Much of the banking sector still relies on so-called ‘legacy systems’. These systems often remain due to the mindset at many banks of, ‘if it works, why change it?’
Moreover, migrating from legacy systems is no easy task.
It’s therefore important to ask why these systems remain, what it takes to modernize them, and what the consequences are of failing to do so, especially in this new age of AI.
Many systems used by large banks today were originally designed and developed in the 1970s, 80s, and 90s, leveraging the cutting-edge technologies of the time.
“These systems were meticulously tailored to meet the requirements and demands of that era,” explained Atmaram Parameshwara, director for digital at Synechron, a New York City-based consultancy and advisory firm, which primarily focuses on driving digital strategies in the financial services space.
However, as time has progressed, both technology and business needs have evolved significantly. “Despite this, one aspect has largely remained unchanged: the reliance on these legacy systems,” Parameshwara said.
There are a number of reasons why they remain in place. Primarily though, these systems are very resilient and have proven their reliability by handling massive transaction volumes and maintaining stability over many years, he noted.
So what does it take to modernise?
“Modernisation hasn’t been easy and has required significant efforts from organizations, but it is essential,” Parameshwara stated. “And now with AI, it’s become easier to update legacy systems.”
He is convinced modernisation requires clear commitment and strategic alignment from leadership. “Secure buy-in from all your stakeholders to ensure unified support and commitment.”
“Some finance organisations lack a clear roadmap for modernisation and fail to allocate the resources.”
– Atmaram Parameshwara
Moreover, comprehensively assessing systems is also crucial, thinks Parameshwara. “Evaluate legacy system, and business needs, then identify the appropriate alternatives.”
AI algorithms, for example, can analyse large volumes of data more efficiently than traditional methods, he pointed out.
“This enables legacy systems to derive insights and make data-driven decisions, improving overall performance. Modernisation is a complex process that demands meticulous planning, precise execution, and ongoing evaluation,” Parameshwara continued.
“It requires support and alignment from every level of an organization. Sometimes it also requires a realignment of strategies, processes, and organizational priorities to ensure success.”
‘Tricky to replace’
Parameshwara stressed that replacing any legacy systems is not easy since there are “a number of complexities involved”, such as high replacement cost, migration complexity, custom-built solutions, dependency on existing ecosystems and a lack of strategic vision.
“Some organisations lack a clear roadmap for modernisation and fail to allocate the resources or leadership needed to initiate and execute such projects,” he remarked.
Then there is also organisational resistance, which should not be underestimated, Parameshwara noted.
“Changing legacy systems within organisations can be challenging, as teams are often resistant to adopting new systems or processes. This resistance sometimes originates from the top of an organisation,” he said.
In fact, “modernising core systems can take years to complete, and long timelines can discourage organisations from starting the process.”
However, at the end of the day, banks and other financial institutions simply have no choice as the consequences of not modernising are significant.
“Modernisation is not mandatory, banks can continue to operate using legacy systems, given their proven reliability in supporting essential operations, so, why should they modernise?”
Parameshwara stressed that not modernising always incurs higher costs over time, compared to the investment required for modernisation.
For example, building an API for external integration on a legacy system often requires additional components from the vendor, leading to increased costs, he pointed out.
“Clinging to legacy systems can prove even costlier for big banks in the long run.”
– Atmaram Parameshwara
Additionally, “the implementation process is time-intensive and demands significant manpower. In contrast, building an API on a modern system is typically faster, requiring minimal effort and resources,” he said.
“Making changes to a legacy system is inherently more expensive, whether it’s for new implementations, scaling, or addressing regulatory and compliance updates.”
Moreover, another factor that plays a role is that finding skilled professionals proficient in legacy systems is becoming increasingly difficult. As a result, hiring and retaining such talent has grown progressively more costly over time.
“There is growing competition from emerging FinTechs that leverage modern technologies and streamlined processes to attract customers with innovative products, superior user experiences, and cost-effective solutions,” Parameshwara observed.
“Achieving this with legacy systems is a monumental challenge, given their limitations in scalability, flexibility, and integration with modern technologies and systems.”
Finally, one of the main reasons organisations hesitate to modernize and continue relying on legacy systems is the fear of disruption, high costs, and the uncertainty associated with large-scale changes.
“But, clinging to legacy systems can prove even costlier for big banks in the long run,” Parameshwara summarised.
“It risks inefficiencies, higher maintenance costs, and the possibility of them being outpaced by agile FinTech competitors – or becoming obsolete altogether,” he concluded.
Why not become a QA Financial subscriber?
It’s entirely FREE
* Receive our weekly newsletter every Wednesday * Get priority invitations to our Forum events *
THIS MONTH
WATCH NOW
READ MORE
- Goldman puts AI coding to the test
- How to test AI models that banks do not control
- OpenAI, Filigran and SunTec: the latest vendor and product news
- Sygnum: Testing AI is ‘a measurement problem’
- Banks’ ‘code for all’ push raises testing risks
