Fintech Futures: Are Legacy Core Systems the Banking Equivalent of “Kodak” and “Blockbuster”?
- The days of overnight processing in the consumer banking sector are over
- About 50% of banks, however, still use overnight batch processing
- Real-time data processing is now not only recommended, but crucial
At Fintech Futures, banking industry veteran Dharmesh Mistry argues that that many banks still rely on outdated legacy core banking systems that do batch overnight processing -- once considered very fast and efficient -- rather than engaging truly real-time analysis.
It’s hard to quantify exactly how many, but a significant number of banks – I would guess at least 50% – are still running on legacy core banking systems that have overnight batch processes. These batch processes update ledgers with charges, interest and transactions that could not be processed in real time or near real time.
Mr. Mistry puts forward reasons he believes some banks are lagging behind, mostly related to the fact that their systems were created when tech was still in development:
- Transactions were carried out in branches so were built to handle the branch staff user base, not the creation of transactions by customers.
- Most transactions used cash, so a significant amount of processing was not done on the core banking system at all.
- Technology and especially storage was expensive and it was cheaper to process transactions in volume overnight when branches were not operational. At this time, branches were connected to central mainframes using expensive private networks, too.
- There was no real-time settlement between banks.
Time for Real Time
Real-time data processing is essential for advanced personalization, fraud detection, and engaging conversations with AI assistants.
Mr. Minstry notes:
Aside from the huge volume increases and the expectations of consumers that banking is 24x7x365, the act of processing transactions has become much more complex. In the very early days, you could only transact what you had in your account. Later you could have overdraft limits based on an offline credit assessment. These days, more advanced platforms are constantly assessing your income/spend to not only understand your risk, but to check for fraud, anti-money laundering and many other processes to meet compliance and reporting requirements.
In other words, the "good old days" of a multi-day "transaction cushion" are over. He concludes:
This week, I’m just saying that it really is time for banks to get ‘real time’ with their core banking solutions. I truly believe that failing to move to the next generation of real-time cores and analytics could be just like the Kodak and Blockbuster moments that were defined by the transition to digital.
Mr. Mistry makes quite a compelling comparison with Kodak and Blockbuster's inability to adopt the latest technologies to stay relevant -- albeit a bit extreme, in our opinion.
However, we do agree that legacy systems are holding banks back from achieving tangible goals, like real-time processing. For instance, technologies for check processing have advanced with the deployment of AI and machine learning, enabling banks to process checks in real-time -- with accuracy and read rates over 99%+ -- effectively automating the process. However, there are still banks out there that rely on legacy OCR technology requiring manual intervention because of their 80-85% accuracy and read rates.
As banks continue a path towards modernization, it's critically important that they are not the last adopters of the latest technologies -- or they could become the next Kodak or Blockbuster.