Version française prochainement disponible
A recent visit with my mother made real for me the significant challenge banks still face in managing data.
Like most retirees, my mother wants to make sure she understands all the details of her hard-earned savings.
So, we set up a meeting with a banker to review her accounts, consolidate funds, and make some simple fixed-income investments. The day should have been informative, productive, and uneventful.
Unfortunately, some of these requests proved to be impossible, even for a senior branch representative from an FDIC-insured institution.
At one point, we needed to print a physical statement for one of her accounts. The statement was not available through the web portal. Then, a limitation in their books and records reporting system and CRM software prevented them from printing the document at the branch office.
For me, a technologist focused on the financial services sector, it was an alarm bell.
Financial services firms are under cost pressure from heightened regulatory oversight and risk controls, placing downward pressure on margins. This, coupled with growing competition from new market entrants and an increasingly diverse customer base, is driving mature firms to reassess their legacy platforms and data requirements.
While firms have spent recent years investing heavily in modernizing their front-end and mobile experiences, they have often neglected the core mainframe systems that process and record most of the world's financial transactions —including 90% of credit card transactions.1
So, while it's clear that banks still have a steep hill to climb to expand services, increase efficiency, reduce risk, and drive innovation, none of this will be possible without first prioritizing work to democratize mainframe data. Democratizing data from the mainframe can transform operational efficiencies and improve compliance by powering access to the valuable information locked in silos.
Organizational silos inhibit data sharing
Numerous factors have led to the current state of data. For starters, most large financial institutions are a collection of diverse businesses created over the past century through M&A activity.
These businesses have built massive and diverse organizations that are highly effective, but in many cases, siloed from each other. The organizational structure has contributed to a complex technology and data stack that makes data sharing across different silos difficult—and, in the case of my mother’s account, impossible for front office staff to access.
In addition, traditional Role Based Access models for data access are proving to be less responsive to business demands than more modern Attribute-based Access Control models, where data and system access is based on business rules instead of more siloed user permissions.
Democratizing mainframe data can transform operational efficiencies and improve compliance by powering access to valuable information locked in silos.
To address the silos, banks have built expansive data warehouses that enable them to consolidate, organize, and analyze data from multiple sources. And while data warehouses have helped institutions in areas like compliance, customer relationship management, and risk, there have been limitations due to factors including data quality, integration, scalability, and governance.
Integrating mainframe data remains a struggle
The power of a data warehouse is its ability to consolidate data from across the organization. Many of our clients, however, struggle to integrate mainframe data.
It’s difficult to diagnose a technical issue while sitting in a branch office in Boca Raton in front of a frustrated customer representative (and my mother). But I assume the product records associated with my mother’s investments were stored on a mainframe and were probably misclassified. That’s why they weren’t available to the reporting systems that produce customer financial statements.
Mainframe datasets exist in many different formats. There are files with fixed and variable length, DB2 and IMS databases, Virtual Storage Access Method (VSAM) files, and COBOL Copybooks, which are sections of code that define the data structure within a COBOL program.
Copybooks are particularly complex as they are metadata blocks that define the physical layout of data but are stored separately in the platform. But, modern data warehouse tools like Hadoop do not understand Copybook data structures, making data conversions difficult.
Other data challenges facing financial institutions
Other data challenges can also create barriers financial institutions, including:
- The lack of single client or legal entity identifiers make it difficult to get a complete view of the client’s holdings, activities, and profitability.
- Gaps in client profile information may cause servicing issues and pose a regulatory risk.
- Inconsistencies between local and global asset and product classifications could impact advisory and reporting functions.
- Different position-keeping standards among regional offices, like time stamps associated with trade reports (trade date vs. settlement date), introduce reporting risk.
Freeing these critical datasets to support business and customer demand will be a key factor in a bank's successful transformation toward a data-driven business model.
Books and records: some of the most siloed datasets in the bank
Books and records, long considered core data in financial services, refer to the financial and operational records that an investment firm maintains in connection with its business activities. These records include client accounts, financial statements, trading records, and other documents that provide evidence of the firm's transactions and operations.
Brokerage firms are required to maintain accurate and up-to-date books and records as part of their regulatory obligations. They are required to provide this data to regulatory authorities upon request to ensure legal compliance.
While books and records data is considered some of the most critical within the organization, it is often the most siloed and least accessible.
Why? Cost and risk.
Programs to expand access to these core systems are often deprioritized as they provide limited ROI to the IT organization charged with transforming and operating the platform.
While the data in books and records is considered some of the most critical in the organization, it is often the most siloed and least accessible.
Since books and records platforms are considered critical, any change will require senior-level approval. Leadership will ask about risk versus reward, which may be difficult to quantify. Should the project team fail to provide clear monetary benefits, an integration project may not receive the necessary approvals.
However, as firms drive transformational digital strategy, access to these core datasets is critical. The ability to access them will be a key factor in successfully navigating shifting market forces over the next decade.
New banking regulations increase need for data access
Regulators in the U.S. have been stepping up oversight to protect banking clients and reduce risk to the economy. To reduce risk in the clearing and settlement of securities, the SEC passed sweeping regulation to shorten the settlement cycle for most securities transactions from two business days after the trade date (T+2) to one business day (T+1) by May 2024.2
Given the massive regulatory drivers and the enormous business value of the data locked in these core books and record platforms, they represent an excellent place for banks to start a data democratization journey.
Bank customer satisfaction is also at stake
New regulations are not the only challenge facing banks avoiding their data democratization journey. Banks that are slow to democratize will feel a growing impact on multiple fronts, including with their customers.
For my mother, data silos on the mainframe prevented a branch representative from producing the documentation needed to complete her most important transaction. After more than an hour, we decided that closing the account and consolidating her money with a competitor was less risky than waiting for the organization to resolve the issue.
Fortunately, banks can avoid disrupting the customer experience and improve operations innovation by prioritizing data democratization, starting with books and records.
Robert Wallos is a Financial Services Technology Strategist at Kyndryl.
1 Are Banks Breaking Up With Mainframes. Forbes. March 2023
2 SEC Finalizes Rules to Reduce Risks in Clearance and Settlement. SEC. February 2023