Incumbent banks may have large pools of existing customers and brand recognition, but as they strive to compete with digital-first challengers, they also suffer from a key disadvantage: dated technology assets.
Their core banking systems are often decades old, inflexible, and laden with technical debt that prevents them optimising and innovating at a time when customers expect a modern digital experience. Fulfilling any new substantial business requirement ends up being a long, potentially multi-year programme executed in a silo. Change is delivered in long infrequent cycles, if not as a one all-or-none release with excessive manual oversight. With such a slow pace of change and high delivery risk, the business risks missing market opportunities and costing the firm their competitive advantage.
At Thoughtworks, we enable leading financial institutions to cut through this complexity and modernize legacy systems into modular, decoupled and evolutionary platforms that leverage modern cloud-based cores. This gives them the flexibility to serve digitally savvy consumers seeking secure, real-time, engaging and seamless experiences across channels, optimise their product and service portfolios, and efficiently target higher business volumes. Higher operating costs on public cloud should be measured against the benefits of lower capital expenditure, and enhanced ability to unlock revenue growth opportunities.
Modernization further needs to lay a foundation for business to build and extend its competitive advantage. This requires taking a holistic end-to-end view of the modernization journey. A clear articulation of the business value case for modernization, a directional and flexible roadmap, and domain aligned incremental slices enable both building a modernization momentum and de-risking a successful delivery. This leads to a successful legacy exit. Therefore, successful core modernization requires a holistic perspective as shown in the 'continuous alignment and optimization' diagram below.
The legacy challenge
Monolithic legacy banking systems are characterised by a tight ‘coupling’ of capabilities, making even the smallest of changes akin to open-heart surgery, with significant risk of unintended side effects. This inhibits progress, as any changes mean that systems must be subjected to long, usually manual test cycles and change approval processes.
This also lowers resilience. A fault in one domain, like loan origination, can take down others like deposits, payments, or insurance, because all of these apparently distinct capabilities rely on the same underlying infrastructure, such as processors, memory and storage. If a bug in loan repayment calculation overloads the central processing unit (CPU), other dependent workflows will be inhibited too.
There are clear commercial consequences of all of these challenges. Outages, even small ones, cause loss of marginal income from transactions. Longer outages trigger regulatory oversight with potential fines, compensation to affected customers and have opportunity cost to the bank as it diverts staff to address regulators’ concerns.
Similarly, a slower time to market means potential retardation of revenue streams because each new product or feature, or a change to existing ones has an associated revenue expectation. Further, these cause erosion of customer trust and confidence. This can result in proportional reduction in active accounts and declining new business from both existing and new customers.
While end-to-end change delivery automation can boost productivity and speed, it is hard to implement in the context of such unwieldy legacy systems because of the sprawling interdependencies. Manual processes slow the development and release of new features and products.
Last but not least, a monolithic core cannot efficiently support variations in transaction volumes. Banking is cyclical and seasonal. There are times of day, month and year when transaction volumes peak well above the average. With monolithic cores, banks caught in these cycles have limited options:
Substantially overprovision capacity to ensure all transactions are processed in real-time.
Smooth peaks by batching or queuing transactions, sometimes processing them days after they were initiated.
Build custom solutions by carving out specific modules to separate stand-alone systems, and then look at options to synchronise data between the two (especially in cases where low value, high-volume transactions start impacting the core system’s performance).
None of these choices are attractive. Overprovisioning is costly. Batching leads to poor customer experience and may have financial consequences. Creating stand-alone solutions is also costly and needs careful design to ensure data integrity across systems.
Most legacy banking cores have accumulated substantial technical debt over the years, and in some cases this can place the solvency of entire banks in question. McKinsey, for example, calculated US$2 billion in tech debt costs across more than 1,000 systems at one large North American bank. New regulations also make technical debt a potential financial liability. Reforms like the European Union’s Digital Operational Resilience Act (DORA) and the Australian Prudential Regulatory Authority’s CPS-230 require a level of operational resilience that those saddled with legacy cores will struggle to meet.
Many financial institutions have therefore layered new systems on top of their monoliths to provide customer-facing features like multi-channel apps with faster transaction capabilities. But these do little to remove the constraints of legacy technology, and may introduce higher overall complexity if approached without engineering rigour.
The cloud solution
A more optimal response is a cloud-native banking solution that is modular, decoupled, domain-driven and composable via application programming interfaces (APIs). Modularity and decoupling limit the ‘blast radius’ of defects or failures, strengthening overall system resilience. These modern banking systems support flexible deployment models such as software-as-a-service (SaaS), public cloud, and private or on-prem cloud, enabling contextually relevant deployment models.
With such systems, failure in a single service or capability cannot bring an entire bank or system down. This approach is also economical because it allows granular scaling of the most in-demand capabilities. The scope of desired change can be limited to specific services and subsystems, thus reducing the testing effort, which can therefore be increasingly automated. Last but not least, composability via APIs enables the reusing of capabilities to assemble new and varied customer journeys, further economising development and reducing time-to-market.
Banks can accelerate their engineering improvements by utilising the platform capabilities offered by public cloud providers. For example, our partner Amazon Web Services (AWS) offers step-by-step guidance to building code-light banking systems on its platform. By separating functions such as payments, loans and deposits, this allows new features to be released rapidly and independently.
And banks are not just limited to consuming the capabilities offered by cloud providers. They can tap into a range of cloud-native SaaS tools offered by third parties, most commonly modern thin core ledgers. This further reduces the heavy lifting that banks would otherwise have to perform, further strengthening their ability to focus on value for customers.
Given the potential gains, it is tempting for banks to ‘lift and shift’ legacy core systems into the cloud wholesale. Yet often this simply reproduces the same complexity in a new environment while creating additional costs and risks, as legacy code and processes consume vast amounts of cloud resources. Some may argue that it is easier to modernize to a cloud-native target once on the cloud. In our experience, it just prolongs the journey with much higher costs to the business.
Further, cost and risk of cloud-native modernization programmes are related to their scope. Many banks approach modernization with the aim to modernize and migrate the entire portfolio of products, features and capabilities. Instead, modernization should be understood as an opportunity to spring-clean your banking stack. This involves decommissioning parts of the system that are providing diminishing returns and using commodity systems for those that provide undifferentiated value. In our experience, this can substantially cut the cost and risk of large-scale modernization programmes. Teams can then focus on remediating and migrating the remaining codebase to leverage the benefits of public cloud. As a consequence, banks improve their bottom line through efficiencies from cloud-native right sizing while pushing the top line through the associated scalability, flexibility and agility.
The Thoughtworks approach
At Thoughtworks, we enable business success through excellence in strategy, technology and execution. With our Financial Services competency, we bring deep industry expertise to drive successful outcomes, and cloud-native transformation of core banking systems is no exception. We aim to deliver early and incrementally increase ROI for our clients, allowing benefits to accrue as reforms progress. This also allows early surfacing and resolution of risks. Together, both early and incrementally increasing value delivery and proactive risk reduction build a modernization momentum that is pivotal to ensure programme completion and successful legacy exit.
We achieve this by building business-aligned momentum through incremental modernization and migration. One by one, we break off capabilities and services from the monolith and move them to the cloud, modernizing each part in the process. Similarly, integrations with third party services like core ledgers and CRMs are also pursued incrementally, just enough to support the business capabilities being built. Working closely with our clients, we decide what order to migrate each increment based on business priorities. This involves weighing savings, new revenue opportunities and the risks that will be mitigated by cloud transition. This approach offers three benefits for business leadership and the IT teams:
Results are immediate. Managers and engineers can see individual capabilities and services moving to production and generating results early, building confidence and motivation to continue the migration.
Cloud costs can be actively managed. Redundant capabilities and services are decommissioned, resource provisioning is optimised and modularisation enables granular, cost-effective scaling.
Organisational culture and operating models are transformed. Thoughtworks introduces operating model transformation alongside engineering effectiveness tailored to the needs of regulated industries like banking, financial services and insurance. These enhance value, reduce friction, minimise waste and increase confidence in delivering change. This is achieved through optimising portfolios, prioritising deliverables on value and risk, and minimising manual testing, releases, handovers, and signoffs.
The long-term payoff
Before starting their transition, we encourage clients to be realistic about the financial implications of cloud migration. The process is costly and carries inevitable risks. But equally importantly, successful cloud migration can change the financial profile of a business. The capital cost of provisioning infrastructure declines, although operational costs rise in line with cloud bills. Given most corporations report earnings before interest, taxes, depreciation and amortization (EBITDA), which factors in operating costs like cloud bills but excludes capital investment in infrastructure, cloud migrations may weigh on financial performance reporting. These operating costs can only be lowered so much before impacting the ability of the bank to operate effectively.
The trade-off, of course, is a cloud-native modern core banking system that allows features to be released faster, potentially accelerating revenue streams. Further, these systems scale economically and efficiently to process greater business volumes, also leading to higher revenues. This transition is not a cost-cutting measure but rather an engine for revenue growth, enabling the quick release of new products and features, and on the bottom line, reducing the cost of serving customers along with the savings that come from risk avoidance thanks to banks’ ability to comply with regulatory regimes.
We think it’s important to work closely with our clients on cloud modernization in order to better identify ways to boost revenues using their new core banking systems, while also optimising costs and increasing resilience. At one North American bank for example, we helped add US$800 million worth of revenue by shifting the payments system to the cloud, cutting outages to zero for more than two years, and accelerating product launches.
Other benefits of cloud-native core modernization include a faster speed to market, the ability to pivot strategy in dynamic market conditions, and the potential for operational efficiencies, which also lead to improved customer experience. Thoughtworks helped Bluestone, an Australian non-bank lender, transition their business from a mainframe to a distributed Mambu-based cloud-native composable architecture on AWS in just seven months. The first loan was settled within 11 days with a saving of over 11 persons hours per day across an average of 30 loan applications per day. Bluestone leveraged this flexibility to swiftly pivot their strategy to capitalise on market opportunities. This resulted in a 25% growth for Bluestone in the quarter following the implementation with an eventual mainframe exit.
Such examples show why public cloud-native core banking should resonate with financial services companies. But banks should bear in mind another Thoughtworks maxim: focus on the ‘stickiness’ of the cloud, not just the speed of getting there. Rather than shifting legacy stacks wholesale and leaving clients with the same level of technical debt as before, we take an incremental approach to ensure modernization is genuinely transformative, laying the groundwork for new revenue opportunities and revitalising the organisation’s engineering culture.