The SME Credit Desert: 2.5 Million Businesses, a $15B Funding Gap, and the Alternative Lenders Building the Bridge

The Thai banking system's SME lending problem isn't malice — it's architecture. Traditional credit underwriting is built around two signals: collateral (land title, property, equipment) and documented income history (tax returns, audited accounts). Thai SMEs — particularly the micro and small businesses that constitute the bottom 80% of the SME pyramid — frequently have neither. They own informal assets (the family home that isn't legally registered, the motorcycle that has no financing trail), they pay taxes inconsistently or not at all, and they have accounting records only to the extent required for their specific business relationships. The bank's systems say no. The actual credit risk is often more benign than the systems suggest.

What Alternative Credit Scoring Actually Uses

The alternative credit scoring approaches gaining traction in Thai SME lending share a common design principle: substitute formal documentation signals with behavioral and transactional data that is equally or more predictive of repayment probability. The most accessible and predictive alternative signals in the Thai context are PromptPay transaction history (payment regularity, income patterns, seasonal cash flow), Shopee and Lazada merchant data (for e-commerce SMEs — revenue trends, customer return rates, dispute rates), SCB Easy and KBTG app activity (spending patterns, savings behavior, bill payment consistency), and government contribution data (social security contributions are a proxy for payroll stability).

None of these signals individually substitute for a full credit assessment. In combination, normalized across a large borrower dataset, they produce credit scores that have been validated in pilot programs by KBTG (Kasikorn's tech arm), Ngern Tid Lor (one of Thailand's largest non-bank lenders), and several fintech startups against actual repayment performance. The machine learning models work — the data exists, the signal is real, and the underwriting capability is buildable. The challenge is assembling the data, building the distribution to reach SME borrowers, and pricing the credit risk correctly in a market where traditional banks have set rate expectations that don't reflect the actual cost of serving this segment.

The Supply Chain Finance Opportunity

Supply chain finance — lending to SME suppliers against receivables owed by creditworthy large corporations — is a particularly elegant solution to the SME credit problem because it substitutes the SME's own credit risk with the credit risk of their large-company buyer. A cassava processor that sells to a Thailand-listed food conglomerate on 60-day payment terms can access early payment of those receivables at a small discount — effectively borrowing against the conglomerate's credit quality rather than their own. The credit risk to the lender is the large company's, not the SME's.

This market is underdeveloped in Thailand. The large corporate anchor buyers exist (CP Group, Thai Union, Mitr Phol, SCG, and many others) and have extensive SME supplier networks. The fintech infrastructure to provide supply chain finance digitally — invoice digitization, real-time payment processing, yield management for the lender — is largely available from regional providers. What's missing is the platform that connects large corporate buyers' accounts payable systems to SME suppliers' working capital needs at sufficient scale and with sufficiently low friction that the economics work for both parties.

The Gig Economy Credit Gap

An adjacent and fast-growing segment of the SME credit problem is the gig economy worker: the Grab driver, the Lineman delivery partner, the Airbnb host, the Shopee seller. These individuals generate formal, documented income through platforms that have detailed earnings histories — but they are classified by most banks as self-employed with irregular income, making standard credit underwriting difficult. Platform-native lending is the solution: Grab Financial's GrabPay Later, Lazada's credit products, and Lineman's emerging financial services are all attempting to build credit products for their gig worker customer base using the platform's own earnings data as the primary underwriting signal.

The market size is substantial: Thailand has an estimated 3–4 million active gig economy workers, the vast majority of whom are underserved by traditional credit products. A lender who can accurately underwrite gig worker credit risk using platform data is accessing a market segment that is growing at 15–20% annually and has minimal formal competition.