Part Two
Data & Digital Market Infrastructure
Fixed income leaders have high future expectations of AI. But they are clear that the data it runs on has to come first.
The foundation comes first
A third of leaders say better data integration is the single most important capability for the trading desk of 2030. Add reference data, lifecycle data and traceability, and data capabilities make up 63% of answers, while only 4% choose AI-enabled workflows.
This is not a rejection of AI but a question of order. AI can help connect data from different sources, but it cannot fix the data itself, so the foundations need to be put in place first.
Swati Bhatia
Head of Fixed Income, SIX
"Respondents ranked data integration as the most important capability for the trading desk of the future. This reflects a broader industry shift away from siloed datasets toward connected information environments that combine reference data, pricing, liquidity and lifecycle events.
High-performing desks will increasingly differentiate themselves through their ability to access a consistent and complete view of an instrument throughout its lifecycle. The competitive advantage will come not from having more data, but from having connected and trusted data that supports better decision-making."
Lars Salmon
Head of Fixed Income Trading, Fidelity
"When it comes to AI, we’re talking more about workflow performance, not necessarily changes in execution performance. I agree data integration is the most critical part, but I’m surprised the survey doesn’t show more expectation for improvement from AI-enabled workflows to bring the increasingly available data together. That’s what we mostly use it for, and I think others will as well.
AI has arguably validated our build-over-buy decision, because building is becoming so much easier. Five years ago every relevant FILS panel talked about trader having to be Python or SQL coders. That’s not the case anymore – but, of course, traders will always have to be able to interpret the data intelligently and make use of that output.”
Where AI will earn its place
Leaders expect AI to deliver the most value in risk management, meanwhile only 4% name risk management as their top strategic priority over the next two years. It suggests that firms want to improve their high-stakes risk management practices through AI, rather than as a separate project. Finding liquidity is both the top execution challenge and a leading AI use case, so leaders clearly see AI as part of the liquidity discovery solution. Notably, workflow automation which is often described as AI’s easy win ranks well behind.
AI/ML integration was already viewed as a leading approach for improving trading performance last year; the question now is where to use it, not whether.
Q: What are the biggest obstacles preventing wider adoption of AI?
Respondents were asked to select up to three options
Q: What data challenges must be solved before AI can deliver significant value?
Respondents were asked to select up to two options
Swati Bhatia
Head of Fixed Income, SIX
"One result stands out clearly: 50% of respondents identified poor data quality as a major barrier to AI adoption. That finding is significant because it suggests the industry’s challenge is not a lack of interest in AI, but a lack of confidence in the underlying data. Firms cannot expect reliable, explainable outcomes from models trained on fragmented or inconsistent information. Before AI can transform fixed income workflows, the industry must first address the quality, governance and transparency of the data that powers it.”
Ines de Tremiolles
Global Head of Trading, BNP Paribas
"It’s not so much poor data quality, as poor data architecture and infrastructure. In fixed income, the amount of data we need to store is humongous. When resources are constrained and tech teams limit how much data you can store, you end up throwing away data that is our wealth, simply because there is no efficient infrastructure to store it.”
Alec Clements
Co-Founder, PricingMonkey
Poor data quality tops the list at 50%, and AI is only as good as what it’s fed. PricingMonkey starts from clean, consistent pricing data, so the answers rest on numbers you can trust. Legacy technology and integration challenges follow, each cited by 29%. Too often, AI arrives as yet another system to bolt onto old infrastructure, with a long project before anyone sees value. We think the bar should be lower: browser-based, nothing to install, and you ask your question in plain English. That shortens the path from pilot to proof, which also helps the 42% who struggle to prove ROI."
The data barrier
Poor data quality is holding back AI adoption currently, as the most consistently cited issue by fixed income leaders.
The next barrier, proving ROI (42%), follows from the first: without good data, it is hard to measure what AI actually delivers. Despite the ongoing struggles with showing real returns of years of AI investment, budget is not a constraint for further investments. This marks the industry’s continued commitment to AI adoption.
Where Technology Could Have the Greatest Impact
Open-ended responses from this section broaden the picture considerably. Asked which emerging technology will have the greatest impact on fixed income trading over the next five years, leaders identify a wide range of developments. Three themes stand out.
1: AI Moves Closer to the Trading Workflow
AI, generative AI, agentic AI and machine learning feature heavily across the responses. Leaders see applications spanning market research, risk analysis, pricing, liquidity discovery and execution decisions.
There is also a strong emphasis on reducing the amount of information traders need to process manually. Natural language tools could help interpret news, research and issuer disclosures, while agentic AI could take on more complex processes across the trading workflow.
“Generative AI can assist with market research, summarize large amounts of information and help traders make quicker decisions while reducing manual effort.”
“AI powered decision tools can support faster trade evaluations by combining market signals, historical trends, and investor behavior.”
“Routine trading processes could be automated by agentic AI, freeing up portfolio managers to concentrate on investment strategy.”
2: Better Data Makes Technology More Useful
Several respondents explicitly connect future technology with trusted data, integration and real-time intelligence. Rather than viewing these capabilities separately, leaders describe systems that combine AI with reliable information and connected workflows.
This echoes the quantitative findings. Better data integration is already the leading capability for the trading desk of 2030, while data quality and fragmentation dominate the obstacles to AI adoption.
“The biggest impact will come from digital technologies that enable quicker, more reliable trading decisions across the investing process by combining AI, reliable reference data, and workflow automation.”
“Because traders require a single, reliable perspective of fragmented markets, integrated pricing and liquidity data platforms will become crucial.”
“Real-time data platforms will provide traders with instant access to market information, helping them react more quickly to changing market conditions.”
3: A More Connected Digital Market
The responses also extend beyond AI. Tokenisation, distributed ledger technology, cloud infrastructure, APIs, digital identity and interoperable trading platforms all feature as technologies that could reshape fixed income.
Connectivity is a recurring theme within these answers. Respondents see digital infrastructure supporting smoother movement of data between systems, easier access to liquidity and more efficient transaction processes. Tokenisation also attracts attention for its potential to improve transferability, settlement and market access.
“Digital market infrastructure developed on interoperable platforms will boost market connectedness and make liquidity more accessible.”
“Tokenization could make financial assets easier to divide, transfer and manage across different types of investors.”
“Open banking and API based connectivity create smoother links between financial systems and allow faster exchange of trading information.”
Building the Foundations for an AI-Enabled Market
Part Two shows an industry with clear expectations for AI.
This is particularly true across risk management, liquidity discovery, execution and pre-trade analysis. Turning those expectations into measurable value will require progress in the data environment surrounding those applications.
That work is already emerging as a strategic priority. Better integration leads the capabilities firms believe will matter most by 2030, while poor quality and fragmented data remain the most prominent obstacles to wider AI adoption.
"The responses reveal a remarkably consistent message. Data quality, fragmentation, standardisation and integration remain the industry’s most pressing challenges. This highlights an important misconception surrounding AI adoption: organisations often focus on selecting the right technology before addressing the readiness of the underlying data. In reality, sustainable AI initiatives depend on trusted data foundations that provide consistency, transparency and context across the entire investment lifecycle. Firms that solve these challenges first will be best positioned to realise long-term value from AI."
Swati Bhatia
Head of Fixed Income, SIX
The open-ended responses point towards an even broader evolution. AI will develop alongside real-time data intelligence, interoperable platforms, tokenisation and increasingly connected digital infrastructure. Together, these technologies could change how information moves through fixed income markets and how quickly firms can act on it.
For trading desks, the next stage will be about bringing these capabilities together. Trusted data, stronger integration and more connected infrastructure will provide the base from which AI can move deeper into everyday fixed income decision-making.
