Banks to the rescue: Can IFRS9 contribute to improved access and lower interest rates for consumer credit in Africa?

On my first visit to Kenya, I was amazed by the comfort of using my phone to pay everywhere. Even the street merchant of mandazi (Kenyan pastry) is being paid via mobile phone. At that time in Germany we had to carry cash on top of our bank card, because even card payments were not widely accepted. Transferring money to your friend at another bank would take 3 days, while the same would be instant with M-Pesa, the market leader for mobile payment in Kenya with more than 80% market share.

With mobile payment came mobile lending. Especially for the low-income population, often unable to access bank loans, mobile lending came as a saviour. They could then enjoy a bit of the privilege of using their future income now to purchase goods and services. But this came at a relatively high cost for them. While middle-income customers and above pay interest rates between 9 and 30% at a bank, low-income borrowers must pay between 40 and 1250% per year for their loans at a mobile money lender.

The high interest rates of banks in developing countries compared to developed countries can be explained by a lack of efficiency and innovation. With increasing efficiency, interest rates should decrease over time and converge with the low interest rates of developed countries.

Mobile lending has brought innovation and efficiency since loan applications are submitted over the phone and processes are semi or fully automated using machine learning for borrower selection. Loan providers face lower costs, as they do not have to maintain physical offices. On the customer side, time restriction due to office opening hours are removed. Despite these innovations and efficiency improvements, we are facing extremely high interest rates in this segment against the expectation of convergence towards lower interest rates.

Some mobile lenders explain their high interest rate with the high risk of the segment they are operating in. However, the overall default rate of mobile lenders in Kenya was only 16% in 2018.[1] If we add a security margin of 4%, we can estimate the risk premium for the expected loss at a maximum of 20%. This estimation of the risk premium is based on the fact that most loans are very short term, from a few days to a maximum of one year and we assume a well-diversified portfolio, which is realistic looking at the number and amount of loans given. Factors such as rating migration are not taken into account, since all borrowers are assumed to be in one rating class with 20% default and loss given default is taken at a maximum of 100%, even if most of defaulters pay some instalments before defaulting. Banks, classified as deposit-taking financial institutions from a regulatory perspective, have to add a risk premium for unexpected loss, which is the same for all under the standard approach of Basel II/III. The Basel II/III capital requirement for unexpected loss is not relevant for unregulated mobile lenders, since they are not taking any clients’ deposits. Also, inflation can be neglected for short-term loans with maturities under one year.

Looking at a risk premium of 20% given expected losses in this segment, interest rates of 3 to 4 digits common in mobile lending cannot be justified. It would be highly desirable to improve access to — relatively affordable — bank loans also to lower-income groups while simultaneously reducing the administrative costs of small-scale loans for banks.

And indeed, the recent introduction of IFRS9 gives banks an opportunity to increase the efficiency of their lending decisions. The new accounting standard requires banks to estimate a forward-looking expected loss, by calculating the probability of default, the loss given default and the exposure at default, while accounting for macroeconomic impacts on the same. Banks now have an incentive to implement an internal credit scoring model instead of relying only on external credit scoring of credit reference bureaus for borrower selection.

With the estimation of expected loss at the individual level, banks can implement risk-adjusted pricing, which will have the effect of 1) reducing the effective interest rate for creditworthy borrowers and 2) increase accessibility to loans for less creditworthy borrowers by assessing their risk precisely and compensating it accordingly with a risk premium.

It is now up to the banks to get on this boat and compete with non-regulated lenders to provide affordable loans to the low-income population.

[1] Micro Save Consulting (2019) “Insights from Analysis of Digital Credit in Kenya”