.

    This is kind of a rare choice, mainly for banks that want to sell the platform to other institutions. There could be only a dozen banks that might fit this category. But for a regional bank that simply wants to offer tokenized deposit services to its corporate clients, it may only add unnecessary costs.

    Source: BCG Analysis

    But one thing to keep in mind is to have an exit option. Use open standards and a ledger you can run in your own environment.  They should make sure they can change vendors or bring more of the infrastructure in-house later, if they want to.

    Zeeve comes in the full-stack category and offers the fastest path to deploying market-ready solutions.

    2. Don’t Spend Years Running Pilots That Never Go Live

      Banks have been experimenting with blockchain for nearly a decade. But not every successful experiment has become a commercial product. Take the repo market, for example.

      According to ICMA’s June 2026 report, there were 34 publicly reported DLT repo tests and transactions between 2017 and 2025. But only Broadridge’s Distributed Ledger Repo and JPMorgan’s Kinexys had reached regular commercial use by the end of that period.

      But why did so many pilots fail to reach production?

      Well, because it’s not just about testing whether blockchain technology worked. They were also trying to understand regulatory treatment, settlement arrangements, and whether other banks would join the network.

      And these questions weren’t always something a technology pilot could answer. For example, the classification of JPM Coin under European regulations has remained contested, with the European banking regulator treating it as an e-money token rather than a conventional deposit.

      However, banks starting today have more clarity than the early movers. Both on the technology and regulation sides.

      In April 2026, the FDIC proposed that deposit insurance treatment should depend on whether an arrangement legally qualifies as a deposit, not on the technology used to record it. The proposal still needs to be distinguished from a final rule, but we have a clear direction.

      So what does a regional bank still need to test? It’s not technology, as there are many production examples that are running at scale. Legal issues are also pretty much solved. So what’s left is whether your own clients will use it and pay for it.

      A small launch with selected corporate clients and a clearly defined use case can help answer these questions.

      Cost is important, too, here. A mid-sized bank can’t fund a three-year experiment the way a global bank can. Correct?  The better path is a product that’s already live and plugs into the core software the bank can use today. The bank turns that product on, and there’s no long build project.

      As an example, take the Cosmos tokenization suite, which had been approved for Fiserv’s app store and was being integrated with Jack Henry. So a bank on Fiserv could add it like any other approved module and skip a multi-year pilot.

      3. Don’t Replace Your Core Banking System Just to Offer Tokenized Deposits

        A bank’s core banking system manages customer accounts, deposits, and transaction records.

        And replacing it can be expensive.

        For banks using older systems such as Fiserv Premier or Hogan, a full core migration may cost US$10 million or more. But does a bank need a new core banking system to introduce tokenized deposits?

        Not necessarily.

        The production examples have shown that tokenized deposit services can work alongside existing banking infrastructure. This is possible through what is known as a sidecar ledger.

        Well, a sidecar ledger is simply a separate ledger that works alongside the existing core banking system. The core continues to maintain the official records, while the sidecar adds capabilities such as programmable transactions, sub-accounts, and 24/7 movement of tokenized deposits.

        For example, a company may want to move funds between two of its accounts on a weekend. A sidecar ledger can support that movement within the bank’s tokenized deposit environment, while the transactions are synchronized and reconciled with the core.

        And banks can continue using their existing compliance systems.

        Transaction information from the ledger can be passed to the bank’s AML and sanctions screening tools before settlement. This allows existing controls to remain part of the process without developing an entirely separate compliance function.

        Of course, integration, reconciliation and operational controls still need proper attention.

        But the bank doesn’t have to undertake a complete core replacement just to add these capabilities.

        A November 2026 survey of 200 senior decision-makers in the US and Canada found that 69% planned to adapt their existing infrastructure rather than build separate systems.

        Source: Broadridge

        There’s another thing worth considering.

        Not every banking service needs blockchain. Some benefits associated with tokenized money can be achieved through conventional systems.

        So, if a bank can already offer a service efficiently through its existing infrastructure, tokenization may not add much value.

        It makes more sense for services that are difficult to support today, such as programmable escrow, automated weekend sweeps, and conditional payments.

        4. Don’t Build Tokenized Services Without a Clear Business Use Case

          Tokenized deposits are often discussed as a way to move money instantly between different banks.

          But that’s not where most of the early transaction volume came from.

          Much of it involved moving funds between accounts within the same bank.

          JPMorgan initially focused on reducing intraday liquidity exposure. Siemens later used its real-time liquidity pooling service to reduce liquidity requirements by 50%, reportedly saving more than US$20 million annually.

          BMW also used tokenized banking infrastructure to move money between its accounts in Frankfurt and New York.

          But why did these banks start with such transactions?

          Well, a tokenized deposit is a claim against the bank that issues it.

          When both accounts are held at the same bank, the institution can manage the transfer within its own ledger.

          But moving a deposit between two different banks requires additional clearing and settlement arrangements. The receiving bank cannot simply accept another bank’s liability as its own.

          And that’s what makes multi-bank tokenized payments more complicated.

          For regional banks, starting with existing corporate clients can be much simpler.

          There are several use cases worth considering.

          Programmable escrow

          Banks can use tokenized ledgers to maintain separate balances and release funds when predefined conditions are met. This can reduce the need to manage multiple accounts and process manual release instructions.

          Corporate treasury sweeps

          Companies can move funds between their own accounts or entities automatically, including during weekends and outside normal banking hours.

          FBO account sub-ledgering

          Banks serving fintech companies often maintain For Benefit Of accounts holding funds attributable to multiple end customers. Tokenized sub-ledgers can help track individual balances and support reconciliation, although banks remain responsible for maintaining accurate records.

          But choosing a use case is only the first step.

          Banks also need to understand whether it makes business sense.

          How many customers would use the service? How much manual work would it remove? Would it generate additional fees or reduce operating costs?

          If there’s no clear answer, the bank may be building a technology platform without knowing what to do with it.

          5. Don’t Make Interoperability More Complicated Than Necessary

            JPMorgan, Citi and other large banks initially built tokenized deposit platforms that worked mainly within their own networks.

            This allowed customers to move funds efficiently within the same banking environment.

            But there was a limitation.

            A tokenized deposit issued on one bank’s ledger couldn’t automatically move to another bank’s ledger.

            And that’s still one of the biggest challenges in tokenized banking.

            Now, several initiatives are working to solve this problem.

            The Clearing House is developing infrastructure to support clearing tokenized deposits between banks.

            SWIFT is also developing a shared ledger that allows banks to maintain deposits on their own systems while coordinating transactions through a common infrastructure.

            In August 2026, HSBC and Standard Chartered completed an initial transaction using SWIFT’s ledger.

            But these networks are still developing, and there isn’t yet a single widely adopted system for cross-border tokenized deposit settlement.

            What about blockchain bridges?

            Well, bridges are commonly used to connect different blockchain networks.

            But they also introduce additional risks.

            Some of the largest security incidents in the cryptocurrency industry have involved blockchain bridges.

            For banks, using a third-party bridge means relying on another organization’s infrastructure and security controls to handle transactions involving customer funds.

            And if something goes wrong, the bank still has responsibilities to its customers and regulators.

            One alternative is Inter-Blockchain Communication, or IBC. It has been operating across blockchain networks for several years and is moving under LF Decentralized Trust governance now.

            IBC is a protocol that allows compatible blockchain networks to communicate directly and verify messages without depending on a centralized bridge operator. In the very recent project Simbolon, Patrior delivered a multi-network orchestration with Deutsche Bank and DBS Bank as participants. They used IBC for it. 

            Source: Simbolon Project Report

            However, using IBC doesn’t automatically solve all the legal, security and settlement requirements involved in interbank transactions.

            So, rather than depending on a single connection method, banks should choose infrastructure that can support different interoperability standards as the industry develops.

            Privacy is equally important.

            Banks participating in shared networks cannot expose their customers’ balances, counterparties or transaction details to competing institutions.

            That’s why privacy and access controls should be considered from the beginning, rather than added after the platform is developed.

            6. Don’t Assume Tokenized Banking Is Only for Large Banks

              The largest banks spent years developing tokenized deposit infrastructure because they had to build and test much of the technology themselves.

              But banks entering the market today have access to existing blockchain frameworks, tokenization platforms and managed infrastructure.

              This can make implementation considerably less expensive.

              Practitioner estimates suggest that a mid-sized bank could begin implementing tokenized banking capabilities with a budget of approximately $200,000 USD to $5 million, depending on its existing infrastructure and the scope of the project.

              These aren’t verified industry-wide cost figures, but they show how different the investment can be from building an entire platform from scratch.

              And there’s another reason why smaller banks should pay attention to tokenized deposits.

              Their business depends heavily on customer deposits.

              Now, with stablecoins becoming increasingly popular for payments and treasury operations, businesses have more options for holding and moving digital dollars.

              But unlike tokenized deposits, stablecoins issued by another institution are not deposits on the customer’s bank balance sheet.

              So, if corporate clients move funds out of their bank accounts into externally issued stablecoins, banks may lose deposits that would otherwise support their lending activities.

              The Independent Community Bankers of America has raised concerns about this risk, referring to a potential US$1.3 trillion in deposit outflows associated with stablecoin adoption.

              This is a risk estimate, not an amount that has already left the banking system.

              Tokenized deposits offer banks a way to provide some of the same digital payment and treasury capabilities while keeping the underlying funds within the banking system.

              And regional banks have already started exploring this.

              The USDF Consortium, a network involving US regional banks, uses Cosmos-based infrastructure for tokenized deposits.

              This means smaller institutions don’t necessarily need to build their own blockchain networks to participate.

              They can use existing infrastructure to offer services that would previously have required much larger technology investments.

              How Can Banks Get Started?

              Now that we’ve looked at these six mistakes, the next question is where banks should actually begin.

              Well, the first step should be finding a problem worth solving.

              It could be a corporate client struggling to move liquidity between accounts outside banking hours. Or a business that depends on manual escrow processes. Or a fintech partner that needs better sub-accounting and reconciliation.

              Once the use case is clear, the bank can evaluate whether tokenization would improve it. If it does, the bank can then look at the infrastructure required.

              This doesn’t necessarily mean developing an entirely new platform. A permissioned sidecar ledger can work alongside the existing core banking system with compliance tools and privacy controls integrated into the setup.

              And since interoperability is still developing in the whole banking ecosystem, the infrastructure should also be flexible enough to connect with other networks in the future.

              The important thing is to start with something that customers actually need and expand as the service proves useful.

              Building Tokenized Banking Infrastructure with Zeeve

              For banks that don’t want to develop their tokenized deposit infrastructure from scratch, Zeeve provides blockchain deployment and management services. Zeeve has supported a major US bank’s tokenization initiative and has experience working with institutional blockchain infrastructure.

              Our infrastructure and processes are ISO 27001 and SOC 2 Type II compliant, which can help banks during security, risk, and vendor assessments.

              Zeeve is also a partner of Cosmos Labs. Through Zeeve, banks can integrate the Cosmos Tokenization Suite to issue tokenized deposits. Banks can deploy this infrastructure within a controlled environment without replacing their existing core banking systems.

              For EVM-compatible networks, Zeeve’s privacy suite offers application, ledger, asset, and network-level privacy.

              And with IBC, compatible ledgers can communicate directly without depending on conventional third-party bridges.

              This allows banks to introduce tokenized banking capabilities while avoiding much of the infrastructure development work that early adopters had to undertake.

              Talk to Zeeve experts today to discuss your requirements.

              About the Author

              Ghan Vashishtha

              Co Founder & CTO

              Ghan Vashishtha is the Co-Founder, CTO of Zeeve with more than a decade of expertise in niche industry sets including Telecommunication, Defense & Aerospace and Media & Broadcasting. He has been providing technology and innovation leadership, building strategic roadmap while building strong partnerships ecosystem He helps build early traction while driving cross technology vertical engagements.

              Subscribe to Zeeve Newsletter!

              Get our latest news, announcements, and other value-packed insights straight to your inbox, join our 30000+ subscribers newsletter .