Credit Scoring On Open Bank Data: A Revolution?

The wave set in motion by the European PSD2 directive has not finished sweeping the hitherto frozen landscape of credit underwriting processes, in particular – but not only – in countries which, like France, do not have a positive credit bureau.

It is now possible to access the borrowers’ bank accounts, which enables the creation of scoring models based on the borrower’s bank transactions in the months preceding the application: the borrower’s risk is assessed on this basis, very different from the elements on which traditional models are based. The latter generally value the stability criteria, while the models developed on open banking data take a much more behavioural and transactional approach directly related to recent financial behaviour. These open banking models make it possible to take into account elements relating to the stability of the borrower’s expenses, available savings and the management of his/her account balance and overdraft.

The results obtained by the pioneers in this area are excellent, performing even better than traditional models. For instance a recent French example reached a Gini of 72. Furthermore, this approach avoids the tedious entry of socio-demographic data, criteria necessary in a traditional approach to risk estimation, but superfluous in this new one, which exploits the potential of banking aggregation. In the end, associated to the power and flexibility of TheLoanFactory, the application process is facilitated and accelerated, the risk is better estimated, for the greater benefit of the borrower as well as the lender. A winning combination that may disrupt loan onboarding habits, subject to the acceptance by the borrowers of entering their banking connection credentials!