Blog article
See all stories »

Improving traditional credit scoring with a forward-looking approach

Question: Do you know how much your total committed credit exposure is?

 

Traditional credit registries

In Belgium when a natural person requests a credit, banks and credit institutions are obliged to consult the central credit register (CKP/CCP) at the National Bank to obtain the existing credit exposure of the individual and check whether the individual does not have any reimbursement issues for existing credits.

Banks are also obliged to inform the National Bank when new credits are contracted and when individuals have reimbursement issues.

The aim of this register is to protect individuals from accumulating excessive debt at different financial institutions.

Current credit risk models start to become insufficient

Despite these good instruments, many cases still slip through the net. First of all several credits, like for example credits below 200 EUR, are unregulated and as such do not need to be registered in this central registry. Furthermore this register only gives a partial view, because:

  • It only includes the liability side, i.e. there is no view on the asset side, potentially negatively impacting people with a lot of debt, but also a lot of illiquid assets
  • In the standard case of no reimbursement issues, the register only contains the credit opening amount and not the current credit outstanding amount. Especially for revolving credits, like credit cards and overdrafts, big differences can be present between the credit line and the actual credited amount
  • As already indicated the register does not contain the unregulated credits, but also other payments for which the individual has made commitments are not registered, such as:
    • Signed upcoming payments, which will only be debited from the customer account in the future
    • After-pay commitments (Buy Now, Pay Later), for example after-pay payment to be made with the popular rising Fintech Klarna
    • Subscriptions and other recurring payments (like rent, child alimony, etc.) which are only charged at the end of the month
    • Large bills coming later, like insurances, utilities, etc.
    • Signed contracts, resulting in future payments (for example a signed contract for executing a home renovation)
    • Tax liabilities (like VAT, personal taxes, etc.)

It is clear that with the ongoing movement of embedded (frictionless) financing and frictionless payments (like automatic subscription payments), the current models to protect consumers against excess debt start to become insufficient and also inefficient. Not only does the current model lead to a lot of false positives (i.e. people for which financing is accepted, while it shouldn’t), but also to a lot of false negatives (i.e. people whose financing is refused, while they would be able to reimburse the credit).

The main inefficiency comes from incomplete credit registers as indicated above, but also because credit legislation is still too much backward-looking (i.e. looking at your credit history), rather than forward-looking. People who have had debt issues in the past, often have great difficulties finding new financing, even if they have completely adjusted their financial situation. As a result, this blocks a large group of people from obtaining financing to execute their future projects.

Forward-looking affordability checks

The current credit history checks should therefore be replaced by (or at least complemented with) an "Affordability Check", based on your current situation: all your expected incoming cash flows (with different levels of certainty of receival) and your committed outgoing cash flows (committed credit exposure) should be taken into account. Such affordability checks allow to be forward-looking, instead of backward-looking.

This requires however a real-time estimation of an individual’s cash flows and balances (and the same applies for  small- and medium-sized companies) - cfr. following blog on personal balance sheet.

The Affordability Check can be obtained through a collection and aggregation of all (expected) incoming and outgoing cash flows and all existing assets and liabilities.

Bank and Fintech PFM an BFM apps (like KBC PFM app or Cake app) and accounting platforms (like Silverfin, TOCO) providing more real-time financial reporting already provide a big part of the puzzle, but a lot of data is still missing to make a proper assessment. Capilever’s NLPT (Non-Liquid Position Tool) tool can help to complete this puzzle.

However in order to really make leaps in this evolution, regulators will need to enforce all institutions where an individual makes a committed credit exposure to share this data with a repository of choice by the customer (e.g. their bank). This way customers have a complete view of their exposure and (current and expected) financial situation. Via Open Banking architecture, the aggregator should furthermore provide customer scoring to counterparties. This way the counterparty gets an idea of the person’s affordability, whilst avoiding to share all details of the customer’s financial situation. The first steps towards this concept are included in Capilever’s CPRA (Counterparty Risk Assessment) tool.

Obviously the bank or Fintech given the consent of the customer to collect all this data will have a super detailed, holistic view on the complete financial situation of the customer and as such can give very personalized advice. This competitive advantage will be enormous, meaning that banks and Fintechs should already start preparing today for this evolution of tomorrow.

1570

Comments: (0)

Thomas Pintelon

Thomas Pintelon

Co-founder

Capilever

Member since

13 Jan 2017

Location

Brussels

Blog posts

23

This post is from a series of posts in the group:

Financial Risk Management

This network brings together professionals involved in the oversight and management of their company's financial risks and exposures as well as solution vendors, in order to discuss risk issues including interest rate risk, foreign exchange risk and commodity price risk, among others.


See all