Core Banking Technology, APIs and Open Banking

A bank is, at bottom, a ledger that the law and the public trust. Every deposit, loan, card swipe, interest accrual and fee is an entry in that ledger, and the bank's promise that "your money is here" is only as good as the systems that record it. Most of those systems are old: many of the largest US banks still run their core accounts on mainframe software whose design dates from the 1970s and 1980s, processed in overnight batches and written in COBOL. Around those cores, banks have added decades of channels, payment hubs, card processors and data warehouses, and since the 2010s a layer of APIs that lets fintechs connect to them. This plumbing decides what a financial company can build, how fast it can change, what it costs to serve a customer, and what happens when something breaks. When it breaks, the consequences are not a slow website but frozen salaries and missing money: a 2015 software failure at a BNY Mellon vendor left ~1,200 mutual funds and ETFs unable to strike accurate prices for days; a Citibank operations error sent about $900 million to Revlon's lenders in 2020; and the 2024 collapse of the banking-as-a-service middleware firm Synapse left tens of thousands of fintech customers' funds frozen and a reported shortfall of roughly $65–96 million that no party's ledger could explain. This section covers the banking technology stack, core systems and their vendors, core modernisation, ledgers and reconciliation, payment rails and messaging, BaaS architecture, APIs and open banking (the US Section 1033 rule and its 2024–26 reversal, with the UK and EU regimes that shaped it as comparisons), operational resilience and cloud, key numbers, cases, pitfalls and expert framing.