Banking
IS GOOD DATA GOVERNANCE THE ANSWER TO MANAGING THE NEW UNKNOWN RISK IN BANKING?
Published by Gbaf News
Posted on May 10, 2017
7 min readLast updated: January 21, 2026
Global Banking & Finance Review®
Company
Published by Gbaf News
Posted on May 10, 2017
7 min readLast updated: January 21, 2026
Lee Thorpe, Head of Risk Business Solutions at SAS UK and Ireland
Last month the European Risk Management Council launched its first flagship event, the UK Risk Management Leadership meeting, which brought together leading executives to discuss some of the most important topics dominating the risk management agenda. One of the things that it highlighted was that nothing is really changing within the risk community. Risk is calculated from historical data and those same risks that executives have been worrying about for several years – low interest rates, challenger banks, regulatory pressures and cyber threats – continue to remain front of mind.
Cultural change
A cultural change is needed that encourages the risk industry to have a more forward-thinking approach. Part of the challenge is that banks are trying to cut corners to keep up. Low interest rates are suppressing margins and new entrants are emerging that are chipping away at the profitability of banks. Take the payment industry as an example. We’ve seen countless examples of technology firms like PayPal, Apple and Google emerging within a niche space that is less regulated to offer services traditionally delivered by the big banks. This is taking profit away that has traditionally supported the large infrastructures and complex processes that banks rely on to demonstrate regulatory compliance. At the same time, the manual and complex processes surrounding regulatory requirements are proving expensive to maintain and creating more areas for processes to break and fail.
Stress test complexity
Regarding the IFRS 9 reporting standard, we’re seeing that financial institutions have spent too long creating complex models and have failed to think about how the processes will be executed on a monthly basis. The next trigger point for the risk community will be around stress testing, which is used to identify unknown risks that financial institutions are facing by testing how different scenarios would impact balance sheets. From next year, stress testing will become more complicated and will integrate IFRS9 forecasts. As the scenarios that banks face as part of stress testing become more exploratory, the complexity of modelling and governance over the process will also become more difficult.
To build powerful and predictive risk models that can be moved into production, there needs to be an underlying data model and structure that is built on the availability of clean data. Both legacy and challenger institutions are collecting increasing volumes of data but need to consider how to ensure data quality. For instance, if you’re using social media to collect customer data then you can’t guarantee that it will always be accurate or represent a sample of typical customers. Only once you have good data, can information be automatically analysed, modelled and executed for better risk management. The danger is that complicated statistical models are conceptual and difficult to repeat time-and-time again for compliance purposes. For example, one company has 150 data scientists that have built 50 models in the last year but haven’t managed to move anything into production.
Data governance key to automation
Yet as the risk industry has moved online and become more automated it has introduced new, unknown threats; for example, the potential for an IT failure or a breakdown in the process between people and technology. To reduce the chance of that risk materialising the process needs to be monitored effectively. Good data governance will create a better risk process – from modelling through to reporting – that requires less manual interaction and ensures financial institutions can apply complex systems in a structured way to reap the benefits.
We’ve now reached a point where regulation risks putting a cap on the progression of the industry. Banks are becoming akin to a utility that is expected to operate with lower risk, lower profitability but a far higher amount of capital. Technological capability needs to be implemented before automated processes that are well controlled can create efficient and agile banks. Automation offers an opportunity to help traditional banks compete in a digital realm and become more efficient. However, only if they have the right data, systems and culture.
Find out more about how analytics can drive intelligent risk management.
Lee Thorpe, Head of Risk Business Solutions at SAS UK and Ireland
Last month the European Risk Management Council launched its first flagship event, the UK Risk Management Leadership meeting, which brought together leading executives to discuss some of the most important topics dominating the risk management agenda. One of the things that it highlighted was that nothing is really changing within the risk community. Risk is calculated from historical data and those same risks that executives have been worrying about for several years – low interest rates, challenger banks, regulatory pressures and cyber threats – continue to remain front of mind.
Cultural change
A cultural change is needed that encourages the risk industry to have a more forward-thinking approach. Part of the challenge is that banks are trying to cut corners to keep up. Low interest rates are suppressing margins and new entrants are emerging that are chipping away at the profitability of banks. Take the payment industry as an example. We’ve seen countless examples of technology firms like PayPal, Apple and Google emerging within a niche space that is less regulated to offer services traditionally delivered by the big banks. This is taking profit away that has traditionally supported the large infrastructures and complex processes that banks rely on to demonstrate regulatory compliance. At the same time, the manual and complex processes surrounding regulatory requirements are proving expensive to maintain and creating more areas for processes to break and fail.
Stress test complexity
Regarding the IFRS 9 reporting standard, we’re seeing that financial institutions have spent too long creating complex models and have failed to think about how the processes will be executed on a monthly basis. The next trigger point for the risk community will be around stress testing, which is used to identify unknown risks that financial institutions are facing by testing how different scenarios would impact balance sheets. From next year, stress testing will become more complicated and will integrate IFRS9 forecasts. As the scenarios that banks face as part of stress testing become more exploratory, the complexity of modelling and governance over the process will also become more difficult.
To build powerful and predictive risk models that can be moved into production, there needs to be an underlying data model and structure that is built on the availability of clean data. Both legacy and challenger institutions are collecting increasing volumes of data but need to consider how to ensure data quality. For instance, if you’re using social media to collect customer data then you can’t guarantee that it will always be accurate or represent a sample of typical customers. Only once you have good data, can information be automatically analysed, modelled and executed for better risk management. The danger is that complicated statistical models are conceptual and difficult to repeat time-and-time again for compliance purposes. For example, one company has 150 data scientists that have built 50 models in the last year but haven’t managed to move anything into production.
Data governance key to automation
Yet as the risk industry has moved online and become more automated it has introduced new, unknown threats; for example, the potential for an IT failure or a breakdown in the process between people and technology. To reduce the chance of that risk materialising the process needs to be monitored effectively. Good data governance will create a better risk process – from modelling through to reporting – that requires less manual interaction and ensures financial institutions can apply complex systems in a structured way to reap the benefits.
We’ve now reached a point where regulation risks putting a cap on the progression of the industry. Banks are becoming akin to a utility that is expected to operate with lower risk, lower profitability but a far higher amount of capital. Technological capability needs to be implemented before automated processes that are well controlled can create efficient and agile banks. Automation offers an opportunity to help traditional banks compete in a digital realm and become more efficient. However, only if they have the right data, systems and culture.
Find out more about how analytics can drive intelligent risk management.
Explore more articles in the Banking category
