Tala has been a pioneer in the alternative lending space. Founded by Shivani Siroya, the Santa Monica based company has demonstrated tremendous success in providing credit to people who are unbanked.
Tala uses machine learning from a user’s smartphone to predict a user’s capacity to borrow and the user’s associated repayment/default patterns[i]. Most individuals in emerging markets have no proper access to financial services. The World Bank estimates that nearly 60% of adults in Sub Saharan Africa, and nearly 45% in all developing countries are either unbanked or severely underbanked[ii]. Existing institutions do not have access to any digital data on potential borrowers, to enable them to understand the risk profile of potential customers. Even when they do have data, they often lack analytical capabilities to properly assess that data.
This environment has made it challenging for banks and microfinance institutions to provide proper access to credit to users. Tala uses an android application that requests users for access to certain information on their phones, including: SMS, call records, location patterns, etc. to get a better sense of the user. It claims to collect over 10,000 unique data points per user. The company then does experiments where it lends out unsecured loans to users and gathers data on repayment patterns from the users. It then combines the data on repayment patterns with the data gathered from the phones to get a better sense of the drivers of the default patterns.
So far, Tala has built over 1 million unique profiles for individuals. Most of its current users are in Kenya, though it has recently started expanding into other developing countries such as Philippines, Nigeria, and Tanzania[iii]. In late 2017, it raised a $30million round for its Series B and claims to be growing at more than 20% per month. The round is notable in that it was the largest Series B by a female founder in the US in the last few years[iv]. It is also cash flow positive, having reported a health profit in the same year. It is using the funds from the Series B to expand into other countries, particularly: India and Mexico.
While Tala has demonstrated early successes in using machine learning to power users, a few questions arise about its business model.
First, the company takes on a lot of balance sheet risk to provide the capital to the users. A large macro event affecting consumers in the countries it operates could lead to a surge in losses that could potentially bankrupt the company. This issue is exacerbated by currency issues since the company issues local currency denominated loans even though most of its funding is USD based.
Secondly, the unsecured nature of the credit makes it impossible for them to force borrowers to pay back. Third, it has proven very for Tala to move upmarket and provide loans to small business owners. The average loan size is currently $50 with a maximum of $500[v]. The unsecured nature of the loans it provides has made it difficult to move upmarket because the likelihood of fraud at this level is substantial.
Finally, Tala’s current model is susceptible to competition. For instance, Branch – which was started by Kiva’s founder Matt Flannery, was started in Nairobi two years ago and just completed a $70million Series B to go after the same market[vi]. The market is unlikely to be a winner-take-all market, and new entrants benefit from learning from the mistakes of Tala and Branch. Finally, changes in the regulatory environments may make it much harder for the likes of Tala to gain or retain access to the amount of information they currently have access to.