New Credit-Risk Models for the Unbanked

> Posted by Tony Goland, McKinsey & Company

This post is part of the Center for Financial Inclusion’s Expert Exchange: Building A Movement Toward Financial Inclusion by 2020, cultivating conversation around the goal of reaching full financial inclusion by 2020. For further questions about this series, write to Sonja E. Kelly, Fellow, Center for Financial Inclusion at ACCION International.

There are more than 2.5 billion people worldwide who live beyond the reach of formal financial services. Most of them come from lower-income households and/or run small and informal enterprises. For these households, credit is an important component of financial independence. Used responsibly, it enables investment in income-generating activities and enables better short-term liquidity management.
Lending to otherwise unserved households is challenging, not only because of their limited familiarity with formal financial services, but also because lenders often have little to none of the data they might traditionally use to assess the risk of lending to them. Conventionally you would look to a borrower’s banking and credit history. This tactic has proved less effective in emerging economies, especially among poorer customers who often have no record of past borrowing or earnings and lack formal savings or assets that could serve as collateral.
An alternative approach has emerged that would help lenders to build better risk models and help borrowers to receive improved and affordable products and services. Being pioneered by a handful of mobile operators, utilities, retailers, and direct-sales companies, the approach entails tapping into new forms data spun off from their core businesses to lend in responsible, low-touch and low-cost ways.
How does it happen? It involves lenders leveraging increased computing power and new sources of information and data, such as mobile-phone usage patterns and utility bill payment history to assess creditworthiness. The potential to reap this information in emerging markets is enormous. For example, by the beginning of 2009, these countries accounted for approximately 75 percent of the world’s four billion mobile phones.
We have found that lenders will benefit from exploring six sources, among others, for this new information and data:  

  • Telecommunications providers (e.g. data about financial transactions done via mobile phone serve as indicators of cash flow)
  • Utilities (e.g. whether or how often bills are paid as a proxy for willingness and ability to repay)
  • Wholesale suppliers (e.g. payment histories for small businesses as a proxy for revenue estimation)
  • Retailers (e.g. data about customer purchases can help to estimate income levels)
  • Government (e.g. demographic and census data can indicate default risk)
  • Financial institutions’ own, previously overlooked data (e.g. paper records that have not been digitized).

Indeed, credit risk innovators such as Brazilian wholesaler Grupo Martins and the Chinese B2B internet company Alibaba have already begun to successfully use such non-traditional data.
Beyond identifying these new data sources, lenders should pursue two additional steps to develop effective credit assessment strategies and approaches for lending to lower-income households (and micro enterprises too). First, they need to secure access to the appropriate data. The simplest – though least cost effective— way is to pay for it. A better solution may be to strike partnerships with mobile companies, retailers, utilities and/or wholesalers to gain access in ways that benefit all parties.
The other requirement is for lenders to convert data into actionable credit insights. Many consumer lenders have advanced credit-risk modeling capabilities, but incorporating new data into strategies, models, and processes will require some major changes in people, technologies, and workflows. Three areas to target for change are talent, information technology, and the collaboration between risk and marketing teams.
Of course this new approach is not without shortcomings. Non-traditional data must sometimes be gathered from diverse sources, and the volume is often several times that of traditional sources. For example, each mobile account may generate hundreds or even thousands of calls and text messages per month, each carrying a rich data set that includes when the call was made, the location of the caller at the time of the call, the type of information accessed via text messaging, and the types and number of payment transactions made through the device. This indicates that successful lenders who lever such data typically are not unsophisticated and in fact string together a very thoughtful business system that is tailored to the peculiar opportunities and constraints of each situation, carefully resolving which partner is doing what with each piece of data.
Another challenge is that many practitioners are not yet skilled in new data standards and protocols and new tools that bring together disparate data sets, matching and comparing them to generate insights. They are also unfamiliar with aggregating diverse and oblique data to derive meaningful insights. And finally, gaining access to data can be difficult as well. In many cases, the data sets that lenders want are owned by entities that may not want—or are not allowed—to share them. In markets where privacy laws are well-established, regulatory requirements and privacy laws may restrict lenders from gaining access to and/or fully using certain types of information.
Challenges notwithstanding, this new approach enables a more complete understanding of lower-income clients’ financial needs and behaviors. With that understanding, providers can move beyond simple lending to help customers make good financial decisions, offer the right noncredit products and conduct marketing and communications in ways that are more likely to resonate for distinct segments.
Until now, neither microcredit nor traditional consumer finance has sufficiently served the diverse needs of economically active lower-income families and smaller/informal businesses on a truly sustainable basis. Today we inarguably have a more viable approach on the horizon.
This blog is based on research conducted by Tobias Baer, Tony Goland and Robert Schiff of McKinsey & Company. The complete report can be found at http://mckinseyonsociety.com/new-credit-risk-models-for-the-unbanked/
For more information, sign up for updates from the Financial Inclusion 2020 campaign.
Tony Goland, McKinsey & CompanyTony Goland is a Director in the Washington, DC Office of McKinsey & Company. He is a senior leader of the Americas Financial Institutions Group and Organization Practice, and he leads the Firm’s global efforts to promote the availability of quality, affordable financial services among the world’s economically active poor.

<noscript><iframe src="https://www.googletagmanager.com/ns.html?id=GTM-W5XP2HF" height="0" width="0" style="display:none;visibility:hidden"></iframe></noscript>