Lenders must obtain more individualised and context driven data for rural smallholder credit schemes to be successful. And with current developments in ‘alternative data’, and ‘data science’, new methods of risk analysis may be the necessary catalyst to strengthen lending opportunities for the more ‘riskier’ borrowers. As discussed in an article by Next Billion.
‘Alternative data’ is the sourcing and collation of data, whilst ‘data science’ refers to how this data is analysed and applied through new methodologies.
Models for measuring credit worthiness have historically relied on credit scoring, and income and relationship driven assessment. Yet data that establishes a borrower’s habits and behaviours can offer a new approach to data collection; subsequently ‘redefining the borrowing landscape’.
With mobile money, e-wallet and mobile utilisation data as their primary sources, data models so far have been largely dependent on heavy mobile phone usage to obtain information. This works well in the tech savvy environments of developed societies, but the limited digital climate within agricultural regions means that models must be adapted to a completely different demographic.
Next Billion identifies increased smartphone usage and improved rural connectivity as trends likely to improve the level of depth and insight into future data.
They also recommend three aspects that, given sufficient focus, could facilitate the necessary adaptation to smallholder borrowers:
- More collaboration with non-traditional actors that may have access to information about agricultural borrowers.
- More digitization of data from subsidy programs, censuses and extension worker programs that could be useful for borrower verification and more relevant credit algorithms.
- More context-driven appropriate investment that accommodates the longer product development and testing cycles of the agriculture sector.
To read the full article, click here.