The structure of Asset Liability Management (ALM) is going through a fundamental shift as banks break down silos between Treasury, Finance, Risk, and business units. Regulatory expectations since the global financial crisis, the widespread adoption of cloud technology, and heightened interest rate volatility since COVID have created a self-reinforcing cycle of change. To remain resilient, it is already a reality that institutions must rethink their approach to ALM, where speed, transparency, and automation are prioritized.
In this CeFPro interview, Eoghan OGriobhtha, UKI Managing Director at Mirai RiskTech, shares his perspective on technology-driven changes in banking ALM.
Eoghan OGriobhtha spoke with CeFPro, and during his interview, he shares his takeaways on the converging forces of regulation, cloud adoption, and automation, and how these dynamics are reshaping the future of ALM.
What are the main insights from his perspective on technology-driven changes in banking ALM?
Together, these insights highlight a clear industry shift: from legacy systems and workarounds to unified, automated, and AI-enabled ALM platforms that are foundational for resilience and competitiveness.
To dive deeper into practical applications, explore our whitepaper "Modernizing BSM: How Big Data, Cloud, and AI Are Shaping the Future of ALM and Liquidity".
Read the full interview below.
This is a phenomenon I've observed over the course of my career. In my mind, there are three key drivers to a unified approach, and in important ways they overlap and fuel each other.
The first key driver is regulatory oversight. Beginning back in the global financial crisis, regulators require banks to provide very detailed information, which is a large overhead to produce.
Secondly, since about 2015, there's been an ever-increasing adoption of cloud computing in banking and operations. So, it started with customer-facing parts of the bank, such as banking apps, and now it's permeating into the interoperation of banks.
We're only really seeing the maturity of this in the last four to five years, and you're seeing a lot of regulations coming out about cloud outsourcing. But finally, we were seeing interest rate volatility in the market since COVID, and this is another key driver.
These factors are all intertwined to create a little bit of a flywheel. So, rate volatility requires more timely information for senior management, getting increasing amounts of information available, asking more questions, and putting more investments into being able to produce more information.
Then, regulators are seeing that there are more technological practices being used. They need to legislate to keep consumers safe with cyber, etc.
They also become more aware of the capabilities, and they themselves begin to see that there's more information that can help. So, it's a self-perpetuating process.
An ALM framework will ultimately be consumed by the senior management of the bank at ALCO or otherwise.
And the senior management at ALCO, they'll always consider these metrics holistically within a risk appetite and regulatory limits. So, they're managing one balance sheet, and they have multiple risks they need to manage; the solid approaches really are technology constraints rather than business decisions.
But I've seen management frustration at the lack of holistic information available. And if we shock liquidity, what happens to IRO or FTP?
The answers would typically be “we'll tell you at the next ALCO, in a month's time”. And the typical thing I saw when things took too long was senior management taking a bunch of high-level information and assumptions based on the metrics that were produced, and they create their own basic model in Excel or otherwise, and have these guardrails to make their decisions.
The obvious win here is to have an infrastructure that can let you run management scenarios on demand, on the granular data. It'll allow you to act quickly and make more precise decisions.
Automation will reduce the time taken to produce information and remove reliance on multiple process hubs. It'll increase confidence in data and information being produced as everything's tracked, logged, and queryable.
It'll make information available to business teams quicker, and ultimately, it'll give the ability to answer more questions and more complex questions.
Automation is interesting; every industry is looking to automate. So, what exactly are we trying to automate in ALM? Most banks typically will close their accounts on a plus five basis, and the risk process is a plus 10, plus 15. This is due to the connectivity of all the systems and pieces. Excel is plugged into multiple steps to assist with the data and plug the gaps.
There is a best practice approach. The first step is to capture and cleanse your data in a well-governed process with a four-eye oversight. These should be logged and traceable throughout the whole of the downstream process. And then the following processing of data, including business logic, should be performed in a controlled system, again, with full lineage and traceability.
So, we need to resist having Excel plugged into multiple hops of the process. And, of course, I will mention that all of this we've built in with the Mirai platform.
I've seen two strong trends that will continue for the three-to-five-year horizon, I believe. One is within the larger D-SIB and G-SIB banks. And the other trend has been the mid-tier tiers.
So on the D-SIBs and G-SIBs, we've been working with some larger banks to increase their modelling sophistication.
As an example, one organization wants to perform the outlier test and ad hoc scenarios on their front office cash flow models. We've been using that platform to call their model libraries and then perform scenario analysis on that cash flow.
Other examples include running machine learning models on back testing. So, the real emphasis here is to increase capabilities and insights in the age of AI.
The second trend is among the mid tiers. Here, predominantly, they're looking to automation while modernizing their architectures.
There's an increased focus on granular-level processing and applications that can exist with lineage and traceability. And, of course, the production of our integrated reports across risk metrics. So, across liquidity, FTP, interest rate, and risk, having one system that can produce all of them together.
Watch the full interview here for Eoghan’s uncut insights on how regulation, cloud, and automation are reshaping ALM.
To dive deeper into practical applications, explore our whitepaper "Modernizing BSM: How Big Data, Cloud, and AI Are Shaping the Future of ALM and Liquidity."