Rethinking The SME Loan Underwriting Model In Southeast Asia

The small business credit gap is a global issue, a problem that banks and FinTechs continue to hack away at, yet one that remains stubbornly persistent.

In markets like Southeast Asia, high rates of credit-invisible, underbanked small businesses make filling the small business credit gap an even more difficult challenge.

At the root of the matter is often a lack of predictive data to ascertain the creditworthiness of these small- to medium-sized businesses (SMBs). Traditional banks will require credit histories and collateral to underwrite a small business loan, both of which many SMBs in Southeast Asian nations lack. It’s a self-perpetuating scenario that keeps barriers to capital in place.

In a recent conversation with PYMNTS, Raghav Mathur, head of data science and analytics at Singapore-based Grab Financial Group, discussed the opportunities in data technology that can address the region’s most pressing SMB lending needs.

Asymmetric Information Exchange

In Southeast Asian markets, banks’ lack of adequate data on SMB loan applicants is perhaps the tallest barrier to connecting SMBs to capital in the region. This asymmetrical information exchange, in which the SMB may hold valuable information that supports its loan applications, yet which is inaccessible or hard to aggregate for the lender, “has led to the [SMBs’] lack of access to formal finance,” according to a 2008 report by the Hachinohe University Research Institute and the Daiwa Institute of Research. The report added that for banks, the costs of gathering that data remains significantly high, with the volume of available information remaining low.

Grab’s “Know Your Business” processes are a mechanism that helps onboard small businesses to the Grab platform to enable Grab to offer value-added services. In exchange, Grab gains greater visibility into SMB operations and performance.

“By taking an interest in these businesses’ growth, we break asymmetric information cycles between platform SMEs and their financial situation [or] creditworthiness,” Mathur said. “For non-platform SMEs, we aim to learn as much about them through traditional sources of data, such as bank statements, credit reports and other available sources of verifiable information.”

Beyond Traditional Data

The data-gathering process certainly relies heavily on traditional sources, like credit bureau reports, to ascertain creditworthiness. But Mathur noted that targeting underbanked SMBs means expanding the scope of data collection and extracting predictive value from such data through technologies like machine learning (ML).

For instance, he said, examining the total monthly earnings in addition to monthly transaction frequencies of an SME on the platform (and its mathematical as well as machine learning derivatives) is an important way in determining an SMEs current financial health.

A merchant’s interaction with the platform, via merchant value-added services and channels including delivery drivers, front end app screens, also creates useful touchpoints that can be leveraged for data collection, verification and document delivery.

While internet presence, social media, digital references about SMEs offer dynamic insights, the use of advanced data science algorithms to discover more and useful data patterns from seemingly traditional data sources like bank statements gives a huge boost in understanding an SMB’s business.

Improving The SMB Lending Experience

A differentiated approach to SMB data collection yields enhanced risk mitigation and underwriting capabilities that can expand capital to SMBs that would otherwise be unable to access funding from a traditional bank.

Mathur emphasized that an augmented data collection approach can do more than simply connect SMBs to capital when they cannot access bank financing.

He pointed to Grab’s ongoing innovation of “bite-sized” products, which enables the company to offer smaller loans to SMBs that traditional banks find it costly to facilitate. These loans are “in-sync with their ‘observed ability and willingness to pay,’” he said, and they help “build trust and credit history, versus an absolute decline to any credit.”

“Offering smaller value loans can support an initial interaction between borrower and lender, help build borrower credit histories, and develop longer-lasting relationships that eventually lead to larger loan values as the business grows,” he explained.

With the small business credit gap continuing to plague SMBs across Southeast Asian markets, the opportunity to fill that gap through data technology is significant. Both traditional and alternative data, as well as data technology, are key to connecting businesses to capital. This is critical, regardless of whether an SMB is “credit invisible” or whether legacy underwriting tools are unable to push an SMB over the threshold to acceptance.

“With the increase in online access and smartphone penetration in [Southeast Asian] developing economies, we will continue to see an increasing demand for digital financial services such as digital lending products,” he said.