logo

Can finance stay connected as geo-economic risks rise?

Add The Asian Banker on Google
Discover more trusted banking and financial services insights by adding The Asian Banker as a preferred source on Google.
Can finance stay connected as geo-economic risks rise?
  • 82

At the Singapore FinTech Festival 2026, policymakers and practitioners argued that banks must build interoperable routes for moving money as geo-economic pressure turns dependencies into leverage and AI changes who can initiate transactions.

Control over chips, data and payments has become a tool of statecraft even as financial institutions build cross-border links. At the Singapore Fintech Festival 2026 opening plenary, Ravi Menon, chairman of the Global Finance and Technology Network, argued that resilience depends on creating alternatives that can still connect.

The discussions that followed—from payment infrastructure to national AI missions and European central banking—showed how shared standards, public-private cooperation and clear accountability could help preserve those links.

Alternatives matter when they remain connected

Menon distinguished geo-economics from geopolitics. Where geopolitics pursues power through armies and alliances, geo-economics uses markets, money and machines. Its tools range from tariffs, sanctions and export controls to trade agreements, common standards and digital corridors. Menon said governments are reaching more often for coercive measures than cooperative ones.

The contest is visible across trade, investment, supply chains, data and finance. Advanced chips and chipmaking equipment have become strategic goods. Export controls can restrict access, but they can also encourage efforts to develop substitutes. Menon pointed to Chinese domestic accelerators and open AI models as evidence that restrictions can prompt a target to seek alternatives.

Investment screening is expanding, too, with governments assessing inbound deals for security risks and outbound investment for possible technology leakage. Menon cited the Netherlands’ intervention at chipmaker Nexperia: control changed, but the company remained dependent on Chinese plants for packaging, and automakers warned of shortages.

AI’s supply chain is concentrated and cross-border. A financial transaction in Singapore might rely on a chip fabricated in Taiwan, a model trained in Texas, power generated from Qatari gas and a cable crossing the South China Sea. “Every link is a geo-economic exposure,” Menon said.

The technology stacks led by the United States and China are not fully separate. A data centre in South Asia could run on American chips while serving Chinese open models. Hardware controls have little reach over software that can be downloaded across borders. Menon also described a divide in data regimes, with the US approach market-first and China’s security-first.

Finance is another source of leverage. The dollar remains dominant, including through dollar-denominated stablecoins, but Menon said alternatives can reduce the force of financial coercion without matching the dominant system’s scale. Citing IMF research, he argued that coercive power falls non-linearly: a target has few options when one system has near-total dominance, but even a smaller alternative can give it room to act. The aim, he said, is targeted diversification at key points in the system—not de-dollarisation or decoupling.

For Menon, this is not the end of globalisation. Trade restrictions have increased alongside trade volumes, while supply chains have been reconfigured rather than withdrawn. The task is to make those connections more resilient through coalitions of willing partners and public-private cooperation where multilateral agreement is difficult.

Payment networks need to connect at every layer

Menon’s example of constructive geo-economics was cross-border payments. Bilateral QR links already let travellers pay through domestic apps while merchants receive local currency, reducing reliance on correspondent banking or international card schemes for some retail transactions. Project Nexus aims to extend this model: a shared gateway would connect national instant-payment systems across five ASEAN countries and India, with a live launch targeted for 2027.

Singapore’s PayNow Generation 2 roadmap addresses the domestic foundations. The Monetary Authority of Singapore (MAS) and the Association of Banks in Singapore plan to introduce SGQR+ in stages from 2027, allowing residents and tourists to pay with their preferred QR app at participating merchants. The roadmap also includes deep-linking for online payments and richer data to support automated reconciliation. A Future Payment Rails Blueprint, due in 2027, will set out how the system can support more connected and programmable payments.

Connecting retail payment networks is only one layer. Menon argued that common standards are also needed so regulated digital money issued in one jurisdiction can settle in another. Singapore’s Project Guardian is testing tokenised finance, including standards work for fixed income and foreign exchange.

A panel featuring Swift chief executive Javier Pérez-Tasso, HSBC co-chief executive for Asia and the Middle East Surendra Rosha, and Ant International chief executive Yang Peng discussed how those connections might work in practice. Swift is developing faster account-to-account and account-to-wallet payments on existing fiat rails, alongside a shared ledger for tokenised bank deposits.

Ant uses tokenised deposits and a settlement token to address cases where banks do not recognise one another’s deposit tokens. Peng said businesses operating around the clock can face limited settlement windows and may need to hold extra working capital.

The panel’s central point was that these systems need to work together. “The ‘versus’ needs to be replaced with an ‘and’,” Pérez-Tasso said, referring to banks and fintechs, public and private infrastructure, and multiple payment rails.

AI agents could make those connections more useful by handling currency conversion and compliance checks as they move between networks. Cross-border use will require common standards for agent identity, authority and transaction limits. Rosha stressed that banks remain accountable for actions taken through their systems.

For banks and payment providers, interoperability is therefore an operating and governance task: connect rails and settlement assets, verify which agent is acting, set its authority and retain accountability for the transaction.

Shared infrastructure can support autonomy

Singapore’s financial AI missions showed how governments and industry might organise adoption around problems individual firms cannot solve alone. Deputy Prime Minister and MAS chairman Gan Kim Yong announced work on financial crime prevention, SME financing and safe agentic transactions.

Five banks are pooling transaction data securely in a proof of value to identify scam patterns that no single institution can see. The findings will inform the design of an industry-wide detection utility. A second mission will test whether richer data and AI can improve lending assessments for SMEs with limited credit histories or collateral. The third will examine how agents can use purpose-bound money within defined limits. The Future of Finance Institute will bring industry, researchers and public agencies together to develop practical solutions and shared standards.

Gan linked the missions to Singapore’s ambition to become a leading place to deploy AI. Kampung AI aims to attract practitioners, while Punggol Digital District will serve as a testbed for applications including robotics. His emphasis was on building trust, attracting talent and preparing workers so businesses and people can use AI responsibly.

Europe faces a related challenge, but across many states: preserving autonomy while sharing infrastructure. A panel featuring José Luis Escrivá, governor of the Bank of Spain; Joachim Nagel, president of the Deutsche Bundesbank; and Ulrik Nødgaard, governor of Danmarks Nationalbank, discussed how to balance innovation, integration and resilience.

Escrivá said the Eurosystem was preparing infrastructure and experimenting with new technologies while regulatory consensus was being built across the European Union’s 27 member states. Nagel countered the idea that Europe only regulates, pointing to the Markets in Crypto-Assets Regulation as a framework that gives businesses direction and to central-bank work on tokenised finance.

Nødgaard explained why Denmark chose to join European payment infrastructure while retaining its own currency. Monetary sovereignty, he said, is distinct from the technical rails used to move money. For a small economy, shared systems offer scale, cybersecurity expertise and capacity to innovate.

Shared infrastructure does not eliminate the value of domestic alternatives. Denmark’s instant-payment system is widely used for transfers but has yet to gain ground in physical stores, where cards and international wallets dominate. Nødgaard said the central bank hopes to make it possible to tap a phone and pay directly from a bank account, giving consumers another route if card networks are disrupted.

The panel also discussed the digital euro as a digital counterpart to cash and a way to preserve a role for central-bank money as payments move online. Central-bank money could also provide a trusted settlement asset for tokenised markets. AI adds risks for central banks to monitor: agents may make deposits easier to move between providers or increase herding if they make similar trading decisions. Shared systems can improve efficiency, but they also change the risks institutions must manage.

AI’s gains must reach beyond the technology sector

The case for AI depends not only on the power of the technology but on how widely it is adopted. In a conversation with Nobel laureate Michael Spence, Menon returned to the “social license” for AI: public acceptance will depend in part on what happens to jobs and wages.

Spence said productivity gains would not come from frontier models alone. Businesses need to experiment with workflows, business models and skills before AI changes performance across the economy. He described a J-curve: investment can weigh on productivity at first, with gains emerging later if adoption and organisational change take hold. If AI use remains concentrated in technology and a few other sectors, its broader economic effect will be limited.

The employment effects are also unsettled. Menon described automation as a force that can displace jobs and augmentation as a way to raise productivity, wages or demand for new work. The same technology can remove entry-level roles or help less experienced workers become more productive. Menon cited a US estimate that employment among 20- to 25-year-olds in the most AI-exposed occupations had fallen about 13% relative to peers since generative AI took off. He also cited a separate customer-service field study in which productivity rose about 15%, with newer hires gaining the most.

Spence argued that firms still need to develop junior talent, even if AI changes how people enter professions. He expects people and machines to work together: agents may save time, but employees remain responsible for checking outputs and accountable for decisions. The transition will affect many jobs even where it does not eliminate them.

Gan acknowledged workers’ anxiety about that transition. “We may not be able to protect every job,” he said, “but we do intend to protect every worker.” That means helping people build skills for changing roles and new opportunities. For children entering an uncertain future, he emphasised curiosity, continued learning and the ability to ask good questions.

The plenary set out two tests for the next phase of financial integration. Shared standards have to give institutions and customers more ways to transact across borders. The gains from AI have to reach businesses and workers beyond the technology sector if adoption is to earn lasting public support. Menon described the work of building those connections as proceeding “standard by standard, corridor by corridor, and now agent by agent.”