In 2022, Standard Chartered needed a digital-only bank for the Hong Kong market. Rather than spend three to five years rebuilding its own core systems, it partnered with Thought Machine, a fintech company founded in 2014, and launched Mox Bank on Thought Machine’s cloud-native platform in a fraction of that time. That partnership model, a traditional bank using fintech-built infrastructure to modernise, is now the dominant pattern in banking technology. The global banking-as-a-service market, which supplies much of this fintech infrastructure, reached $18.6 billion in 2024, according to Global Market Insights, growing at 15.1% annually toward $73.7 billion by 2034.
Why Banks Cannot Modernise Alone
Most large banks run core systems built between 1975 and 1995. These mainframes use COBOL, a programming language from 1959, and process transactions in overnight batches. They work. They have worked for decades. But they cannot do what modern banking requires: process payments in real time, expose data through APIs, integrate with third-party services, or launch new products without months of custom development.
Replacing a core banking system is the most complex technology project a bank can undertake. Every product the bank offers (deposits, loans, payments, cards, trade finance) connects to the core. Every regulatory report draws data from it. Every branch and every ATM depends on it. Changing the core while the bank continues to operate is comparable to replacing the foundation of a building while people are still working on every floor.
The Boston Consulting Group projects fintech revenues will reach $1.5 trillion by 2030, with embedded finance and digital lending accounting for the largest share of projected growth.
According to CB Insights’ 2024 fintech report, global fintech funding declined 40 percent between 2022 and 2024, pushing the sector toward consolidation and a sharper focus on profitability over growth at all costs.
Banks tried for years to do this with internal teams and consulting firms. The results were poor. IBM, Accenture, and Infosys ran multi-year, multi-billion-dollar core modernisation programmes at major banks throughout the 2010s. Many ran over budget, over timeline, or were abandoned entirely. The complexity of legacy migration exceeded what traditional technology vendors could manage.
What Fintech Companies Built Instead
Fintech companies approached the problem differently. Rather than trying to migrate existing systems, they built new core banking platforms from scratch using modern cloud architecture. Thought Machine (founded 2014, London), Mambu (founded 2011, Berlin), and 10x Banking (founded 2016, London) each built cloud-native cores designed to run on public cloud infrastructure (AWS, Google Cloud, Azure).
These platforms share several characteristics that legacy systems lack. They process transactions in real time, not in overnight batches. They expose every function through APIs, allowing banks to connect new services in days rather than months. They scale automatically with transaction volume, eliminating the capacity planning that mainframe operations require. And they use modern programming languages (Java, Go, Python) that today’s software engineers actually know, unlike COBOL, where the average practitioner is over 55 years old.
Platform-based banking-as-a-service models account for 69% of the BaaS market, according to Global Market Insights, and cloud deployment holds a 67% share. Those numbers reflect the industry’s direction: the future of core banking runs on cloud platforms built by fintech companies, not on mainframes maintained by in-house teams.
The Partnership Models
Banks and fintech companies work together through several distinct models, each suited to different levels of ambition and risk tolerance.
The first is full core replacement. A bank migrates its entire operation from a legacy mainframe to a fintech-built cloud platform. This is the most ambitious approach and typically takes two to four years. JPMorgan Chase, which spent $17.1 billion on technology in 2024, is pursuing elements of this strategy with internal builds supplemented by fintech partnerships.
The second is the sidecar model. A bank launches a new digital product or sub-brand on a fintech platform while keeping its legacy core for existing operations. Standard Chartered’s Mox Bank is an example. So is Goldman Sachs’s Marcus, which runs on a modern stack separate from Goldman’s institutional trading infrastructure. This model limits risk because the legacy system continues to serve existing customers while the new platform handles growth.
The third is API-layer modernisation. A bank places an API gateway on top of its legacy core, allowing it to connect to fintech services without replacing the underlying system. This is the fastest and cheapest approach, but it creates a dependency: the legacy core remains, and every API call must translate between modern and legacy formats. Over time, this layer becomes its own maintenance burden.
Banks globally now process over 2 billion API calls daily, handling $676 billion in transaction value, according to Coinlaw. That volume reflects how deeply fintech-built API infrastructure has penetrated traditional banking operations, even at institutions that have not yet replaced their cores.
Specific Areas Where Fintech Is Modernising Banks
Beyond core banking, fintech companies are modernising specific banking functions that have resisted internal improvement.
Payments infrastructure is the largest category. Traditional correspondent banking for international transfers involves multiple intermediary banks, each taking a fee and adding processing time. Fintech companies like CurrencyCloud (acquired by Visa), Wise, and Banking Circle have built direct connections to local payment networks in dozens of countries, enabling banks to offer faster, cheaper cross-border payments. The global cross-border payments market reached $371.59 billion in 2025, according to Fortune Business Insights, and fintech infrastructure handles a growing share of that volume.
Identity verification and KYC (know your customer) processes have been transformed by fintech providers like Onfido, Jumio, and Sumsub. These companies use machine learning to verify identity documents and match them to biometric data in seconds, replacing manual review processes that took days. Banks integrate these services through APIs, adding them to customer onboarding flows without building the technology themselves.
Fraud detection has shifted from rules-based systems (if a transaction exceeds $10,000 from a new country, flag it) to machine learning models that analyse hundreds of variables in real time. Fintech companies like Featurespace and Feedzai provide these models as cloud services that banks connect to through APIs. The models learn from each bank’s transaction patterns and improve continuously, catching sophisticated fraud that static rules miss.
The Risks of Fintech Dependence
The bank-fintech partnership model creates dependencies that regulators are increasingly concerned about. When a bank relies on a fintech company for its core processing, a failure at the fintech can disrupt the bank’s operations. The Synapse collapse in 2024 illustrated this risk: when the BaaS middleware provider failed, thousands of customers temporarily lost access to their funds.
The EU’s DORA regulation, effective January 2025, addresses this by requiring banks to assess and manage the risks of all technology partners in their supply chain. Banks must demonstrate that they can continue operating if a fintech partner fails, which means maintaining contingency plans, backup systems, and the contractual right to access their own data.
Fintech companies are helping banks modernise faster than banks could modernise themselves. The trade-off is a new form of operational dependency that both parties, and their regulators, are still learning to manage.
