Leveraging Open Finance Data for Credit Underwriting in the UAE

The Underwriting Opportunity

Traditional credit underwriting in the UAE relies heavily on credit bureau data and salary transfer records. While effective for salaried employees with established credit histories, this approach creates barriers for:

• SME owners whose business cash flows don’t translate to personal credit scores
• Gig economy workers with irregular but substantial income
• New-to-bank customers without local credit history
• Young professionals early in their financial journey

Open Finance unlocks a new dimension of data: real transaction behavior.

The Data Advantage

What Open Finance Reveals

Through consent-driven access to transaction data, underwriters gain visibility into:

Data Signal	Underwriting Insight
Regular income deposits	Income stability verification
Rent/utility payments	Payment discipline patterns
Savings behavior	Financial buffer assessment
Spending categories	Lifestyle and risk indicators
Account balances over time	Cash flow management capability

Complementing Traditional Sources

Open Finance data works best when combined with existing data sources:

┌─────────────────────────────────────────────────────────────┐
│                   UNDERWRITING DATA FUSION                   │
├─────────────────────────────────────────────────────────────┤
│                                                              │
│   Credit Bureau  ──┐                                         │
│   (AECB/Etihad)    │                                         │
│                    ├──► Unified Risk    ──► Credit Decision  │
│   Salary Transfer ─┤    Assessment                           │
│   (WPS)            │                                         │
│                    │                                         │
│   Open Finance  ───┘                                         │
│   (Transaction Data)                                         │
│                                                              │
└─────────────────────────────────────────────────────────────┘

Practical Implementation Considerations

1. Consent Architecture

The customer must explicitly consent to data sharing. The consent flow should:

• Clearly state which accounts will be accessed
• Specify the purpose (credit assessment)
• Define the data retention period
• Allow granular control where possible

2. Data Categorization

Raw transaction data needs enrichment through:

• Merchant Category Codes (MCC) mapping
• Counterparty identification for recurring payments
• Income source classification
• Expense categorization algorithms

3. Risk Signal Extraction

Key signals to derive from transaction data:

// Example: Income Stability Score
const incomeStabilityScore = calculateScore({
  regularityOfDeposits: 0.3,    // How consistent are income deposits?
  varianceInAmounts: 0.2,       // How much does income fluctuate?
  numberOfSources: 0.2,         // Diversified or single source?
  trendOverTime: 0.3            // Growing, stable, or declining?
});

4. Model Integration

Open Finance features can be integrated into existing credit models as:

• Additional input features in ML models
• Override signals for edge cases
• Segmentation criteria for thin-file applicants
• Validation layer for stated income verification

Regulatory Alignment

This approach aligns with CBUAE’s Open Finance objectives:

“Enable consumers to share their financial data securely with third-party providers to access better financial products and services.”

The framework supports financial inclusion by enabling credit access for underserved segments while maintaining robust risk management.

Next Steps for Implementation

1. Define use cases - Start with a specific product (e.g., personal loans for thin-file customers)
2. Map data requirements - Identify which transaction data points are needed
3. Build enrichment pipeline - Categorize and normalize raw transaction data
4. Develop risk features - Create derived variables for model consumption
5. Test incrementally - A/B test against traditional underwriting
6. Iterate based on performance - Refine signals based on loan performance

Related Resources

• CBUAE Open Finance Regulation
• Al Tareq API Documentation
• UAE Al Etihad Credit Bureau integration patterns