The Use of Data in Traditional Credit Decisions

Enhancing Financial Inclusion through Advanced Credit Decisioning

Financial inclusion aims to provide accessible and affordable financial products and services to everyone, regardless of their financial standing. To achieve this, countries are expanding access to formal credit systems. In this context, advanced credit decisioning models, leveraging data, AI, and machine learning, play a crucial role.

Traditional Credit Assessment vs. Modern Approaches

Traditionally, credit institutions relied on Credit Information Companies (CICs) to assess creditworthiness. CICs provided information on loan types, liabilities, default history, and credit scores. However, with the rise of digitalization, financial institutions are adopting automated and data-driven approaches.

Advancements in Credit Decisioning

• Data, AI, and Machine Learning: These technologies enable banks and fintech companies to improve credit decisions and streamline lending processes.
• Automated Credit Decision Models: These models automate data storage and analysis, facilitating accurate and timely decisions.
• Digital Performance Models: These models help identify and reject high-risk customers, reducing credit losses.
• Benefits:
  • Increased revenue.
  • Improved customer experience.
  • Reduced acquisition costs.
  • Significant reduction in credit losses (20-40%).
  • Enhanced efficiency through automated data extraction and case prioritization.

Challenges in Implementation

Despite the advantages, banks and financial institutions face several challenges:

• Cultural Issues: Resistance to change and adoption of new technologies.
• Data Handling Problems: Ensuring data quality, security, and compliance.
• Lack of Technological Education: Insufficient skills and expertise in using advanced technologies.
• Inflexible Models: Difficulty in adapting to changing market conditions.
• Regulatory Reviews: Navigating complex regulatory requirements.
• Lengthy Implementation Timings: Time-consuming processes for deploying new systems.

Best Practices for Designing Credit Decisioning Models (McKinsey)

To overcome these challenges and build robust credit decisioning models, companies like McKinsey has identified the following best practices:

(a) Establish a Modular Architecture

• Concept: Create a "meta-credit signal" based on data, coverage, and industry information.
• Benefits:
  • Flexibility to add or remove modules.
  • Robustness to adapt to economic disruptions (e.g., pandemics).
  • Customer-centric models that react to diverse customer interactions.
  • Example: During a pandemic a module related to pandemic economic impact can be added and weighted more heavily.
• Explanation: A modular architecture allows for the easy modification of the model, making it adaptable to new situations.

(b) Increase Data Sources

• Concept: Tap into various internal and external data sources, including non-traditional data.
• Benefits:
  • Improved accuracy and reliability of credit decisions.
  • Access to information about individuals with limited credit history.
  • Examples of non-traditional data:
    • Telecom data (mobile usage, bill payments).
    • Social and networking site information (travel, job status).
• Explanation: More data leads to a clearer picture of an individual's financial behavior.

(c) Use Business Expertise

• Concept: Incorporate internal business expertise into the model development process.
• Benefits:
  • Identification of missing credit signals.
  • Validation of new credit signals.
  • Helps to ensure the model is practical and realistic.
• Explanation: Business experts understand the nuances of credit assessment and can contribute valuable insights.

Impact on Financial Inclusion

By implementing these best practices, financial institutions can:

• Enhance financial inclusion by providing access to credit for underserved populations.
• Reduce credit losses and improve profitability.
• Improve operational efficiency and customer satisfaction.

In conclusion, advanced credit decisioning models are essential for promoting financial inclusion in the digital age. By leveraging data, AI, and machine learning, financial institutions can make more informed credit decisions and expand access to financial services for everyone.