Page tree
Skip to end of metadata
Go to start of metadata

FlexFinance supports the consideration of credit risk throughout all the stages in the life cycle of a loan. From deal origination date onwards, the solution tracks the development of credit risks.

Risk-based pricing in the past began by calculating the revenues associated with a loan, including the interest income and any fee-based income. 

These revenues were spread over the life of the loan.


To “risk-adjust” the interest income, several factors need to be considered.

  • Funding costs
  • Administrative and overhead expenses associated with the product.
  • Credit risk cost


For the loan application process, FlexFinance provides a credit spread (CS) calculated on PD and LGD in a neuronal network. This CS can be used as a component for calculating the interest during the appliction process.

Besides optimising the fulfilment of requirements implemented by legal authorities regarding financial accounting and capital requirements, the higher accuracy in calculating expected credit losses combined with improved segmentation supports a pro rata allocation of expected default risks in the internal business segment calculation. This offers new options while defining interest conditions: in determining the interest rate conditions in the context of a loan application, the allocation of credit default risks in line with the originator leads to a "fair" interest rate. From the point of view of individual business, the use of machine learning means that, in the case of higher credit default risks, the associated costs can be identified precisely for the first time when applying for a loan and can be claimed as a contribution to the interest rate conditions. This tends to result in higher net interest income for transactions with greater default risks. This improvement in net interest income is offset by higher loan loss provisions in the income statement.

Of course, the enforceability of higher interest rates for loans with a higher default risk is initially theoretical, since it depends on the framework conditions of the respective market.

The cause-related allocation opens up the possibility of deciding which products and conditions are offered to which individual customers and at which price. The use of machine learning makes it possible to accurately identify the appropriate credit risk for a customer. This in turn creates the basis to avoid situations, in which

  • Too much credit spread leads to an interest rate that exceeds the market interest rate, the condition contribution required to cover default risks cannot be implemented on the market.
  • Too little credit spread might lead to a successful customer acquisition, but the expected increase in interest income will be consumed by the impact on future profit and loss due to credit losses.









  • No labels