MICROFINANCE PAPER WRAP-UP: Banks and Microbanks, by Robert Cull, Asli Demirgüç-Kunt & Jonathan Morduch

Written by Robert Cull, Asli Demirgüç-Kunt & Jonathan Morduch, Published by the Financial Access Initiative in September 2009, 54 pages, available at: http://financialaccess.org/sites/default/files/CP%20Banks%20and%20Microb…

This paper is intended to determine whether or not the presence of banks in 38 developing countries affects the profitability and outreach of microfinance institutions (MFIs). In other words, does competition from larger banks cause MFIs to lose some of their more “profitable” customers, resulting in a smaller average loan size and depressed profits? A smaller average loan size, as well as an increase in women customers as a result of competition, would be indicators of a greater depth of outreach in terms of “poor and excluded groups,” according to the authors. Bank penetration is measured in both “branches per capita” and “branches per square kilometer.” Data from 328 MFIs was used for this study.

The main results are as follows: An increase in bank penetration, when measured by branches per capita, is associated with an decrease in loan size for MFIs. There was no significant result for branches per square kilometer. Conversely, bank penetration, when measured by branches per square kilometer, is significantly correlated with an increase in the share of women borrowers for MFIs, with no significant result for the “branches per capita” measure. As stated, the authors consider both of these results to be evidence that competition from banks causes MFIs to have a greater outreach to the niche markets of poor and excluded borrowers. In terms of profitability, increased bank penetration (both measures) is weakly correlated with return on assets for MFIs, and not significantly correlated with a “financial self-sufficiency index” based on whether nor not revenue covers costs. This does not provide particularly strong evidence for the hypothesis that competition decreases MFI profits.

Certain controls used for the regressions also had important results. For example, MFIs that employ village banking or group lending models in which there is joint liability among clients have smaller loans and a higher share of women clients to start. Therefore, they are less likely to be affected by bank penetration (both measures) in terms of profitability and outreach as compared to those MFIs that employ traditional, bilateral lending contracts.

The authors then run several other regressions to determine if there are alternative explanations for their results. For instance, because MFIs may choose to enter markets for reasons related to pre-existing high or low bank penetration, a regression was run using only “established” (since 1996) MFIs to avoid the possible affects of such selection. They also add controls for characteristics of banks in the countries (such as foreign-owned, state-owned, and asset concentration) to determine if said characteristics, as opposed to competition, would affect results. Lastly, the authors used measures of bank regulation and supervision to determine if these variable could somehow affect outreach. The authors find that the results of these tests vary little from the original findings.

Therefore, the authors believe that they have reasonably established a relationship between bank penetration and competition with MFIs. They associate such competition with a decrease in loan size, an increase in loans to women, and, weakly, with a decrease in return on assets.

By Christopher Maggio, Research Assistant

Similar Posts: