The increase in the cost of global credit sets the scene for financial institutions to create risk models that are more exploratory, adaptive and responsive to changes in real time
The current economic landscape requires banks to use their financial data appropriately to adjust their models and keep delinquency indicators under control, without slowing their core business: lending.
The Federal Reserve Board (FED) of the United States increased the reference rate four times this year and its decisions have pushed the rest of the world's central banks to align themselves, resulting in an increase in the cost of global credit.
This poses a challenge for banks wanting to continue offering loans without increasing their exposure to risk, but they have data at their disposal that they can use to improve credit analyses.
The design of credit risk models based on legacy systems is proving deficient when it comes to automating workflows and the analysis of data, which is often housed in different servers and lacks correct labeling, cross-referencing and utilization.
With that in mind, financial organizations need faster and more exploratory forecasts. They also require better management of their data by integrating it into agile servers in the cloud, where they can identify the behavioral and transactional patterns of their customers.
Financial companies can cross-reference information between their different subsidiaries, drawing on multiple data sources such as transaction volumes, purchases, savings, and insurance products to generate credit profiles for their customers.
For example, the Brazilian neobank Nubank has an algorithm it calls Betty that is able to evaluate transactional data and thereby give more people the chance of obtaining for a Nu credit card. Initially, Betty takes information from traditional credit bureaus and then uses machine learning to establish whether the customer makes debt payments to other institutions on time. This indicator opens up the possibility of improved credit scores.
This intelligent risk model can be adjusted to potentially reduce defaults or default rates and consistently score applicants. The engine needs to be actively adjusted in order to synchronize with real-time data, as the traditional credit analysis parameters are no longer sufficient.
While using credit bureaus is an initial step, diverse financial firms agree that alone they are not sufficient or fully aligned with today's economic reality
Enter gig economy workers and the self-employed, who often face obstacles in accessing credit bureaus.
Self-employment platforms have increased tenfold in the last decade and banks urgently need to adapt their risk models to cater to this sector. The global ride-hailing platform DiDi, for example, takes into account the usage history of passengers and drivers to offer loans of up to US$1,450 that are financed by the company itself. The length of time using the application and the frequency of trips is fundamental.
In Latin America, fintechs such as Lana, Ábaco and Migrante offer loans to the employees of platforms such as Rappi, Pedidos Ya and Uber for the purchase of the motorcycles and cars that are so critical to the gig economy.
For their part, banks can obtain information on applicants not only from their servers but also alternative sources. Alternative indicators such as rent payments, telephone bills, monthly payments to children's schools are elements that can provide a more complete picture of people's financial behavior. And it is the job of banks to generate the technological capacity to group, analyze and add value to this data.
Technology allows banking data to generate appropriate credit profiles for each customer. This is how digital channels can speed up the approval and disbursement of a personalized loan.
This automation requires banks to implement agile, robust technologies using artificial intelligence to identify who they should lend to, how much and at what interest rate.
To reach this point, it is essential to have an adequate Know Your Client (KYC) system and effective onboarding, which allows banks to cross information on the identity and financial histories of applicants.
It is the only way for banks to calibrate their powerful data generation abilities and give them the agility to deliver the loans needed by today's economy.
Join our online community and stay up to date with the latest news from the world of technology.