Digital Banking, Alternative Data, and Advances in Credit Modeling
An Interview with Roberto Jabali, CRO @Banco BV.
The fast growing trend of digital banking is a shift away from traditional brick and mortar financial services to services that are offered over the internet. Digital banking is not only more convenient and accessible for customers, but it also exposes financial institutions to new data and technologies that are gradually transforming the ways many of them do business. One such example is in the field of credit and lending, where alternative credit data are driving more advanced risk models and financial inclusion, and few countries around the world are moving as fast on this as Brazil.
That is why we were very excited to have had the recent opportunity to chat with Roberto Jabali. Roberto is currently the Chief Risk Officer at Banco BV, Brazil’s 5th largest bank, and prior to that he held senior executive positions in credit and risk analytics for years at Santander and Citibanamex. In these roles, Roberto led policy management and risk modeling for both the retail and wholesale sides of banking, and has seen firsthand the changing trends in this industry around the world, and in Mexico and Brazil in particular. Moreover, he is now doing some very interesting work with BV, who recently made the move to 100% digital accounts!
Finally, we at IA have known Roberto for several years, and are thrilled to speak with him about how digital banking and alternative data are changing credit modeling.
Welcome Roberto, and thank you for spending time with us today. Let’s jump right in. What does BV’s move to digital banking mean for the bank’s risk models?
“BV has significantly invested in its approach towards the digital world by the establishment of our digital bank. Our ambition is to occupy a relevant position among other digital banks, and we are doing this. The fact that we now have a digital bank reinforces the need to make digital decisions, and the need to consider more digital information, variables, and characteristics than we did before. This doesn’t only apply to the digital bank, but rather the whole process of making good decisions at BV.
Brazil has made good progress from a macroeconomic perspective. In recent years, the country established a positive credit bureau, for example. And so, the composition of using different information sources to make better decisions has now evolved to include digital information.”
“The final objective is to sustainably include new populations from the credit market, and generate more value by making better decisions.”
What role does alternative credit data play in your digital banking platform?
“Incorporating different sources of information opens the level of decisions with new characteristics that we didn’t have before, and that creates the opportunity for us to approach new markets and new populations that were not included previously.
There are trends showing that the mobile phone is becoming the new credit bureau. I think that is a little overstated, but there is certainly a lot of value coming from different sources of information like the mobile phone. The final objective is to sustainably include new populations from the credit market, and generate more value by making better decisions.
For example, we have established contact with you [IAssessments] for a long time now, in order to capture the value of psychometric scores, and that is something that we absolutely want to include into our decision modeling. In a market like Brazil, that is something that is proving to be valid.”
“I see the psychometric scores as a different and unique source of information that when combined with other sources of information, is going to give a much better compounded result for our model.”
How much emphasis should alternative data actually be given in the final risk models?
“The final proportional value that alternative data will have in relation to traditional data is still uncertain. That is a question that is not yet answered, but is something that needs to be intentionally pursued.
Alternative data is going to gain more and more relevance in the process of taking good credit decisions. It’s not that alternative data by itself is relevant, but when you combine it with traditional data from a whole model’s perspective; that is where we see the data aggregating into more value, in terms of the quality of the model that you can develop.
We have seen this process evolving in recent years in the Mexican and Brazilian markets. In the last decade or so, risk models reached a plateau with their standard traditional data, until new and different alternative sources arrived. For example, geolocation, telco data, and now also psychometric scores. These are all new and different sources of information, which no one had before.
But, just like when other traditional data sources first came to the scene, this is something that needs to mature. And that maturing is something that needs to happen intentionally, by going after that information.
Psychometrics is one example. We tested with you years ago in Mexico, and the results were interesting. I’m sure this next round of testing that we are doing now is going to be on another level. Because the tools have evolved, and because we can now combine this information with other sources. I see the psychometric scores as a different and unique source of information that when combined with other sources of information, is going to give a much better compounded result for our model.”
“Credit is one of the pillars that we need to apply to support the recovery from the pandemic. The fact that you can combine traditional data with new data is certainly something that will enhance the quality of credit, and this will help people recover.”
How can digital banking be used to promote financial inclusion?
“Let’s first consider the context. We are exiting a pandemic (hopefully), and so the whole world needs to be supported in terms of economic recovery, which will be different for different segments. There are segments that will recover better, and segments that need real support for recovering.
Credit is one of the pillars that we need to apply to support the recovery from the pandemic. The fact that you can combine traditional data with new data is certainly something that will enhance the quality of credit, and this will help people recover.
We are now dedicating time and resources to see what are the most relevant population segments, which can benefit from new models. This is similar to an approach, which we discussed in the past, using alternative information for those who are non-banked and outside the banking system.
One of the challenges that we have when talking about different sources of information is how to get that information. As banking evolves to a much more digital framework, we can get closer access to digital information, which is difficult to get in a brick and mortar bank – that is clear. And that translates into something much more robust. So implementation is still one of the main challenges, but as long as the business keeping evolving, the access to this information will get easier and easier.”
How do customers react to banks using alternative data?
“We are exactly at this point in the discussion, since we are still in the early phases of open banking. But it seems to be a matter of managing authorizations. That is, the agreement of the customer to provide certain information, or not. My view is that this is something that needs to be incorporated into the culture of the population, and that is not easy, and takes time.
The banks need to make sure they are working to towards building that culture, by showing customers what kinds of benefits they might experience by providing additional information. That applies to open banking and alternative sources in general, as well as the psychometric data. The early readings that we had from our previous tests showed that people’s acceptance of providing that information was surprisingly not bad at all.
We also always need to be mindful about what regulations apply. Here in Brazil we have LGPD regulations, which is related to the roles and responsibilities of how we treat personal data. As long as we strictly follow what the regulations say, we are going to be working in favor of the customer, and contributing for him or her to have a better credit. So that in the end of the day, it will help them do whatever they want to do.”
That all makes a lot of sense, and sounds really exciting. Thank you, Roberto. Do you have any final comments?
“BV is not the biggest bank in the country, and it is not a fintech startup either. Because of this, it is uniquely positioned in a way that gives it the agility to make decisions quickly, while also having the capital to compete in digital banking. The decisions that are made from a strategic side are not easy, but the bank is already delivering on what it promised a few years ago. I think there is a really bright future for the bank in this respect!”