Temenos at Sibos 2026: Customer Churn Risk, Agentic Payments and Tokenised Money

For banks, customer retention can mask a gradual loss of business, while new forms of digital money are testing the systems that move and safeguard funds. Those two pressures shaped Temenos’s conversations at Sibos 2026 in Miami: Will Moroney discussed customer expectations and core modernisation, while Mick Fennell examined agentic payments, tokenised deposits and other digital currencies.

The customers who may already be leaving

Research produced by Celent and released at Sibos surveyed more than 2,500 banking consumers across five regions. Nearly three quarters said they were, at best, moderately satisfied with their main bank. Temenos calls this group the “switchable middle,” a customer base that may appear stable in account data while shifting activity to other providers.

“Retention doesn’t necessarily mean loyalty,” said Moroney, Temenos’s chief revenue officer. Customers can keep an account open while routing more payments or borrowing through fintechs. For banks, that makes account closure a lagging signal: by the time a customer leaves outright, the institution may already have lost part of the relationship.

Moroney sees everyday payments as a particularly exposed point. Customers make them frequently and often under time pressure, whether paying a bill or sending money to a friend. Their service expectations are shaped by digital platforms outside banking, including social media and ride-hailing apps, where quick, simple interactions have become familiar.

Younger customers are more likely to maintain relationships with several banks, Moroney said, and can switch between providers with relative ease. A moderate level of satisfaction may not be enough to keep their full range of financial activity. That leaves banks with a difficult measurement question: does a strong retention rate reflect genuine preference, or simply customers’ willingness to keep an account they use less?

Why core systems are back in focus

Banks have made visible progress in digital channels and instant payments, Moroney said, crediting institutions including Swift for advances in the latter. But improvements at the customer-facing end can place more demand on back-office systems that were built for older processes. If a digital payment moves straight through the front end but slows down behind it, customers still experience friction.

Moroney’s case for core modernisation rests partly on a change in what banks can learn before they replace legacy platforms. He argued that AI can help institutions examine decades of code, data and business processes, giving them a clearer view of what existing systems do and how they support operations. In his view, that understanding could reduce a major source of uncertainty in large-scale migration and help banks move to newer core systems faster.

The connection between customer satisfaction and infrastructure is direct. Banks can improve digital interfaces, but delivering consistently fast service depends on the systems beneath them. Moroney’s message to Sibos delegates was that a large share of customers remain open to alternatives, even where conventional retention figures look healthy.

Agentic payments put controls under pressure

Fennell, Temenos’s business line director for payments, described agentic commerce as a shift from AI supporting payment activity to machines making decisions and payments with other machines. Early activity is concentrated in corporate settings, where proofs of concept and pilots are under way. To support those flows, he said, banks need cloud-native, event-driven and API-first infrastructure; older systems may struggle to respond at the required pace.

Speed also changes the demands on fraud controls. A human reviewer cannot assess each transaction in a high-volume, machine-led flow. Fennell’s approach is to agree in advance what an agent can do for a customer, including limits on the amount and frequency of payments, then monitor activity against those terms. He compared the model with direct debit mandates: a known form of customer authorisation adapted to faster, more frequent payments.

Fennell sees this as an extension of existing payment infrastructure, with stronger controls layered onto familiar account and transaction processes. “We’re still moving money,” he said. “There’s still an account at the end. There’s still going to be a debit and a credit.” The operational shift lies in the speed and volume of activity, and the need to set clear boundaries before an automated agent acts.

Four forms of digital money, four different roles

Fennell used the “four musketeers” to distinguish the main digital currency types banks are weighing. In his analogy, central bank digital currencies are Athos, representing public money; stablecoins are Porthos, private money with momentum but risks to consider; tokenised deposits are Aramis, bank money operating within existing rules; and cryptocurrency is D’Artagnan, which he characterised as exciting, unbacked and a case for caution.

Tokenised deposits are the most direct fit for banks, Fennell argued. They represent customer deposits in tokenised form for use on blockchain networks, while remaining bank money managed by the institution. He said they follow the same rails and regulatory framework as existing deposits, with tokenisation adding a new way to use and transfer those funds.

That does not mean every bank will support every form of digital money. Fennell expects institutions to focus on tokenised deposits and, where there are live use cases, central bank digital currencies. He pointed to the digital euro as the largest CBDC effort in train. Banks will need to assess the risks, services and redemption arrangements associated with each asset, including whether customers can redeem at par under stress.

His analogy also underscores that these categories serve different purposes and carry different responsibilities. Banks may be able to manage tokenised deposits as an extension of their existing customer relationships. Other forms of digital currency can raise distinct questions about backing, liquidity and the bank’s role in maintaining trust.

Stablecoins show promise within corporate networks

Fennell was more cautious about stablecoins as a broad solution for cross-border payments. Around 97 per cent are denominated in US dollars, he said. A payment between Malawi and Japan would still need to be converted into yen, and the on-ramp and off-ramp infrastructure needed to handle that process at scale is not yet in place.

Use is more practical today within large companies and multinational groups that move money between their own entities. When both sides of a transaction can be managed, stablecoins can offer real-time settlement and a shared ledger that makes payment activity visible. That narrower use case avoids some of the conversion and access challenges faced in open cross-border networks.

Fennell said customers care about cost, access, speed and transparency, the objectives set for cross-border payments by the G20, rather than which payment rail processes a transaction. In his assessment, tokenised money and smart contracts already help with three of those four goals. He expects lower costs to depend on deeper liquidity pools.

He also expects tokenisation to extend beyond payments into lending and trade finance. For platform providers, the challenge is to let banks manage traditional and tokenised finance together. “The future of banking is tokenised, or at least a significant part of banking,” Fennell said, describing a single platform for both forms of finance as the objective.

Liquidity and trust remain the practical tests

Fennell identified liquidity, custody and trust as unresolved considerations alongside technology choices. Banks already manage pools of liquidity across many currencies through correspondent relationships. Holding liquidity across hundreds of stablecoins could fragment balance sheets, and he said no one has fully answered that challenge.

Adoption also varies by market, including in places where rules exist. Fennell said governments and regulators are concerned about control over money movement within national borders, which affects how quickly digital currencies gain traction. For banks, the question is therefore wider than whether a technology can process a transaction: they must decide how to manage liquidity and safeguard customer confidence as new forms of money develop.

Moroney’s research and Fennell’s payments analysis point to the same operational priority. Banks need to make everyday services dependable while preparing systems for faster, more automated transactions and tokenised assets. The competitive test will be whether their infrastructure can keep pace with customer expectations without losing control of the money moving through it.

The interviews with Moroney and Fennell were recorded at Sibos 2026 at the Miami Beach Convention Center, 28 September to 1 October. The full conversations are available at Will Moroney’s interview and Mick Fennell’s interview. Temenos’s research was produced by Celent and covered in The Fintech Times’ pre-event feature.