This website uses cookies for analytics purposes. Learn more
Vestigo Scorecard Support
- Client: Any Lender
- Project Length: Variable
- Personnel: Flexible, depending on need
Short-term responses
Challenge
COVID-19 has significantly altered the overall lending environment, impacting both consumer incomes and borrower behaviours
Without sufficient data to inform an entirely new set of scorecards, lenders must leverage the available data dynamically to successfully manage risk and continue to operate effectively
Solution
We help lenders to gain a clear understanding of the weaknesses of current application tools, allowing clients to:
- Understand drivers of consumer stress
- Leverage additional tools, such as Open Banking, to address under-performance of existing scorecards
Implementing more conservative measures, such as additional policy rules to be used alongside an existing scorecard or adjusted score cut-offs, help to increase control of the lending decision
Forward-Looking Scorecard & Origination Analysis
Challenge
Scorecard performance changes and shifting patterns in the underlying variables leave lenders struggling to balance the conflicting requirements of reduced risk appetite, while maintaining a decent volume of lending
For example, increased use of Payment Holidays may have hidden or delayed potential arrears cases, resulting in a misalignment of credit bureau arrears variables. Differing applications of Payment Holiday guidance per lender may have also lead to conflicting information
Solution
In order to react proactively to the changing environment, shorter outcome windows and increased weighting towards recent case performance is vital to ensure a lender can cope with customer behaviour shifts
We help lenders to closely monitor population shifts to highlight any potential issues, using analysis such as:
- Population Stability Index, by understanding score distribution shifts of the current population vs the development sample
- Characteristic Misalignment, analysing any changes in distribution at a characteristic level
- Gini and KS statistics to monitor scorecard predictive power
Our granular characteristic analysis uses shorter outcomes that are correlated to the wider outcome in order to react more quickly as a lender
Using these short outcome windows and early performance origination analysis, we suggest improvements such as the recommendation of policy rules that may be cautiously lifted in the future as normality returns
We also provide support at risk committees, allowing lenders the use of our wider market knowledge to benchmark their portfolio performance
Self-Learning Decision Support Layer
Challenge
The economic shock caused by the unprecedented COVID-19 pandemic, and associated lockdown, has resulted in scorecard performance deterioration for many lenders
Insufficient performance outcome of the current state-of-play means that lenders are unable to begin development on an entirely new scorecard suite. The challenge is therefore for businesses to respond to economic changes quickly, while maintaining the existing scoring framework
Solution
The Vestigo Self-Learning Decision Support Layer enables our clients to overlay many of the benefits of dynamic Machine Learning techniques onto existing, well-established decision processes
This layer sits after the core decision process and acts as a ‘safety net’:
- The ML model assesses cases based on the very latest information and flags where the decision could be improved
- These can then be referred for manual review, sent down an alternative strategy, or receive additional conditions
- Short horizon performance data is fed back into the self-learning layer
This solution allows lenders to adapt to continually shifting market conditions by capturing the latest customer behaviours, without the need to redevelop the existing decision-making framework
See how we can help
Contact the
Vestigo team
Experienced risk and analytics consultancy
Building analytical solutions that are focused on your real-world objectives
Get the latest news from Vestigo