An Open Letter to Central Bank of Kenya Governor Patrick Njoroge

Dear Sir,

We were delighted to see Reuters report that you believe that “the risks to Kenyans showed there was a need for regulation in the booming [digital credit] sector” https://reut.rs/2H0SRZW. While many of these small digital credit loans are immensely valuable for people facing emergencies, managing cashflow problems or for small scale trading, there are clearly significant downsides that deserve and need our collective attention.

We sympathise with your concerns that you do not like the idea of your country “being a guinea pig for new technology deployed by foreign companies”. However, the three largest digital credit lenders (M-Shwari, EazzyLoan and KCB-M-PESA) are very much Kenyan entities. It is fair to say that in Kenya, as in many other African countries, the development of digital credit was sponsored by international development agencies. This was done with good intentions, if little foresight of the unintended consequences. However, the continued celebration of the quantity of loans issued without reference to their quality is alarming.

And as MicroSave (https://bit.ly/2v5mbck) and CGAP (https://bit.ly/2ppVcXz) have shown, the quality of this portfolio is frighteningly poor. There is growing evidence that a worrying proportion of these digital credit loans used to finance the sports betting epidemic that is sweeping the continent (https://bit.ly/2LENfYG). Perhaps as a result, we saw in a recent study that borrowers default on a third of first cycle loans and are thus negatively listed on the credit reference bureaux – see graph.

As a result, as on March 31st 2017 10% of the adult population of Kenya was negatively listed on the credit reference bureaux – nearly a million of these for amounts of less than $10 – see graph below. Combining this data with that in the graph above we can surmise that now, more than a year later the number of people negatively listed will exceed 3 million as a bare minimum. Many remain negatively listed because of the Ksh.2,200 ($22) required for a clearance certificate to remove their names from the bureaux lists.

We have seen much discussion of the interest rates associated with these loans. However, these largely reflect the risk associated with making them given the high rates of default. Nonetheless, the effective interest rates for many are even higher than those published by many of the lenders. This is because the majority of borrowers repay well in advance of the monthly duration of the loans – see graph below. Sadly this is particularly the case for the one enterprise use case for these loans that is so often cited – including by officials at the Central Bank of Kenya (CBK) – that of day traders borrowing in morning to buy goods that they will often sell the same day. The typical month-long loan, with a fixed interest rate for that month, is simply inappropriate for these borrowers. Small wonder that 36% of these loans are repaid within a week – and sad that these borrowers receive no interest rebate.

As can be surmised from the above, there are many opportunities to re-engineer these loans to make them much more fit for purpose. Relatively small tweaks to the structure, marketing and pricing of the loans could increase transparency, reduce delinquency, help borrowers and thus actually increase the returns for the lending institutions. MicroSave outlined these in How Can Providers Make Digital Credit More Profitable?

Given the alarming number of Kenyans now carrying the “negatively listed” stamp, it’s time for policymakers and digital lenders to act. As a matter of priority, the Central Bank of Kenya (CBK) must insist on financial institutions using the credit reference bureaux as intended, as the reality of digital micro-loans amplify the impact of existing bad practices (e.g. incomplete submissions, human error, binary credit checks) on borrowers. The CBK might consider introducing and enforcing:

·        incentives for auto-submissions from bank databases to CRB to improve accuracy of the bureau data (TransUnion has already introduced an API for submissions)

·        an amendment for digital credit to the Credit Reference Bureau Regulations (2013) to incorporate a minimum loan amount for the listing given that the risk for digital credit is currently covered by the risk premium folded into the pricing

·        (as CBK has already stressed) requirement that lenders use nuanced credit scoring mechanisms basis the “balance” (value of the loan outstanding) and “number of days in arrears” fields in the CRB’s database, (instead of the binary “yes or no” check on negative listings), to prevent denial of credit by lenders. This would allow lenders to vary interest rates and loan sizes according to the credit history/risk profile of the borrower – as M-Shwari already does with loan sizes.

·        mandate for all off-shore, app-based lenders to report into the system (note that some, including Branch and Tala, do already report to the bureaux) on both negative AND positive results (many lenders report only those that default - the negative)

·        a requirement to that bureaux classify loans into more detailed categories to allow better understanding of the performance of different players and models (currently only M-Shwari is classified separately from all other loans)

·        regular audits of the CRB digital credit database for accuracy and completeness

·        accessible, efficient and cost effective mechanisms for consumers to check their credit history and correct any errors

·        clear guidance on clearance certificate requirements for financial institutions to follow (e.g. circumstances when those may be requested, the certificate fee schedule etc.)

·        mandatory requirements to ensure transparent terms and conditions are communicated to potential borrowers so that customers can make informed choices

·        minimum standards for customer recourse channels and coordination by partners to address issues/complaints raised by customers and thus drive long-term usage and customer loyalty

·        a curb on the aggressive, SMS-based, push marketing that is so common, thus reducing predatory lending

Kenya appears to have already started down this path with the proposals to restructure the financial sector through the Financial Markets Conduct Bill.

In anticipation of these changes in Kenya, and indeed throughout Africa, digital lenders should seize the opportunity to re-engineer their loan products to better cater to diverse needs of all potential digital borrowers. Some avenues for improvement include:

·        greater transparency and readability of Terms and Conditions (T&C)

·        examine options to improve understanding of T&C and reduce delinquency through more intuitive, user-friendly interface for all and oral customers, in particular

·        safeguards to prevent over-indebtedness and/or providing inaccurate information on the credit application (e.g. systematic use of CRB database to check a client’s loan burden)

·        credit scoring that allows lending with lower limits or higher interest rates to defaulters who have paid off their balances rather than shutting them out 

·        increased flexibility and utility to users, such as overdraft facilities, interest rebates for prior repayment, longer repayment terms, higher amounts, and interest rates that reflect good credit history

·        greater human touch in credit checks, disbursement, collections, and duty of care in case of delinquency all of which could be implemented through agents

Many of the digital credit providers (including those in Kenya and those based in South Africa and Silicon Valley) are already working on or have begun to implement these types of improvements. The IFC is working on responsible lending guidelines that will be valuable for the sector.

But a little regulatory push may help accelerate the pace of positive change.

Yours faithfully,

Anup Singh, Vera Bersudskaya, Zeituna Mustafa Abdi and Graham A.N. Wright