Looking Beyond Machine Learning To Balance Out Trade Finance Supply And Demand

The International Chamber of Commerce Banking Commission recently released a report that found an imbalance between supply and demand of trade finance services. Its Global Survey on Trade Finance, which surveyed financial institutions across 98 countries, found that 61 percent of banks say they face greater demand for trade finance than they can supply.

More than two-thirds told researchers that compliance and regulatory requirements are holding them back from providing more trade finance in the short term, while cost control pressures were identified as the top challenge for FIs’ (financial institutions) trade finance operations.

Indeed, banks must tread carefully in the world of trade finance, and with such little room for error and financial losses, risk management is critical. In many ways, collaboration with FinTechs has become a key part of risk mitigation for banks, with researchers finding that only 1.4 percent of bank respondents said they consider FinTech’s competitive offerings as posing a threat to traditional banks’ position as trade finance providers.

“The discourse around FinTechs is evolving from competition to collaboration,” the International Chamber of Commerce Banking Commission (ICC) concluded.

One of those collaborators is CRiskCo, a credit risk management company that deploys Big Data analytics to provide a business credit score to lenders and trade finance providers. The firm’s President, Erez Saf, recently spoke to PYMNTS about how these FIs can wield powerful technologies to mitigate risk and remain profitable as the industry works to meet that demand gap.

According to Saf, the adoption of technology by trading partners can be a significant aid to both CRiskCo and the FIs with which it works.

“Since the majority of SMBs’ accounting systems moved to cloud-based solutions, such as QuickBooks or Xero, our ability to identify, calculate and apply complex mathematical models on the data improved dramatically,” he recently told PYMNTS. “As the world becomes more digitalized and connected, our abilities will grow, and we expect the data flow of connected transactions will cover many of the industries that have been left out until today.”

SMEs are often some of those left out from access to trade finance, especially as tighter capital controls, risk requirements and regulations force larger banks to pull back from servicing small businesses.

CRiskCo recently announced news of a partnership with Euler Hermes Digital Agency (EHDA), a unit of trade credit insurance company Euler Hermes that focuses on technology and disruption of the trade finance space. CRiskCo is integrating its analytics capabilities into the EHDA platform, aggregating accounting data and deploying machine learning to predict whether a small business borrower may default.

Together, the businesses are going straight to the trading partners to protect a supplier against the risk of non-payment. CRiskCo’s underwriting capabilities have been integrated into EHDA’s Single Invoice Cover solution, which protects businesses against the threat of non-payment on a single bill, with coverage provided on a transaction-by-transaction basis.

For smaller suppliers especially, that risk can be detrimental. But it is also complex, with the risk mitigation process having to take into account both the challenge for a company to pay its supplier and the challenge for a supplier to be paid, as financing can be placed on either end.

“From a risk perspective, we pay more attention to the business’ ability to have good customers with low-risk and steady payment behavior,” Saf said. “As the business receivables look better, it’s usually easier to help the business with insurance products like ‘Single Invoice Cover’ … or a financing vehicle such as factoring or commercial loans, than a business with fewer customers.”

CRiskCo takes into account a number of factors when assessing this risk, the executive explained.

“We find that more traditional methods of assessing risk tend to use historical data that is outdated and irrelevant,” he noted, adding that the company looks at the stability of client payments, the correlation between sales and debts, sale trends and more.

“For example, a client with Net+60 credit terms that paid every month on day +75 to +85 has a lower risk than another client with Net+60 that has a payment history in the past four months of days +55, +68, +75 and +93,” Saf said.

In other words, consistency may be indicative of less risk than a business that pays inconsistently, even if they sometimes pay sooner.

“Most CEOs, CFOs and credit managers will identify by a hunch their riskier customers,” he continued, “but without a continuous system that can demonstrate and compare all the factors together on a daily basis, it is tough to identify a problem at the right time and stop sales or change the credit terms before it’s too late.”

CRiskCo already deploys machine learning and predictive analytics into its solution, but, according to Saf, there is potential for other technologies to disrupt the credit underwriting process for the trade finance market.

“We are just scratching the surface of ML (machine learning) and data analytics, and we have a lot more to do to achieve a fully learning machine [that] can set its own goals for improvement and identify what it misses to be better,” he said. “But we are keen to explore more technology, and the next technology for us to explore is Deep Learning and Neuro Networks, which can help us better differentiate different industries, markets and locations.”

“If we look one step further, the entry of blockchain technology will gain us the ability to track a full history and path of each invoice/product and will reduce our efforts on fraud detection and identity identification dramatically,” he continued. “It will grow to be very exciting in the coming years.”

According to the ICC, traditional banks have a slightly different view of the future of trade finance: Most said the industry will exhibit little or no growth, and just 12 percent said they are anticipating market update. But considering the rise in FinTech collaboration for trade finance operations, it may be up to FinTechs like CRiskCo, Euler Hermes and others to pull the banks into an era of more efficient, less risky trade finance and potentially address that supply-and-demand imbalance.