Search
Close this search box.

Insights

STUART THEOBALD: Big data and AI could open credit doors for informal sector

This column was first published in Business Day 

A big part of our economy is untouched by registered lenders: the informal economy. That is an unintended consequence of credit regulations that make it impossible to lend to anyone who can’t produce a payslip and three months’ bank statements. If you’re an informal entrepreneur, you either must fund yourself or turn to illegal and often unscrupulous lenders.

The informal economy could be an important driver of jobs and economic growth. It makes up 30% of the national economy and provides 31% of employment. Given the oft-stated policy intention of stimulating small businesses, the informal sector represents a huge opportunity for development.

The National Credit Act and its regulations stipulate that lenders must undertake an affordability assessment before granting a loan. That applies to consumers and to businesses that have a turnover of less than R1m, but because informal businesses are not registered they are effectively sole proprietors and therefore seen as individuals.

Because those individuals are not formally employed (unless their business is a side hustle) they have no way of producing the required documents. Street-side hawkers, for example, have to have working capital to fund their stock. The absence of any formal lending makes it impossible for many people to enter the market, or to do so at expensive rates.

Historically there have not been big protests from the lending industry. That’s because the informal marketplace has been difficult to lend to anyway. Records are often scarce, there is limited accounting or assets to hold as security, and informal businesses can be hard to identify or locate. Thay are not creditworthy on traditional ways of measuring it.

But technology is changing all of that. Tencent, in which Naspers is a major shareholder via its 74% stake in Prosus, owns WePay in China which has developed credit-scoring techniques that rely on their vast data on users’ social, purchasing and payment behaviour.

Facebook has a patent that allows it to use a person’s social network to evaluate a loan application (if your peers are good payers, you are likely to be too). Amazon has millions of small business customers and offers them loans based on a long history of purchase and payment behaviour.

Banks in SA have also tried to crack the informal market. Standard Bank experimented with a smartphone app that tracked informal traders’ cash flows and geolocated them to make loans, but it did that in Ghana rather than SA.

FNB and Standard Bank offer small business loans that are based on cash-flow analysis. Capitec is getting into the SME market after its acquisition of Mercantile Bank but it will focus on formal businesses that are above the R1m National Credit Act cap. It would love to get into the informal sector if credit regulations made it possible.

As in the rest of the world, alternative credit assessments relying on big data may happen outside the banking industry. Retailers, for instance, know what you are buying, information which could be relevant to creditworthiness. Is a shopper with a basket of fresh vegetables a better credit risk than one with two bottles of wine?

Ratios key

Affordability vetting is an odd thing. No lender wants to lend to someone who can’t pay it back. But in SA the efficient legal system means that you can pursue defaulting borrowers cheaply and recover your money at great harm to the borrower.

So, the affordability assessment provides a layer of comfort to block consumer exploitation by lenders. However, the national credit regulator then had to provide a detailed stipulation of how the affordability assessment should be done, otherwise lenders could simply cook up ones that give the answers they want. And so the payslips and bank statements became a requirement.

With the development of big data and artificial intelligence (AI) in finance, the affordability assessment is old fashioned and a big barrier for perfectly creditworthy borrowers in the informal sector. The problem is that there does need to be an objective way of determining if borrowers can afford their loan.

But that should be possible by stipulating a different sort of of assessment approach for the informal sector. Just as with big companies, ratios are key — the ratio of debt service costs to profit, for example. A lender relying on big data should be able to draw up a likely income statement and balance sheet from their analysis.

This is something that the department of trade & industry, which oversees the National Credit Act, and the department of small business development should be pursuing with gusto.

Freeing up the formal lending industry to support a sector that makes up 30% of the economy could have dramatic effects on overall growth. While there is resistance in some government quarters, in that the informal economy is also hard to tax or enforce employment regulations, the economic growth upside is potentially huge.