(May 26): ING Groep NV is turning to “vibe coding” to build electronic trading tools for currencies and credit, its latest use of artificial intelligence to compete with larger peers.
“Vibe coding” — telling AI what you want to create so it can do so, rather than coding manually — is compressing work into hours that may have taken a team of developers weeks, said Simon Bevan, its global head of electronic trading. Recently the Dutch lender has used it to create analytics dashboards to show real-time pricing, incoming trades and performance metrics.
“We recently used vibe coding quite significantly to build an entire credit e-trading system,” Bevan said in an interview. The firm uses external models and has found the most success with Anthropic PBC, he said.
With fewer resources than some global rivals, ING is doubling down on technology to win more business, and Bevan credits an in-house AI currency pricing model for an initial 50% increase in big-ticket trades. Yet it could be a bold bet to increase its reliance on AI in a division overseeing “tens of billions of dollars” of trading a day, in case of flawed output or unforeseen behaviour.
“We’ve also started to use it in actual trading systems where that has a lot more risk,” he said.
Rapid developments in AI are reshaping the financial industry and risk creating winners and losers, depending who can best leverage technology. While many banks have been vocal about the benefits for their back and middle offices to save costs, there’s been less said on the impact for front office trading, often the profit centre.
See also: AI rally drives best run in global momentum stocks on record
The race to automate can come with hefty costs and it may be harder for smaller players to fund the kind of computing power needed to stay competitive. XTX Markets, a quantitative-trading firm that has no human traders, last year invested more than €1 billion building five data centres in Finland to underpin its growing use of machine learning.
“Some of the bigger banks can just throw money and people at these projects,” Bevan said, adding that his team of 10 have to be more creative with strategy. “We don’t have huge numbers of people.”
The potential for AI leading to job losses has become a hot issue. Standard Chartered plc’s CEO Bill Winters apologised for his comments last week about how AI will affect “lower-value human capital” after the remarks spurred a public backlash.
See also: Anthropic to close over US$30 bil round as soon as next week — Bloomberg
ING will cut about 1,250 operations roles globally in 2026 for efficiency as it responds to digitalisation, AI and evolving customer needs. Bevan said AI is reshaping the role of traders and the types of people they want to hire, with the future workforce likely to be focused around fewer “super” employees supervising complex systems.
“We’ve seen over the years it’s become more and more about tech,” he said. “Twenty years ago you’d have expected a charismatic trader or sales person. Now the markets are full of geeks with computer science or physics PhDs.”
Uploaded by Arion Yeow
