In the evolving landscape of financial services, fintechs and lenders continually seek innovative approaches to assess and mitigate risks associated with lending. Traditional credit risk models, heavily reliant on credit scores and banking history, face significant challenges in accurately evaluating the creditworthiness of individuals with thin or nonexistent credit files. This group often includes newcomers, undocumented individuals, and the unbanked population. The limitations of conventional models underscore the need for alternative data sources to enrich risk assessment processes. Utility data, encompassing electricity, water, and telecom payments, emerges as a promising avenue. These recurring service payments play a critical role in our everyday finances, offering insights into an individual’s financial behavior and stability.

The Current State

• Late payments on utility bills often result in collections marks on credit reports, being the fourth most common reason. Yet, only around 3% of people with utility accounts have their consistent payment history documented in their credit files.
• Study showed that consumers with thin files, 89% could move to a score above 620 if Utility/Telecom payment data were to be included, indicating a shift from subprime to near-prime or prime risk levels.
• Utility bills in the U.S. have surged, with a 7% increase year over year, pushing the average cost to $351 monthly, consuming 7% of the average consumer's income.
• Nearly 20 million American households, or roughly 16% of homes, face an average utility debt of $788, signaling a crisis as energy costs soar to $16.1 billion in total owed.

Guide to Advancing Credit Risk Models with Utility Data

The integration of utility data into credit risk assessment frameworks presents a transformative opportunity for fintechs and lenders. This nuanced approach allows for a more comprehensive evaluation of an individual’s or business’s financial and payment behavior. By dissecting utility data into distinct categories, lenders can derive specific indicators that offer a deeper insight into a borrower’s creditworthiness and proactively identify fraudulent applications.

‍

1. KYC (Know Your Customer) Data

Data provided: Name, address, phone number, and email.

Value: This foundational data strengthens identity verification, crucial in combating fraud and ensuring compliance. Matching one or multiple identity sources to utility data is a reliable way to confirm principal residency and contact information like phone number or email.

• Identity Verification: Success RateMeasures the percentage of loan applicants whose identities have been successfully verified using utility data against the total number of verification attempts. This KPI helps assess the effectiveness of using utility data in the KYC process.‍
• Conversion Rate: Measures the percentage of total pre-qualified loan requests that actually convert to verified applications using utility data connectivity, compared to applicants that manually upload documents.‍
• Fraud Detection Rate: Tracks the number of fraud cases identified through inconsistencies or anomalies in KYC utility data compared to the total number of applications reviewed. A higher rate indicates effective fraud prevention.
  

2. Account-Level Data

Data provided: Utility provider name, account type (e.g., residential vs. commercial), account number, account age and the range of services subscribed to.

Value: The depth of an individual’s or business’s engagement with utility services can serve as a proxy for their economic stability. For instance, an older utility account with a history of multiple services will suggest a stable living or operational condition. In contrast, accounts with a high number of services and consistent or increasing utility spending could be seen as expanding or financially robust entities, indicative of lower credit risk.

Enhanced Application:

• Age of Account: A rule-based system can be implemented to flag accounts with an age under a specified threshold (e.g., 12 months). This can help identify newer, potentially less stable accounts. Flag accounts with an Age of Account less than 6 months as “New/Unstable.” and accounts older than 12 months can be considered “Established/Stable.”
• Utility Update Frequency: A rule-based system can identify consumers or businesses that frequently connect to new utility providers for repeated loan requests, which may indicate higher volatility due to changes in location or provider.

3. Payment Behavior

Data provided: Up to 24 months of invoice history, including invoice date, current balance, previous balance, new charges, last payment, payment date, and late fees.

Value: Analyzing patterns of invoice payments—including timeliness, frequency of missed payments, average outstanding amounts, and late fees incurred—yields insights into an applicant’s creditworthiness and potential early signs of financial distress. These indicators can be quantitatively assessed in predictive modeling to forecast future payment behaviors and the likelihood of default. Analyzing payment trends over time enables lenders to distinguish between a one-time financial mishap and a persistent trend of deteriorating financial health.

Enhanced Application:

• Payment Timeliness Ratio: Measures the proportion of payments made on time to total payments for utility bills, reflecting the applicant’s punctuality in payments. Calculation: (Number of On-time Payments / Total Payments) x 100

• Delinquency Trend Indicator: Tracks changes in the frequency of late payments over time, aiding in the identification of either deteriorating or improving financial behaviors.No specific calculation formula provided, but involves analyzing the change in late payments over selected periods.
• Average Payment Delay: Determines the average delay in days for utility bill payments, providing a quantifiable insight into payment behavior and financial liquidity. Calculation: (Sum of All Payment Delays in Days / Number of Late Payments)

• Utility Bill Coverage Ratio: (if income is known) Calculates the percentage of income dedicated to covering utility bills, with a higher ratio indicating more disposable income post-utility payments and thus, less financial strain. Calculation: (Total Income / Total Utility Bill Amounts) x 100

• Consistency in Utility Payments: Evaluates the regularity of payment amounts, with higher consistency suggesting stable consumption and financial stability. Calculation: (Number of Months with Consistent Payment Amounts / Total Number of Months Observed)

• ​Late Fees Incidence Rate: Measures the frequency at which late fees are incurred over a specified period, reflecting the regularity of late payments.This helps understand how often an applicant incurs additional charges due to late payments, which can indicate financial management issues or cash flow inconsistencies. Calculation: (Number of Payments Incurring Late Fees / Total Number of Payments) x 100

• Average Late Fee Amount: Calculates the average amount of late fees incurred per late payment, offering insights into the financial impact of late payments on the borrower. This metric allows lenders to assess the typical cost to the borrower for late payments, potentially reflecting the severity of the borrower’s late payment habits. Calculation: Total Amount of Late Fees Incurred / Number of Payments Incurring Late Fees

• Trailing Percentage of Amount Due: Evaluates the percentage of the total amount due that remains unpaid at the end of a trailing period, such as the last 3, 6, or 12 months. This KPI offers insights into the borrower’s ability to reduce their outstanding balances over time, indicating their financial health and capacity to meet financial obligations. Calculation: ((Total Amount Due at End of Period – Total Amount Paid during Period) / Total Amount Due at End of Period) x 100

• Average Due Amount: Measures the average amount due on utility invoices over a specified period, providing insights into the regular financial obligations faced by the borrower. Calculation: (Sum of All Invoice Amounts Due / Total Number of Invoices)

• Rolling Payment Performance Index: Tracks the change in payment performance over time by comparing the current period’s payment behavior to historical data. This KPI helps lenders identify trends in payment behavior, such as improving or worsening punctuality, which could impact credit risk assessment. Calculation: ((Current Period On-time Payment Rate – Previous Period On-time Payment Rate) / Previous Period On-time Payment Rate) x 100

4. Usage Data

Data provided: Detailed consumption data, either in terms of monetary expenditure or physical usage metrics (e.g., kilowatt-hours for electricity, gallons for water).

Value: Usage data reveals lifestyle patterns and financial health. High or low utility usage can indicate changes in a household’s size or a business’s operational scale, correlating with financial stress or prosperity. Consistent utility usage suggests a stable household for individuals, while for businesses, it can reveal trends, seasonal variations in volume, or growth phases, all critical for credit risk assessment.

‍

Enhanced Application:

• Growth in Utility Usage: Reflects the percentage change in utility usage, signifying household changes or business expansion/contraction. A notable increase for businesses may imply operational growth. Calculation: ((Utility Usage in the Current Period – Utility Usage in the Previous Period) / Utility Usage in the Previous Period) x 100

• Utility Consumption Stability Metric: The standard deviation of utility usage over a specified period, with a lower value indicating stable consumption patterns, a sign of financial or operational stability.
• Financial Commitment Ratio: Evaluates the financial burden of utility commitments by comparing utility expenses to income or revenue. Higher ratios could signal financial stress. Calculation: (Total Utility Expenses / Total Income or Revenue) x 100

‍

Technical and Analytical Integration with Datadeck

Datadeck transforms the way lenders connect to end-users utility data with a wide coverage of 3,000 providers across 30 countries. This section outlines a phased approach recommended to make the most out of this wealth of data to quickly enhance loan application experiences, reducing fraud, and gaining a competitive edge.

‍

Phase 1: Quick Wins with KYC and Basic Insights

The journey begins with integrating immediate value-adding features into the loan application process. Key among these is the utilization of Know Your Customer (KYC) data for robust identity verification. Additionally, basic insights extracted from utility data can serve as early indicators or flags for underwriters, simplifying the detection of potential risks. These insights are mostly rule-based, allowing for easy adoption and immediate application in the evaluation process.

By incorporating essential data features early, lenders can quickly capitalize on Datadeck’s offerings to decrease instances of fraud and loan defaults to grow their business and a valuable utility database.

‍

Phase 2: Leveraging Historical Data to advance modeling

As lenders grow more comfortable with the integration of utility data, the focus shifts to exploiting historical borrower utility datasets. This phase involves correlating borrowers’ utility data and past loan outcomes to develop sophisticated ratios and scoring models.

Utilizing historical utility data sets, lenders can construct complex ratios and scores that are more indicative of a borrower’s financial behavior and risk profile.

By analyzing the accuracy and utility of insights in conjunction with past loan performance, lenders can refine their models to ensure they are employing the most predictive and relevant data points for their specific market.

‍

Phase 3: Breaking data silos and ongoing monitoring

Building on the foundation of utility data integration and advanced analytics, Phase 3 enhances credit assessment with more data sources like telecom, banking, and income details, offering a comprehensive view of a borrower’s financial status.

This phase can also introduce proactive monitoring for high-risk or valuable customers, ensuring lenders can quickly identify and respond to signs of financial distress. By blending these diverse data streams, lenders not only boost the precision of their risk models but also tailor financial products more closely to individual borrower needs, maintaining current profiles for timely intervention.

‍