For capital markets firms with older technology, keeping the systems running is the near-term path of least resistance, but the implicit costs are considerable.

Deepak Dhayatker, CTO at Rapid Addition, noted that a hallmark of IT systems, platforms and databases built in or around the 2000s is that they are inflexible in 2025.

“They were fit for purpose back in their day, but as time has passed, infrastructure has changed, networking has changed, and data centers have changed,” Dhayatker told Markets Media. “Most of the systems do not perform in the way that’s required for today’s client base and regulatory environment.”

“Maintenance and support costs for legacy systems are high,” Dhayatker continued. “But a bigger problem for companies that don’t invest in modernization is that they are slow to respond to business opportunities and to disruptors in their industry.”

Legacy technology is an evergreen issue for trading and investing firms, as today’s leading-edge innovation will be tomorrow’s legacy technology. However, today’s technology should be easier to shed when the time comes, as it’s typically more agile, built on open architecture, and runs off the cloud, versus the more monolithic, proprietary, and hardware-based systems of yesterday.

Wedbush Securities recently advanced its trading infrastructure, enabling a streamlining of multi-asset class trading, risk management, and post-trade processes.

“Modernizing legacy platforms isn’t just about upgrading technology — it’s about fundamentally reshaping how we deliver value to clients,” said Rodrigo Parrode, EVP and Chief Operating Officer at Wedbush Securities. “At Wedbush, we see modernization as essential to staying agile, scalable, and responsive in an increasingly complex market environment. By investing in cloud-ready, open architectures, we’re positioning ourselves to innovate continuously and build for the future with flexibility and speed.”

Rapid Addition’s Dhayatker said a rule of thumb is that a technology can be used for about a decade before it is past its prime. “Then it’s time to rewrite it, either whole or in big chunks, because everything will have changed,” he said. “There are new programming languages and interoperability standards, while underlying infrastructure and CPUs have changed dramatically. Everything is changing underneath you, so basically you cannot still be running an application today that was built 10 years ago.”

A 2021 report by the Financial Conduct Authority showed the breadth of the legacy technology problem in financial services at the time, as 33% of firms said most of their production infrastructure and applications were legacy, 58% said some, and just 8% said none. “Having a significant proportion of legacy infrastructure within an IT estate limits the flexibility of processes and prevents firms from taking advantage of new developments, release and deployment methodologies,” the FCA report stated.

To be sure, some financial services firms have updated and are updating technology, but four years after the FCA report, there is no indication that legacy technology isn’t still a huge burden for the industry.

Innovations such as those Monzo and Revolut have brought to digital retail banking are crossing over to capital markets, and investment banks saddled with legacy technology will find it difficult to provide institutional clients with services they’re increasingly aware of and asking about.

Founding Principles

“Modernization is not just moving into microservices – it is about understanding what the business needs by looking at your application as a whole, and seeing where value can be derived,” Dhayatker said. “Also, with modernization, you want to build in some founding principles. For example, in today’s architecture, observability and monitoring are first-class citizens; five or ten years ago that was an afterthought.”

High-performing trading systems need to be available all the time and be as fast as possible, while capable of handling ever-increasing volumes of data. “How do you build recoverability and scalability without compromising speed?” Dhayatker asked. “There are a few design principles that we follow quite strictly: we look at writing deterministic code from day one, and we look at using a low-level messaging transport which supports our principles, which needs to be secure, fast, deterministic, and scalable. And then we build a platform on top.”

Regarding legacy platform modernization itself, Dhayatker said the notion of a ‘big bang’ release of taking all old systems down for a weekend and starting anew on Monday is frowned upon as too risky. Instead, the protocol is to use a ‘straggler’ pattern of decommissioning small parts of the overall tech system, adding in new microservices, and then repeating. “So basically, you replace it piecemeal over time,” Dhayatker said.

Risks of a technology modernization include project risk, technology risk, rollout risk, and production risk, all of which need to be assessed, monitored and understood before, during and after the project. “These migrations are critical to get right,” Dhayatker said. “A big investment bank needs zero downtime or disruption for clients.”

Rapid Addition recently published a whitepaper entitled Modernising at Scale: Re-Platforming Legacy Trading Infrastructure. “Modernising legacy trading infrastructure is complex, costly, and risky—but the cost of inaction is far greater,” the whitepaper stated. “Success requires clarity of vision, alignment across business and technology, and a relentless focus on execution. With the right strategy, tools, and team, even the most entrenched legacy platforms can be transformed to power the next generation of capital markets innovation.”