In a traditional lending model, banks and lenders have largely relied on financial data from balance sheets and credit reports to make lending decisions, but smaller businesses who have minimal credit history end up being ineligible. Incorporating alternative payment reporting data into credit scoring processes can provide a powerful tool for driving better financial inclusion and expand the number of undeserved businesses who qualify for loans.
MSMEs struggle to access credit in the traditional financing system
Micro, small and medium-sized enterprises (MSMEs) are an integral part of the Malaysian economy, providing goods and services, offering job opportunities, boosting market competition and enhancing innovation in the business landscape, among others. In 2022, MSMEs represented 97.4% of total businesses, contributing 38.4% of Malaysia’s gross domestic product and 48.2% of total employment.
Yet, in spite of their vital role in supporting economic activity, access to finance by most MSMEs remains an engrained problem. A recent survey conducted in 2023 by Small and Medium Enterprises Association (SAMENTA) found that roughly 35%, or one out of three SME companies had issues in securing funding due to limited collateral, a complex application process, high interest rates and lack of a strong credit history to demonstrate financial responsibility and ability to repay debts on time.
Restrictions in traditional credit scoring techniques
These challenges are associated with the norms of the traditional credit scoring approach, which uses the narrow metric of historical financial data such as payment history, cash flow and past business performance records to ascertain a potential borrower’s financial health and future ability to repay debt. Under this system, a business that manages its finances well and makes timely repayments will be more likely to have a good credit score; conversely, unsettled or outstanding debts and lay payments can indicate that the business may not be reliable or capable of making monthly loan repayments on time, leading to a poorer credit rating.
While the traditional credit scoring model has undeniable risk management properties, its limitation of analysis to narrow financial indicators such as past credit performance also means that it is unable to account for any contextual information that may be relevant to an applicant’s credit profile. Compared to larger, more established companies, MSMEs that have not built up sufficient volumes of data or a ‘credit history’ may find it harder to prove their reliability to repay a loan, and register poorer credit scores. This, in turn, leads to greater exclusion of MSMEs in the formal lending market, when ironically, these are the businesses in most need of financing support.
In the past three years, the COVID-19 pandemic aftershocks have left many MSMEs in Malaysia negatively impacted in terms of their operations and finances, making it difficult for them to survive and expand. These companies remain one of the most economically important segments of the business landscape, driving growth revenues and generating employment for the country. Against this, the importance of addressing the gap for MSMEs in mainstream financial services and unlocking sources of capital to better meet their financing needs cannot be underestimated.
New trends and opportunities in credit scoring
One way that financiers are beginning to innovate their lending systems is to explore new credit scoring methods alongside more traditional approaches, which make use of advances in big data analytics and emerging technologies.
In recent years, an approach known as “alternative credit scoring” has been gaining traction among financiers as a solution that can help identify unbanked or underbanked groups that do not meet the criteria for a traditional credit score but could otherwise prove to be good candidates of loan repayment.
How does alternative credit scoring work, and what data does it use?
As its name suggests, alternative credit scoring makes use of data beyond the parameters of conventional financial data such as a company’s financial statements and business plans. While the goal of alternative credit scoring is the same, namely, to underwrite risk and understand the likelihood of a borrower defaulting on a loan payment, the difference lies in the information and (sometimes) analytical techniques that inform these predictions.
Alternative credit scores draw from non-traditional data sources, such as utility bill and rental payments, purchase and payment data, psychometric profiling, social media profiles, sales trends, phone data and so on. These are data that can generate deeper insights into a company’s financial behaviour, going beyond the traditional financial statements and business plans to identify creditworthy borrowers who might otherwise be unqualified for traditional credit scoring. For example, a company with a strong track record of online business transactions and consistent payment of utility bills may reflect strong financial discipline and signal that it is reliable in repaying loans, even if it lacks volume in historical financial data.
Versus the traditional approach, a significant differentiator and strength of alternative credit scoring lies in its use of AI-powered machine learning to perform risk analysis. In the alternative credit scoring approach, alternative “data footprints” left by companies on online third-party providers, such as telecom companies, utility companies and social media platforms, are evaluated using AI-powered machine learning (ML) technologies. These technologies serve to extract value and find patterns from a broad range of data points, incorporating a wider range of variables and data to analyse and predict future financial behaviour. The overall effect is one of unlocking more nuanced insights on a company’s creditworthiness, based on its behaviour. Over time, as the algorithm is trained with more data, it will obtain better prediction accuracy on a company’s financial soundness and repayment capability.
Benefits of alternative credit scoring for borrowers and lenders
The benefits of using alternative credit scores mainly lies in its potential to provide greater access to loans and credit to a broader range of small businesses who may otherwise not qualify under traditional scoring methods.
For such companies, alternative credit scoring allows for them to add value to their existing credit profiles, by including additional financial information and data transactions that can demonstrate their repayment ability and willingness to pay. In the larger picture, alternative credit scoring can help to augment more traditional credit assessments, and solve the problem of borrowers who have a ‘thin’ credit file.
For lenders, there is an additional advantage of accessing new borrowers and expanding their loan market to include smaller companies that may struggle to qualify for loans under the conventional credit scoring system. The increased access to credit may also align with lenders’ financial inclusion goals.
Challenges and risks with alternative credit scoring
While alternative credit scoring brings innovation and the potential for improved accuracy in assessing creditworthiness, it also introduces certain risks that will need to be managed. Questions of transparency, bias and data privacy and security will remain paramount. The potential for bias and discrimination if the training data is skewed or unrepresentative, privacy and data security concerns, as well as data quality constitute some of the key concerns. To address this, it is important for regulatory bodies to establish clear guidelines and enforce stringent data sharing frameworks to create reliable and secure systems and ensure that alternative credit scoring is practised responsibly and ethically.
The future of credit scoring: A blend of the old and new
It should be recognised that alternative credit scoring models are not meant to scrap the historical data that underpin traditional credit scores; instead, they serve to expand and enrich the set of assessment criteria for financiers to rely on when processing loan applications.
While these are still early days, the scope of alternate credit scoring will likely widen with increasing data availability, advanced analytics and computing power. Significantly, the potential to provide a more inclusive credit evaluation system – one that considers the underbanked and expands the reach of loans to MSMEs who lack a traditional credit history – will likely be a driving force for the continued development of these models.
In the long run, broader use of alternative data in lending decisions could not only help bridge the funding gap for MSMEs but also benefit banks and other financiers, by expanding their loan portfolios and allowing them to extend services to underserved clients, while also empowering them to be more accurate and comprehensive in assessing the credit risk of existing clients. This will help improve automation and efficiency throughout the customer lifecycle, creating a mutually beneficial situation for both lenders and borrowers alike.
