Chiara PESCATORI opened the session by briefly introducing the panellists and asking them to explain what digital credit means to them and what they believe are the challenges faced by the industry.
Michael ROTHE was the first to answer, stating that digital credit has the potential to make microfinance more inclusive. He added that digital technologies can address several of the barriers faced by traditional FSPs when aiming to reach more clients. For example, digital technologies can reduce the need for collateral, as still required by traditional banks. Additionally, they can reduce operational costs of delivering services to the target group.
Anup SINGH noted that the digital credit sector still has some key concerns to resolve. According to Singh, many users of digital credit in Kenya are negatively listed and are now reluctant to use digital credit again. Another concern is that interest rates of digital credit are still high. The benefits of digitisation and automation in terms of cost reduction has not yet been passed on to the end-users. As confirmed by a member of the audience, Joris CRISÀ, digitalisation of financial services was expected to lead to a better understanding of clients, a reduction in transaction costs and a reduction of interest rates in the medium term. Based on high prevailing interest rates, it seems that this has not happened.
According to Ben WALLINGFORD, one of the main challenges is to make sure that FSPs are providing their services responsibly to attract more funding and to grow as an industry. Digital financial services face new types of risk including fraud, breach of data privacy and over-indebtedness due to aggressive marketing.
Wallingford stressed the importance of FSPs making proper use of low touch (little or no human interaction) and high touch (frequent and intensive human interaction). Depending on their business model, they must determine the extent of human interaction necessary for each of their procedures to provide a good service. Investors can encourage providers to apply best practices through their investments. Rothe confirmed the need for promoting responsible digital credit, as some investors currently seem to perceive digital credit as a dangerous investment due to activities of some irresponsible FSPs.
Pescatori then asked the panellists how digitalisation of financial services and client protection are best combined. Wallingford argued that the industry should make use of guidelines and standards for client protection, such as the Consumer Protection Standards for Digital Credit, developed by the SMART Campaign in collaboration with MFR. He provided some examples of best practices: Marketing staff should not employ predatory sales techniques; customer service staff should be available to customers not only by chat but also by phone; algorithms for automated loans have to be evaluated and proper systems must be in place to track algorithm effectiveness in predicting repayments; FSPs should ask for clear consent for data sharing such as geo location data.
Wallingford called for more partnerships between FSPs and third parties that can support to deliver digital services to ensure that the design of digital financial services improves. Singh responded that many FSPs are reluctant to share data with partners as they consider that data to be their secret recipe.
By using an example, Rothe explained how well-designed digital financial services can even enable lending during times of crisis. Digital technologies can make use of business data such as transactions of mobile money agents in Uganda, which have excellent predictive power. He added that the transaction data from these agents showed that they used funding from their investors to grow their business. This allowed digital FSPs to continue lending during the pandemic, while many traditional FSPs stopped their lending activities due to several limitations, including a lack of data for proper risk assessment.
Following the discussion on current opportunities and threats to the industry, Pescatori asked the panellists to share their views on the future of digital financial services. Rothe pointed at the opportunities for lending by digital financial service providers, highlighting that, when large investors become active in the digital financial services industry, the industry can grow much faster, as happened to microfinance.
Productive lending will require higher touch business models or a hybrid model combining high and low touch elements, according to Wallingford. For example, producers can submit an electronic application followed, by a bank staff visit in person to build a relationship and perform a qualitative assessment. He added that credit-scoring algorithms should include disposable income and other loans to prevent over-indebtedness, and that financial information such as mobile money transactions is useful for credit scoring. Wallingford also promoted verification of loan use, stating that the use of tick-boxes for self-reporting by customer may not be reliable. In summary, a more transparent and responsible industry applying client protection guidelines, endorsed by investors, to distinguish between good and bad practices will increase funding to the industry.
Singh mentioned several focus areas for improvement of design and delivery of digital financial services: FSPs require advice from partners on the use of hybrid models combining high and low touch to help them with the design of responsible credit delivery services; there is a need for insights into customer behaviour and the creation of a digital identity, as many small customers do not have a digital footprint yet; FSPs must apply appropriate marketing that protects customers; FSPs need to work with regulators on a coordinated approach to client protection through development of principles.
The panellists also shared their knowledge on the impact of COVID-19 on digital financial services. In Singh’s experience, the funding for digital service providers has been reduced and FIs have become more selective. Rothe explained the need for thorough customer research to understand working capital needs during the crisis. This requires an assessment methodology that uses local knowledge and qualitative field data.
Wallingford then shared results of a global survey on the impact of COVID-19. It showed a lower demand for new loans and more conservative financial service provision. He added that the crisis has had more impact on traditional financial services than on digital financial services. One of the problems for traditional service providers was the cancellation of group meetings with clients due to COVID-related restrictions, a problem that digital service providers did not have.
Audience member Alexandra SANCHEZ asked the panellists to share their thoughts on FSPs operating in countries that do not have a central credit bureau. Rothe responded that 92 percent of FLOW customers have never taken bank loans, and that FLOW is creating credit histories which could be used together with other information to prepare traditional credit scores.
Audience member Chris CZERWONKA wrote: ‘I hear about lots of good and responsible data collection that can help the use case of agents because their transactions and performance are the most readily available via MNOs. What about the wide range of other MSME business types where that data isn't so easily sitting on the surface for algorithmic use?’ Wallingford agreed that MSMEs have a smaller digital footprint compared to SMEs, and explained that algorithms can benefit from other kinds of data on business, such as typical financial returns for certain types of business in different countries. Singh added that SMEs can be supported to generate data for their digital footprint, and Rothe gave the example of a digital platform that registers every trip of a motorcycle taxi, which generates useful data for financing motorcycle taxis.