The Evolution of Electronification in Fixed Income: A Conversation with UBS’s Deniz Mace-Jones

Over the past decade, fixed income markets have undergone a fundamental shift. What was once dominated by voice trading, spreadsheets, and manual booking is now increasingly digital, data-driven, and automated. To understand where we are and where we’re heading, I sat down with Deniz Mace-Jones, Global Head of Rates and Credit Products and Electronic Distribution at UBS, for a candid discussion on the realities of electronification, workflow transformation, and the role of AI in the next phase of market evolution.

From Fragmentation to Automation

When Deniz first started her career, paper blotters were still commonplace. Trades were recorded by hand, calls were constant, and data was scattered. Fast-forward to today, and markets like Euro government bonds are 80–90% electronic. Even interest rate swaps, once viewed as too complex for electronic execution, are now largely traded electronically.

That progress hasn’t happened by chance. Deniz attributes the acceleration of the past five years to three main drivers:

1. Regulation: Reforms like Dodd-Frank and MiFID II pushed more trading on-venue, reshaping market structure.

2. Client adoption: Firms tested and refined technology that worked, driving organic adoption.

3. Platform innovation: New trading protocols, portfolio trading in credit being a prime example, changed how the market operates.

As she put it, what we call “illiquid” or “manual” today will likely look very different in just a few years’ time.

The Liquidity Divide

In highly liquid areas like Euro government bonds, automation is now the default. In contrast, illiquid products, high-yield credit, and certain currencies in interest rate swaps remain more voice-driven.

Liquidity dictates workflow. Where markets are deep and prices are standardised, automation thrives. Where trades are bespoke or infrequent, the human layer remains essential.

The Human Edge in an Automated World

For Deniz, automation is not about replacing people; it’s about redeploying their focus. In her early days, the bulk of her effort was devoted to checking trade bookings and reconciling blotters. That effort, she notes, should now be redirected toward client value.

“Automation should remove the BAU work,” she said. “That energy is better spent delivering insight, content, and market colour.”

This echoes a broader shift I see across the industry: human input is becoming more valuable precisely because technology is handling the repetitive mechanics.

Chat Data, The Unused Goldmine

Roughly 90% of OTC fixed income communication now happens through chat platforms such as Bloomberg. Deniz described this as the next frontier: the vast reservoir of unstructured data hidden in chat transcripts. The challenge is making that information accessible and actionable, turning conversation into intelligence.

That’s an area where ipushpull and others are now innovating: connecting structured and unstructured data to create a seamless, live workflow between systems and humans.

Building on Strong Foundations

Deniz made a critical point: “You can’t build the spaceship without the engine.” The most advanced front-end tools are worthless without solid data plumbing behind them. Future-proofing infrastructure, clean data, reliable pipelines, and standardised formats is what will separate short-term success from long-term resilience.

Buy vs Build: The Pragmatic Approach

The days when every bank built everything in-house are fading. Deniz compared it to a builder: no one makes their own tools. The key is knowing what to own (your IP, your differentiation) and what to source externally (the utilities that enable scale).

The buy-versus-build decision, she noted, must always be made case-by-case, informed by client needs, strategic direction, and long-term flexibility.

AI’s Role in Sales and Trading

AI is everywhere in the conversation today, but Deniz’s approach is measured. UBS is experimenting broadly, from chatbots querying large data tables to internal AI tools that streamline repetitive tasks. But she’s clear: it’s about targeted application, not hype.

“Just because you can, doesn’t mean you should,” she said. The goal is efficiency and insight, not novelty for its own sake.

Digitised vs Electronified

One of the most important distinctions Deniz made is between digitisation and electronification. A chat-based trade is digitised; it’s happening through digital channels, but not necessarily electronified, which implies structured data flow and automated execution.

Every market will become digitised. Full electronification depends on how standardised and replicable the product is. Government bonds and vanilla swaps fit that profile. Bespoke structured products likely won’t, at least not yet.

The Road Ahead

Electronification in fixed income is not a destination but an evolution. Regulation, infrastructure, and technology will continue to push the boundaries, but the differentiator will remain how intelligently firms apply them.

As Deniz summed up, markets that master both human judgment and machine efficiency will lead. That’s the next competitive edge, and it’s already taking shape.

Want to discuss a chat automation project? Contact us to chat further.