Buy & Build 2025: Three Key Takeaways from London’s TradingTech Summit
- The Buy & Build 2025 conference highlighted that incremental modernization using a 'strangler-fig' approach—wrapping new modular components around legacy systems—is the most pragmatic strategy for capital markets firms seeking agility with lower risk.
- Data quality was identified as the ultimate ROI driver, with leading firms adopting a 'buy-to-build' approach that combines curated datasets and platform components with proprietary analytics layered on top.
- AI is already delivering measurable results in trader workflows, such as reducing bespoke TCA report generation from a week to hours, but interoperability standards like FDC3 and MCP are the real unlock for enabling AI agents to act within applications.
- Panelists emphasized that AI governance should mirror the supervision of a junior trader, with limited permissions, constant oversight, and full audit trails to give compliance teams confidence to adopt these tools.
- The overarching industry consensus is that 'buy and build' is now a philosophy of controlled evolution—firms should buy commoditized capabilities and build where differentiation matters, all within a unified, interoperable framework.
The Buy & Build: The future of Capital Markets Technology conference took place on October 2nd in London gathered the leading minds in trading technology to discuss how capital markets firms are modernizing their infrastructures balancing innovation, interoperability, and AI.
From legacy transformation to curated market data platforms and AI-driven trader workflows, three panels captured where our industry truly stands today and where it’s headed next.
Buy and Build: The Future of Capital Markets Technology
The opening panel tackled one of the industry’s hardest questions: How do we modernize trading systems built 20 years ago without breaking them?
The consensus was clear incremental modernization beats “rip and replace.”
The so-called “strangler-fig” approach, where firms wrap new modular components around old systems and phase them out gradually, remains the most pragmatic route to agility and lower risk.
Speakers emphasized that modernization is as much about culture as technology. Silos, procurement bottlenecks, and legacy mindsets slow down progress more than technical limitations. The path forward relies on open APIs, flexible architectures, and collaboration between business and tech teams.
Key takeaway: Modernization isn’t just a tech upgrade it’s a mindset shift toward agility, modularity, and continuous innovation.
Best of Breed: How to Build a Curated Market Data Platform
If modernization is about agility, data is the fuel.
The second panel focused on how firms can build curated, high-quality market data platforms that balance performance, cost, and compliance.
The panelists agreed that data quality is the ultimate ROI driver inaccurate or inconsistent data inflates operational costs far beyond licensing fees.
Leading firms now evaluate providers not just on price, but on accuracy, completeness, timeliness, normalization success, and user satisfaction.
The debate between build vs. buy leaned toward “buy-to-build”: buying curated datasets and platform components, then layering proprietary models, analytics, and insights on top. Hybrid cloud architectures are now the norm, with historical data in cloud environments and low-latency workloads still managed on-premise.
Key takeaway: Data quality, strong governance, and a “buy-to-build” approach are essential for unlocking value not just storing data, but turning it into actionable intelligence.
Transforming Trader Workflows with AI and Interoperability
The final panel explored how AI and interoperability are transforming the trader’s desktop and what that means for workflow design and compliance.
Real-world use cases showed that AI is already driving measurable results:
- Bespoke TCA reports that once took a week now take hours; trading desks use AI to automatically build meeting decks, summarize insights, and draft client notes.
But the real unlock comes from interoperability connecting every desktop app through standards like FDC3 and MCP (Model Context Protocol) so AI agents can act within applications, not just chat about them. - Speakers also addressed governance: AI should be treated like a junior trader limited permissions, constant supervision, and audit trails for every action and response. This approach gives compliance teams confidence to say “yes” to AI.
Finally, on the perennial “buy or build” debate, firms agreed on a hybrid view: buy where commoditized (e.g., connectivity), build where it differentiates (e.g., workflows, analytics, UX).
Key takeaway: AI won’t replace traders but traders using interoperable, AI-powered tools will replace those who don’t.
Final Thoughts
Across all three discussions, one message stood out:
The firms winning in trading technology are those combining modular architecture, trusted data, and AI-driven workflows under a unified, interoperable framework.
“Buy and build” is no longer a binary choice it’s a philosophy of controlled evolution: building agility, interoperability, and intelligence into every layer of the trading stack.
Horizon’s Perspective
Some parts of an electronic trading platform have become commodities rebuilding them internally rarely makes sense. But true differentiation lies in the ability to develop your own secret sauce: rapidly designing and deploying new trading strategies, adapting to market conditions, and creating unique client offerings particularly critical for brokers.
That’s why Horizon’s philosophy is centered on flexibility:
With Horizon Extend, our open framework, clients leverage all existing features of our robust trading platform.
They can then focus on building what truly sets them apart custom behaviors, strategies, and workflows without reinventing the wheel.
In a world where interoperability, data quality, and AI innovation are reshaping trading, Horizon provides the foundation for firms to move faster and trade smarter.
Frequently Asked Questions
The strangler-fig approach involves wrapping new modular components around legacy trading systems and gradually phasing out the old infrastructure. It is considered the most pragmatic route to modernization, offering greater agility and lower risk compared to a full 'rip and replace' strategy.
The buy-to-build approach means purchasing curated datasets and platform components, then layering proprietary models, analytics, and insights on top. This strategy allows firms to unlock actionable intelligence from data while balancing performance, cost, and compliance.
AI is already delivering measurable results, such as reducing bespoke TCA report generation from a week to hours and automating meeting decks, insight summaries, and client notes. Combined with interoperability standards like FDC3 and MCP, AI agents can act within trading applications to significantly enhance workflow efficiency.
Speakers at the Buy & Build 2025 summit recommended treating AI like a junior trader—with limited permissions, constant supervision, and full audit trails for every action and response. This governance framework gives compliance teams the confidence to approve AI integration across trading operations.
Buy and build is no longer a binary choice but a philosophy of controlled evolution: firms should buy commoditized components like connectivity and build where it differentiates, such as custom workflows, analytics, and trading strategies. The winning approach combines modular architecture, trusted data, and AI-driven workflows under a unified, interoperable framework.
