Credit ScorecardApril 30, 2021
Commercial Banks in Nepal have been endeavouring to expand their presence in the Small and Medium Enterprise (SME) lending space. This is driven by various aspects, important among them are increasing competition in corporate lending, bank’s strategic shift, the need to diversify risk, pressure on profit and regulatory push. The increased branch network of commercial banks in Nepal, especially after the country went into the federal structure, has also added an impetus for banks to expand SME lending. Most of the borrowing demands in these new locations, mainly remote areas, come from micro and small enterprises.
Like in many other developing markets, loan to corporate sectors is the mainstay of bank lending in Nepal. Some of the key reasons cited by banks on why they find it difficult to build a profitable SME portfolio are the high cost of service associated with SME lending and lack of information about the borrowers’ businesses. “Building a sizable SME portfolio is challenging because SMEs lack financial skills and documentation of proper income statements, balance sheet and operating performance,” said Niraj Kumar Basnet, Head SME & MF at Nabil Bank. Echoing Basnet’s view, Arjun Bhadra Khanal, Head SME and CBD, Siddhartha Bank (SBL) added heterogeneity of SMEs, lack of proper documentation, required knowledge for business sustainability and lack of additional acceptable securities as some of the problems banks face in developing a robust SME portfolio.
To build a profitable SME portfolio, banks have to underwrite a large volume of low value (ticket size) loans. Which in turn, leads to an increase in the cost of sourcing, underwriting and administration. The banks also have been following a lengthy process that is not customised to suit the SME segment.
At SBL, the SME lending process included a detailed credit assessment and the proposal for loan approval with a detailed customer business profile by a Relationship Officer (RO). The proposal would be prepared in a word document. “Lending was based on the personal judgment of the ROs who took into account the nature and performance of the business, promoter’s background, cash flow generated by the business and the availability of sufficient collateral coverage to cover the loan amount. A detailed analysis of financials where minimum standard ratios must be complied with would be done.”
Meanwhile, at Nabil, the process for SME lending was similar to the one practiced for mid-corporate and corporate loans. The process was manual and paper-intensive. Basnet further adds, “Owing to SMEs’ heterogeneity, we were unable to develop an optimal standard for assessing the loan application with ideal Turn Around Time (TAT). Other key challenges were lack of process and workflow automation and absence of an online platform.”
To overcome these challenges, banks use various tools to standardise their processes, bring objectivity in decision-making and automate processes. One of the tools often used by lenders to bring efficiency, objectivity and consistency in credit decision-making is ‘Credit Scorecard’.
One of the core objectives of the UKaid Sakchyam Access to Finance Programme is to facilitate SME financing and it has been working closely with market players to expand SME lending. As part of its SME financing initiatives, Sakchyam launched partnerships with two prominent commercial banks of Nepal — Nabil Bank and Siddhartha Bank – with the focus on approaching SME financing in a more structured way. The partnership with each bank includes establishing/expanding a dedicated unit for SME financing, and the development and rollout of the Credit Scoring model.
The Credit Scoring model is a statistical tool that uses the quantitative measures of the performance and characteristics of past loans to predict the future performance of loans with similar characteristics. Scoring calculations are based on payment record, frequency of payments, amount of debts, credit charge-offs and number of credit cards held. A weight is assigned to each factor considered in the model’s formula, and a credit score is assigned based on the evaluation. The factors can be changed and updated as per the necessity in the market.
Lack of high-quality data on the past loans of SME segment as well as low level of ‘bad’ loans led to the decision of adopting a scorecard based on ’Expert’ or ‘Judgemental’ model. As the name suggests, the model is based on the expertise of experienced lenders in identifying key characteristics associated with loan repayment. For this purpose, a Working Group was formed in each partner bank. The group included staff from Credit Risk, Credit Operation, Operational Risk, IT and Relationship Management. The Working Group of each bank was guided by a consultant from Sakchyam, Dean Caire, who has decades of experience in designing credit scoring models around the world.
Talking about the work process followed during the development phase of the product, Basnet said, “The group first studied the results of statistical analysis of internal tools and after detail deliberations formulated factors based on perceived importance and reliability. The scorecard was integrated with Management Information System (MIS), and User Acceptance Testing (UAT) was carried out before implementing it to users.”
Similarly, the SBL team identified and selected a set of quantitative and qualitative objective borrower characteristics that could best measure the credit risk of borrowers in the micro SME segment. “The identified multiple risk measures were used by the credit scoring consultant to build an equally-weighted scorecard for backtesting and expert validation,” added Khanal.
SBL did backtesting of the models on 50 ‘good accounts’ and 40 ‘bad accounts’. Further, for a period of around three months, all incoming credit applications were scored with the scorecard in parallel with the standard lending procedure. With the help of Caire, the model was further refined through several iterations. For the pilot testing, SBL used the Credit Scorecard for its SME product, ‘Siddhartha Saral SME Karja’. The product was launched in November 2020 with the scorecard throughout the country. “We plan to use the scorecard model for other products as well,” Khanal added.
Similarly, Nabil has been using the Credit Scorecard since January 2020 for all of its incoming application under the loan product, Nabil Sajilo Express Karja.
Understanding the Model
Introduction and integration of any innovative product which requires procedural changes bring its own set of challenges. Lack of expertise to operate and monitor the new product was a major challenge for the banks. Khanal from SBL said, “Initially, the staff were not familiar with the Credit Scoring model and there were also some elements of doubts on whether the model would provide accurate predictions. There were also difficulties in mapping the assessment of ROs to the decision made by Credit Scorecard.” With the help of Caire, both banks held training and orientation sessions for their ROs, Risk Analyst and sales staff.
Simplifying the Process
The Credit Scoring model has helped the banks simplify and streamline the lengthy SME lending process, with a significant reduction in TAT.
“The model has provided greater objectivity and transparency when approving loans since all borrowers must meet the same requirements. The quantitative database gathered through this model is helping us analyse the portfolio in many ways that were not possible before, and also automate the approval process”, said Basnet. Nabil has also expanded the use of the Credit Scoring model for a credit product launched in collaboration with a leading online platform Daraz, targeting their vendors. The model has helped the bank increase the volume of its SME portfolio in a short period.
The use of the scorecard has helped SBL classify the loan clients into five risk groups and design the loan facility accordingly. “Overall it’s expected to improve the accuracy of credit decisions as the bank makes uniform decisions whereby the possibility of rejection of creditworthy applicants has been significantly reduced,” mentioned Khanal. SBL has been generating monthly reports to monitor and evaluate the new borrowers against those studied during the development phase of the product. He maintains that such reports give the bank an opportunity “to check for anomalies and discrepancies in the model itself”.
A faster and simplified process that provides loan based on an objective and standardised data-driven decision-making method has been a welcome change for SMEs.
Nabil has onboarded more than 100 small borrowers through the Credit Scorecard product and has disbursed around NPR 200 million in loans. “The model has helped to improve the TAT of the bank from days to hours and has ultimately helped us achieve a larger SME portfolio,” Basnet added.
SBL has onboarded more than 600 small borrowers with sales below NPR 40 million and loan ticket size up to NPR 10 million through the model and has disbursed around NPR 2 billion in loans. The bank aims to cater to this particular segment as it is under-served/unserved and presents a potential for future growth.
For SBL, in the past, the lack of sufficient acceptable securities among SMEs had been a major hindrance in providing loans and thus preventing the bank from catering to numerous small businesses with promising prospects. “Objectivity and uniformity brought about by the scorecard have allowed the bank to experiment with the modality of linking the credit score with collateral matrix, where SMEs with a higher credit score are required to provide less real estate collateral coverage”, added Khanal.
Text by Nirmal Dahal, Team Leader, Sakchyam Access to Finance Programme