Egypt is moving towards a new phase in credit assessment as the Central Bank of Egypt expands testing of a behavioural credit scoring model involving five banks, while i-Score estimates that the approach could bring approximately 16 million additional citizens into the credit market and reduce default rates by around 50%, based on the company’s simulation results.
The developments were highlighted during a session titled “From Vision to Testing to Scaling: Behavioural Scoring as a New Path to Financial Inclusion” at the 13th edition of the Payments and Financial Inclusion Conference (PAFIX), held alongside the 30th Cairo ICT 2026 exhibition and conference. The session brought together representatives from the Central Bank of Egypt, i-Score, commercial banks and non-bank financial institutions to discuss the development, testing and potential expansion of behavioural credit assessment in Egypt.
The discussions focused on how alternative data and behavioural analysis could help financial institutions assess individuals who lack conventional credit histories but demonstrate consistent financial habits and the ability to meet their obligations. The approach is intended to complement traditional credit assessment methods, improve lending decisions and extend access to financing to previously underserved segments.
Ahmed Mounir, Head of the Regulatory Sandbox Unit at the Central Bank of Egypt, explained that testing the behavioural scoring model reflects the Central Bank’s commitment to supporting financial innovation, particularly applications involving artificial intelligence and alternative data, to improve banking services and expand financial inclusion.
He noted that the Regulatory Sandbox serves as more than an experimental environment for new financial technologies, functioning instead as an integrated supervisory framework through which innovative products and applications can be assessed before their wider market introduction. The process involves specialised teams from different Central Bank departments to evaluate the regulatory, supervisory and technical aspects of each initiative.
According to Mounir, behavioural scoring relies on analysing alternative customer data and identifying financial behaviour patterns to estimate an individual’s willingness and ability to meet repayment obligations. The model addresses two fundamental elements of creditworthiness: Ability to Pay and Willingness to Pay. While conventional credit assessment primarily focuses on financial capacity, behavioural scoring introduces additional indicators that help determine how consistently customers fulfil their financial commitments.
These indicators include payment patterns for utility bills, recurring obligations and other financial commitments, providing lenders with a broader understanding of customers who may not have previously accessed formal credit facilities.
The testing programme initially began on 15 February with Qatar National Bank Alahli (QNB Alahli), focusing on a limited group of customers, including payroll customers and individuals with guarantees or regular income sources. As the testing progressed and results were reviewed, the programme gradually expanded to include customers outside these traditional categories.
The initiative subsequently attracted additional participating banks, including Mashreq Bank, Commercial International Bank (CIB) and the National Bank of Egypt, bringing the total number of participating banks to five.
Mounir revealed that the initial target was to involve 1,500 customers in the regulatory testing programme, but actual participation exceeded that target by approximately four to five times. He stressed, however, that the success of the initiative would not be measured solely by the number of participants, but primarily by the accuracy and reliability of the model’s results, particularly its ability to assess creditworthiness and predict repayment behaviour across different customer segments.
The broader potential of the initiative was outlined by Sayonara El-Asmar, Deputy CEO and Managing Director of i-Score, who said the company’s simulations indicate that behavioural credit scoring could add approximately 16 million citizens to Egypt’s potential credit beneficiary base while reducing default rates by around 50% compared with current levels, without increasing credit risk.
El-Asmar explained that although Egypt’s formal credit bureau infrastructure has made customer information more accessible, conventional credit assessment remains heavily dependent on individuals with previous borrowing or credit-related transactions. This has limited financial institutions’ ability to evaluate people who have never established a traditional credit record, despite potentially demonstrating positive financial behaviour.
To address this gap, i-Score has worked on developing and improving its databases, data quality and analytical capabilities to support more advanced credit assessment models. The objective is to move beyond using customer information primarily for Know Your Customer (KYC) procedures and administrative requirements towards using it as an effective tool for evaluating creditworthiness.
She emphasised that the behavioural scoring model relies on trusted government databases and information connected to the country’s digital infrastructure, rather than social media platforms or unofficial data sources. Before being incorporated into analytical models, the information passes through several Quality Gates designed to assess its completeness, accuracy, consistency and frequency of updates.
The alternative data used in the model includes information related to customers’ residential addresses, workplaces, employment duration and social insurance records, as well as whether they own agricultural land or participate in commercial and industrial activities. It also incorporates payment behaviour related to electricity, gas and telecommunications bills, providing additional indicators of how consistently individuals manage their financial obligations.
To verify the model’s reliability, i-Score conducted simulations using available customer information and compared its outputs with those generated by conventional credit scoring models. The company also conducted Out-of-Sample testing across different periods to assess whether the results remained consistent and were not influenced by seasonal variations or exceptional circumstances.
The testing also included No-Hit Customers, referring to individuals who had no previous credit history when their information first entered the company’s database. According to El-Asmar, comparisons demonstrated significant alignment between the outputs of traditional and behavioural assessment models, supporting the potential use of alternative data in developing credit evaluation methodologies.
Regarding the potential size of the market, El-Asmar pointed out that approximately 78 million Egyptians are between the ages of 18 and 65, representing a substantial population that could initially be assessed for potential credit eligibility based on age. However, she clarified that i-Score did not automatically classify all these individuals as eligible borrowers, instead selecting specific segments according to behavioural patterns and characteristics that demonstrated suitability for the model.
She added that the initiative’s importance extends beyond increasing the number of borrowers, as it could also improve the accuracy of credit assessments conducted by banks and non-bank financial institutions, supporting responsible lending and deeper financial inclusion.
From the banking sector’s perspective, Mohamed Ragai, Head of Retail Banking Risk at QNB, said the bank had participated in the initiative from its early stages, recognising the importance of developing credit assessment mechanisms and expanding access to financial services.
He explained that the banking industry has gradually moved away from relying primarily on experience-based assessments and traditional lending practices towards more sophisticated data-driven models and analytical tools, while maintaining established risk management standards and individual banks’ credit policies.
QNB’s participation initially involved providing the necessary customer data for research and testing in cooperation with i-Score’s specialised teams. The process included analysing historical customer information and examining previous financial behaviour to understand how payment patterns and financial commitments could influence lending decisions and portfolio risks.
Ragai identified determining the appropriate target customer segments as one of the main challenges, noting that banks differ in their risk appetite and the criteria they apply when approving financing applications.
He stressed that behavioural scoring should serve as an additional component of credit assessment rather than replacing traditional models. Banks will continue to rely on their existing credit policies and assessment tools while incorporating behavioural data as another input into lending decisions.
According to Ragai, regular payment of financial obligations, including banking instalments, utility bills and other recurring commitments, can provide valuable indicators of an individual’s financial discipline. He added that expanding access to financing requires carefully identifying target groups according to factors such as age, income levels and customer characteristics, while ensuring that financial products are aligned with borrowers’ actual repayment capacities.
Dalia El-Shawaat, Head of Retail Banking Risk at CIB, highlighted four principal benefits of behavioural credit scoring for the banking sector: expanding customer bases, improving the prediction of future defaults, increasing financing approval rates and supporting sustainable banking growth.
She explained that alternative data allows banks to reach individuals who may have previously been excluded from financing, not necessarily because they represent higher credit risks, but because traditional assessment methods lacked sufficient information to evaluate their financial capacity.
The model could also improve banks’ ability to classify customers according to expected risk levels, establish more accurate default targets and increase financing approvals while prioritising the quality of approved applications rather than simply expanding lending volumes.
El-Shawaat added that behavioural data could help banks determine appropriate credit limits according to individual repayment capacity. Under a gradual approach, newly approved customers could initially receive relatively low credit limits, which would increase as their repayment behaviour is monitored and their creditworthiness becomes clearer.
This Test and Learn methodology allows financial institutions to build customers’ credit records progressively and assess their actual performance before extending larger financing facilities. She also noted that alternative data could benefit existing bank customers by providing additional information that helps institutions better understand their financial circumstances, particularly when conventional documentation does not provide a complete picture.
The potential impact extends beyond the banking sector to non-bank financial institutions, which could benefit from more comprehensive customer information to improve lending decisions and manage risks.
Amr Aboul Azm, CEO of Erada Finance, said behavioural scoring and alternative data represent an important development for non-bank financial institutions, particularly given their need for more detailed customer assessments and their comparatively limited ability to distribute risks across different sectors and portfolios.
He explained that non-bank financial institutions operate under a dual regulatory framework, with direct supervision from the Financial Regulatory Authority and indirect oversight from the Central Bank of Egypt through the requirements governing banks’ financing of the sector. Banks currently provide more than 85% of the sector’s financing needs, making the quality of non-bank financial institutions’ credit portfolios particularly important.
Aboul Azm noted that alternative data provides companies with a more comprehensive understanding of customers through approximately 360 data points covering financial, social and legal information, alongside indicators related to their daily financial behaviour and obligations.
These indicators can include regular payments of electricity and gas bills, as well as other financial and legal information that may help institutions identify potential risks and assess customers more accurately.
He added that non-bank financial institutions require sophisticated risk assessment systems because they do not enjoy the same flexibility as banks in diversifying risks across large and varied portfolios. More accurate assessment tools could therefore help them reach new customer segments, develop financing products suited to different needs and strengthen the confidence of banks that provide their funding.
Aboul Azm also highlighted Erada Finance’s XFin system, developed in cooperation with i-Score, which brings together financial, social and legal information and monitors customer-related alerts to convert different data sources into indicators that support financing decisions.
The system builds on developments in data management over the past decade, particularly within the microfinance sector, reducing dependence on individual indicators and enabling a broader assessment of customer creditworthiness.
Erada Finance offers several financing products, including microfinance, small and medium-sized enterprise financing, Islamic financing, and Supply Chain Finance. According to Aboul Azm, alternative data allows funding banks to assess each product separately, identify suitable sectors and customer segments, and determine appropriate financing amounts and costs according to their respective risk profiles.
He concluded that the value of alternative data extends beyond profitability, as it can help address the information gaps that prevent traditional banks from reaching certain population segments, particularly where the cost of assessing potential borrowers is high or conventional credit information is insufficient.
The discussions at PAFIX highlighted how behavioural credit scoring could become an important component of Egypt’s evolving financial infrastructure, combining government-backed data, artificial intelligence and advanced analytics to improve lending decisions. With five banks participating in the Central Bank’s testing programme and i-Score identifying the potential to bring millions of additional citizens into the credit market, the next phase will depend on validating the model’s performance across different customer categories and ensuring that wider adoption maintains an appropriate balance between financial inclusion and credit risk management.
PAFIX, the Payments and Financial Inclusion Conference, is one of the specialised events accompanying Cairo ICT and is held under the patronage of the Central Bank of Egypt. The conference focuses on developments in digital payments, financial services and banking technology, providing a platform for banks, financial institutions, technology companies and payment service providers to showcase innovations, explore cooperation opportunities and discuss solutions that support the development of Egypt’s financial ecosystem.
