PROJECT

CREDIT SCORING MEETS SMALL AGRICULTURAL LENDING: EXONGENOUS SHOCKS, RECOVERY, AND ACCESS TO FORMAL CREDIT

Researcher: Nicolás de Roux, CDEP Fellow and PhD Student in Economics

Starting in the 1950s in the U.S., credit scoring spread quickly across the banking industry as a tool to assess the likelihood of repayment by prospective clients. Today credit scoring is used in many segments of the lending industry, including small business lending. Formal lending institutions in developing countries followed suit and to date many banks and micro-finance institutions targeting small enterprises collect credit histories and scores.

This project investigates the implications of the use of credit histories and credit scores for small agricultural lending. Credit scoring is used to assess the risk of prospective borrowers, but in the context of small agricultural production part of this risk is outside the farmer’s control. The reason is that the productivity of crops is highly dependent on weather and in many scenarios, on rainfall realizations. Rainfall shocks lead to volatile income streams and, for a farmer who has a formal loan, it can precipitate loan default causing worse credit histories and credit scores and higher chances of future loan denial. This can carry costs both to the farmer and the lender if on average the income of the farmer recovers from rainfall shocks faster than his credit access.

To empirically study these issues, this project uses data from formal agricultural loans for coffee production in the country of Colombia and shows that excessive rainfall shocks cause lower loan repayment, lower credit scores, and more frequent denial of subsequent loan applications. Furthermore, drawing on the agronomic literature on coffee production and using survey data, it shows that income recovers faster from these shocks than farmers' credit histories. Since rainfall shocks are unrelated to a farmer’s ability, productive farmers can be driven out of the credit market leading to costs both to the farmer and the lender that could be in principle avoided. These results suggest that credit scoring should be adjusted to account for the more volatile nature of income of small scale agricultural production in developing countries.