Credit Risk Modeling and Decisioning

Hello, and welcome to the New Age of Risk Analytics podcast. Today we speak on credit risk modeling and decisioning with Ivy Tan, EY Senior Manager specializing in IFRS 9 and credit risk for the financial services sector, and Vikas Veep Sharma, an Associate Partner at EY, also. I'd like to welcome you both to the New Age of Risk Analytics podcast. I'd like to just ask you both for some background on your current roles and responsibilities and how it relates to our topic today.

Let's start with Ivy Tan, EY Senior Manager specializing in IFRS 9 and credit risk for the financial services sector. Can you tell us a little bit about your current role and responsibilities? Sure. Thank you, Lisa, for inviting me to this podcast.

I am a credit risk specialist based in Ernst & Young's Singapore office. My role is primarily to assist the EY assurance team in credit risk-related audit and advisory work for the banking clients, hence my title, which sounds quite a mouthful. Thank you. And now Vikas, an Associate Partner at EY, who we are pleased to note is a certified FRM.

Vikas, can you tell us about your current role and responsibilities? Thank you, Lisa, for your time and welcoming me here. I lead risk data analytics and reporting solutions for financial risk advisory practice in EY Singapore. Now through the years, I have advised financial institutions on credit risk, primarily covering credit risk modeling and credit risk process.

I also cover other financial risk topics such as market risk, liquidity risk, and operational risk. More recently, though, in line with how our client needs are evolving, my focus has moved to data governance, regulatory reporting, and now, more recently, use of AI and machine learning in risk and compliance. Well, I'm sure that's a fascinating trajectory. And, you know, we certainly have, you know, GARP has obviously a broad base in both the financial and energy sector of risk practitioners who work in various arenas, both at market risk and operational risk and enterprise.

From your perspective and your expertise, I'll start with Vikas. Can you give us an overview of credit risk modeling and understanding of the risk profile of a firm? Credit modeling is, in simple terms, translating the demographic and financial data of a customer to a creditworthiness metric. This aids in credit decisioning and portfolio management, leading to the desired risk profile of a firm's portfolio.

The metric can be used by relationship managers to propose the right transactions, by credit officers in assessing creditworthiness of the borrowers and then making approved choices, and then by portfolio managers in recommending the right mix of sectors, geographies, customers, and products. That's very interesting. Ivy, do you have anything to add to that? Well, for my banking clients, credit risk modeling to them is just an analytical tool that they use.

They basically just organize data in a statistically meaningful way so that it helps them in making better credit decisions. Absolutely. Vikas, you mentioned, you know, machine learning and AI, and we are hearing from so many risk practitioners that the landscape is changing substantially at the moment as a result of technology, also regulation, and other issues. But let's start with technology.

Are you finding that your role and responsibilities are really shifting because of this new technology? Yes, certainly. I think changes in the technology are bringing innovation to every aspect of credit origination and management. In today's world, technology is creating new channels for us to reach the customers and enriching the data sets, while from a process point of view, a number of our clients are now focusing on automating and digitizing their legacy credit process.

Today, we have access to alternative data sources if you want to build scorecards for the unbanked population. At the same time, we're also pursuing more updated credit modeling techniques, although I have to admit that there are problems with explaining the outcomes and relating those outcomes to the inputs. But I also firmly believe that this is just the beginning of a change. I think that's certainly the case.

Ivy, do you want to talk a little bit about some of the regulation changes that are occurring that are – has your job shifted in the past few years because of things like IFRS 9 and CECL compliance? Yes, definitely. IFRS 9 has been impacting the financial institution quite a fair bit. From what I'm seeing, the biggest challenge of IFRS 9 is it requires cross-functional teams to work together because it's no longer a silo function where credit does their own thing, finance does their own thing, operation does their own thing.

IFRS 9 actually calls for the integration and the cooperation of these various teams. You need your finance guys to work with your credit guys, to work with your ops guys. So that, in fact, is a bit of a challenge that a lot of my banking clients are trying to work through and trying to get all these teams to align over a long period of time. So we can certainly see that regulatory and accounting changes sometimes does affect how an organization think about new ways of doing things.

Yeah, and I actually have a question along those lines. You know, I think the nature and profile of teams are changing and their sort of skill sets are changing. What are you finding about as these regulations are implemented? What do you find and are the biggest challenges in getting teams to communicate and sort of break down those silos?

Well, if you have any advice along those lines from your wisdom. Okay, not so much of advice, but more of what I'm seeing from the big MNCs banking clients that I have, a lot of time to encourage communication across teams. They have to promote agility and they have to promote what they call cross-functional teams. So the way a team is set up traditionally where basically you have that's a finance person reporting into a finance person, probably doesn't quite work in this new era where you need communication and cooperation across different functions.

So what I'm seeing now is more of like setting up teams in line of projects or like for a particular task, like for reporting. And then you pull different people from IT, from the risk guys, from finance, and everybody just get together in a team where everybody is a specialist, but yet at the same time working towards a common goal and reporting to maybe a senior manager who is like overseeing the entire project. So the traditional way of looking at function, looking at team is actually changing in this age. And also with digitalization, a lot of automation tools are being used.

So it helps in working in this kind of cross-functional team where everyone works on different things, but basically trying to achieve the same objectives. Are you finding that as well? Are you seeing those changes as well, Vikas? When we talk about IFRS 9 or any big transformational regulatory changes, to operationalizing these kind of processes, you need to maintain a minimum bench strength.

And I think the teams that you have trained for such processes or running such processes, the same teams could then be also leveraged to run other credit models or other reporting processes. Now I'll give you an example. If you have invested in infrastructure and you have this infrastructure ready to provide the right data in the right form, the same data then can be used for internal risk reporting. And the teams that have been working on IFRS 9 could also be then leveraged to report internally to risk committees.

That makes terrific sense. I'd like to talk a little bit more about, you know, data integrity and integration of systems. You know, we've talked about how the teams, and I do think that there's some adaptation, but, you know, in automating these processes, you really have to worry about the data integrity. What are you finding are the biggest areas of focus when you're addressing those kinds of issues and ramping up with these new technologies and complying with these new implementations of regulation?

Data integrity is certainly critical in ensuring that you have the right data in the right form. And accuracy and completeness of the data cannot be ensured if you do not choose the right data sources with adequate controls. Certainly when you automate and you centralize the process, you have to think about the data governance around those processes to ensure that you have these controls in place and then the data that you are using for any measurement methods or reporting has met the standards that you have set for yourself. So essentially, that's the biggest concern that I would say that the industry would have.

Do you agree with that from your perspective, Ivy? Yes, I agree that data integrity is critical, in fact, for my banking clients being the regulated entities that they are. A lot of times, they place a lot of emphasis on governance because they need to be able to trust the results that's generated from all those models that they use. And most importantly, for the financial institutions, they need to use these results to do financial reporting, to do regulatory reporting.

So they need to be sure that the result they are using is something that they can trust and they believe in. Do you have anything in particular that you think is really critical for our listeners to know about data integrity in this work and how you approach governance with regard to that element? Are you finding that there are certain techniques in looking at governance that are particularly helpful or are there areas which you feel like people need to really keep a bigger focus on? If you take an example, small errors in the data sets could be easily fixed when you're using these data sets for credit modeling.

But if you have larger data sets, larger gaps in the data sets, they could be very material. And in such a case, you almost always have to go to the sources to obtain raw information, which could be very time consuming, and then to tie up your resources. Now, in such a case, the only option left would be then to drop the entire records of information or drop the attributes, the columns from your data sets. And the data set then you produce may not be representative of the population and hence any models that are built on these data sets may not lead to accurate decisions.

Now, Look at, if you look at how BCBS 2.9, for example, has provided this level of awareness around accuracy and completeness of the reporting processes and whether this data set is used for credit risk modeling or any other purposes, the set of standards and principles are about the same. You could potentially apply the same principles to how the data sets are being produced for your models versus how these data sets are being consumed in your financial reporting or regulatory reporting or even internal risk reporting processes. So this provides an excellent opportunity to financial institutions then to establish a common set of principles and then translate them to a common set of standards that are applicable to a wider range of functions within the institution. I think that's a really great point, and I think Ivy, let me see if you, I mean, one of the things that I, when I talk to people about any kind of regulation or standard, is there really opportunities for people to do their jobs better and more completely and, you know, with greater cognizance of the impact?

Are you finding that, you know, in implementing things like IFRS 9 and BCBS 29, that these are really helping your organization function optimally? Okay, in a way, yes, because when there's regulatory changes and when there's guidelines that is given by some authority, a lot of time it forces the people doing the job to re-examine what they are doing. In a big organization like a bank, there are always a lot of processes, there are always a lot of procedures that sometimes it becomes routine and everyone just basically does it step by step day in and day out. So with all this implementation of new rules or new regulations, like for instance IFRS 9, traditionally when we talk about accounting standards, we always think that the finance guys are the one that will be doing it and they'll be passing the entries and everybody else is like, you know, totally clueless as long as the accounting entries gets passed.

But with IFRS 9, what happens is it just forces the various departments like the credit people need to get involved because they need to start talking about staging, provisioning. And then the finance guys suddenly realize that, hey, it's not just about passing accounting entries, they need to know about the criteria, they need to know about the definition. So it actually not just forces people to think differently, it actually equips people with additional knowledge and creates more awareness. So at the end of the day, everyone is wiser, so to speak, and they're able to do their job better.

That's my view. And do you have anything to add on that, Vikas? Yeah, I think these regulations definitely, I think, as I mentioned before, financial institutions have to maintain a certain bench strength to run these processes. And as Ivy mentioned, that it impacts a broader set of functions within the firm.

So if you train your people with the right skills and expertise, it helps leveraging the same team to run the processes, whether it's for IFRS reporting, measurement, or whether it's for CCAR measurement reporting. There's a certain commonality across these standards. And I believe that the same skill set can be leveraged across them. So I think those provide an opportunity to people to upskill themselves, learn these new governance standards, including new technologies and new methodologies for credit risk modeling and decision making.

And I think this is an excellent opportunity today that these regulatory changes or those technology changes present to the people in general. You've answered one of my questions, which is, you know, essentially, how are people's jobs changing because of these implementations and how do they affect other systems of work? And I think that they do. It seems to me that when I talk to, you know, risk practitioners in this arena, that they are finding that the lessons they learn from implementing any new standard or regulation or having to ramp up for a new technology really give them a skill set that works across areas of focus.

Am I hearing that from you as well, clearly? Is that, am I understanding that correctly from you? And if so, you know, do you think that there's anything we've learned from the implementation and planning of IFRS 9 or any of these other standards that, you know, has specifically helped teams that are not necessarily tasked immediately with implementing them? I agree with the observation.

In fact, you're right. You know, for a lot of people, having to implement and learn new things arising from all these changes actually acquires them new knowledge and new skill set which is transferable. So it's not just like, you know, when you're doing the IFRS 9 implementation and you learn all about IFRS 9 and that's it. Once IFRS 9 is over, you know, this knowledge is no longer relevant.

It's the skill set that you learn in, for instance, cooperating and cooperating with other departments, learning about new things. Like for instance, you're a finance person and suddenly you get exposed to all the credit knowledge. All this comes in useful. And these are transferable skills that going forward, especially, which I believe that cross-functional team is the way to go, it will actually equip people with the skill set to function more efficiently.

So for other teams that have not gone through this regulatory process from the outside, when a person looks at it, it may be a little bit overwhelmed at first to say that, you know, there are so many new areas that I need to learn. There are so many new areas that I need to adapt. But it's actually a good adaptation because it opens up your knowledge. It actually allows you to acquire new skill set.

And I would say that from my perspective, it's your job is changing for the better because you're learning more things, you're being more agile. You are, in fact, contributing in a different way, not just by, you know, processing certain things, but more of like contributing to a bigger goal than just getting the report right, for instance. That's how I look at it. And I think that that's what I'm hearing from people in other areas as well.

Vikas, did you have something to add? Yeah, so I'd like to add to that. I think it goes beyond also the team skills and expertise that you've built. Now, if you look at it from a bank point of view, and I'm sitting in a bank and taking decisions about investing my budget across various budgets, I would say that IFRS 9 and other transformational regulatory changes allowed me to upgrade IT infrastructure, set up data governance, and train my people to run the new process.

So I would then like to make sure that the investment I have made in these processes would be useful for the other initiatives in the future. So, for example, the same sources that are being used for obtaining data for IFRS 9 could also be leveraged for CCAR or other regulatory reporting processes. The same teams that have worked on IFRS 9, they could then be used to build other credit models and run other credit models. And I think one of the important elements of that, the firms themselves gain immense knowledge and significant knowledge about their own data and risk profile that is quite valid in the long term, especially in improving the credit decision-making process or any other management process through which firms make decisions.

You both are very articulately stating this, and I really, truly appreciate it. Just a couple more things I want to ask before we wrap up. As you are based in Singapore, and we've heard a lot, our listeners have heard a lot, and I've had a lot of contact with people who are based in North America and EMEA. Do you find that there are any challenges that you're dealing with in Singapore that come across regulatory or technology changes that are occurring in North America?

Particularly a really good example of this would be Cecil versus IFRS 9. When you're dealing with a multinational corporation, they may fall under various jurisdictions and have to comply with both. Are you guys finding any challenges along those lines, or are you finding that these processes are robust enough to be applicable to whatever challenge comes down the pike? Fundamentally, if you just take the example of Cecil and IFRS 9, they are based on the same set of principles.

Now, there are definitely differences in the way you measure expected credit losses. However, if you look at any regulations, the regulatory changes are generally, since they're based on the same principles, you could potentially leverage the same knowledge and infrastructure. What you're seeing in some of our clients as large financial institutions, they have started to build strong centralized regulatory compliance functions that are able to manage and report on this change. I mean, certainly, there are challenges because you have to, again, ensure that the people in these functions are adequately trained.

You hire the right talent. You retain them. However, this knowledge is essential to understanding the cross-border differences across the regulations. Having said that, I'd like to mention one thing, though, that the technology has started to show quite a bit of promise in managing and especially identifying this change and then assigning these changes to the right parties who would then have to address the regulatory change, be it IFRS 9 or any other regulatory change for that matter.

I'm not sure if that answered your question, Louise. No, you did. And I think that's a really good point about the technology. And I think that's part of the hope of it is that it will allow for efficiencies.

Ivy, did you want to comment on that, on sort of the different regulations that come from different parts of a firm? Yeah. I mean, yes. For my banking clients, especially the big MNC banks, yes, they do have the challenge of having to manage all the different accounting standards and different regulations from different regions.

Even let's talk about ASEAN itself. It's like there are all the local standards versus the international standards and so on and so forth. What I'm seeing is also, as per mentioned by Vikas, a lot of them, they use technology to try and streamline as much as possible their basic operations between the You have inputted into the system. That's very well put.

Thank you both. I thank you both very much for your time today. I hope that you have a great rest of your day. Thank you very much for your time and I appreciate the opportunity of speaking on these topics.

Thank you and have a good day. Actually, it's a good evening to be up indeed. And thank you for joining us on The New Age of Risk Analytics. Please subscribe on your favorite podcast platform to be notified of future episodes and visit sas.com slash risk for information on these and other risk topics.