AI Could Make Small Trade Loans Pay Off Again

When a company sells goods to a buyer overseas, it often has to wait 30, 60 or 90 days to get paid. In the meantime, it needs cash to keep running. Banks have historically been reluctant to lend against that wait, especially for smaller companies, because the cost of checking a borrower’s paperwork and creditworthiness can outweigh the profit on a small loan. That gap between what businesses need and what they can actually borrow reached $2.5 trillion worldwide in 2025, unchanged from 2023, according to an Asian Development Bank survey of trade finance providers.

Smaller exporters get turned down 41% of the time, a rate that has nearly caught up to the 40% rejection rate larger companies face. The the bank cautions that convergence may partly reflect smaller firms giving up on applying rather than a real improvement in access, according to the same ADB survey.

Small Loans Just Got a New Set of Lenders

A group of FinTech startups is trying to solve the same problem faster than banks can. Comfi, based in the UAE, raised $65 million in April to expand across the Middle East and North Africa, the company said in its announcement. The company pays suppliers within 24 hours of an invoice, while giving their customers up to 90 days to actually pay, and has processed more than 15,000 invoices for over 1,000 clients.

“By combining AI-driven underwriting with disciplined risk management, the team has built a scalable platform designed for real-economy impact,” said Christos Mastoras, whose firm, Iliad Partners, led the round.

Hokodo and Treyd are doing something similar in different markets. U.K.-based Hokodo can approve a business buyer instantly, even on its first purchase, and offers payment terms up to 90 days, according to its website.

Stockholm-based Treyd focuses on helping retail brands pay their overseas suppliers upfront, giving the brand itself up to five months to repay, and makes credit decisions by connecting directly to a business’s accounting data.

Both companies, like Comfi, read a business’s invoices, payment history and other transaction data instead of relying on the slow paperwork review banks traditionally use.

The Real Test Is Whether the Math Actually Works

Making loans cheaper to process is only half the bet. The harder question is whether these companies can actually tell which borrowers are safe better than a bank can, or whether they’re simply lending faster to businesses banks already turned down for good reason.

If these companies really can spot good borrowers that banks miss, cheaper loan processing could open up trade financing to millions of small businesses currently shut out. If they can’t, faster and cheaper lending to riskier borrowers could turn into real losses once conditions get harder, the exact risk that made banks cautious about small trade loans in the first place.

Small businesses are already showing real appetite for these tools. Seventy-five percent of SMBs said they would use at least one AI feature offered by their financial institution within the next two years, rising to 83% among businesses generating more than $1 million in annual revenue, according to PYMNTS Intelligence’s “Credit Union Tracker Series,” PYMNTS reported. The demand is there. Whether the underwriting behind it holds up once the economy turns is still an open question.

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

The post AI Could Make Small Trade Loans Pay Off Again appeared first on PYMNTS.com.