First, let me say I applaud Kiva’s addition of the five-star
Field Partner Risk Rating system. This system provides valuable and useful information for lenders to help them judge the risk of lending to businesses offered by the Field Partners. The rating system is comprehensive in that it evaluates a wide variety of financial and performance information about a partner’s lending practices and track record. It is consistent in that it communicates all ratings in terms of the “amount of evidence demonstrating high likelihood of repayment.”
However, the main weakness of this rating system, as I see it, is in its consistency. In effect, a low rating of “1” or “2” is communicated as being the result of
the absence of positive information. These low ratings are described as they are “often a result of the Field Partner being a very young organization or the lack of immediately available, low cost of third-party verification options.” Unfortunately, that is not always the case. For certain Field Partners, the lowest rating is the result of
the presence of negative information, specifically, very high delinquency rates.
The net effect of the current rating system is that where it needs to be the most sensitive – at the lowest ratings – it confounds qualitatively different circumstances. Both young, relatively untested partners, such as
FDM (Fundo De Desenvolvimento Mulher), and established partners with excessive delinquency rates, such as
Women Initiative to Eradicate Poverty receive the same rating of “1.” Consequently, new partners that easily can be given the benefit of the doubt are subjected to “guilt by association” by being placed in the same category as partners in jeopardy of being “deactivated.”
So, what is the solution to this bad apples and green oranges living together dilemma? I believe the first step is to acknowledge when negative information, such as high delinquency/default rates are present. I fully expect this information already is taken into account to determine the ratings. It just isn’t communicated by the current rating definitions. The second step would be to separate the “very little positive information” partners from the “substantive negative information” partners. The “simplest” way to do this would be to create a “six-star” system. Add a “0” star rating that means, “a significant amount of evidence demonstrating
low likelihood of repayment.”
Of course, more or fewer than six stars could be used, but the key to fixing the current “problem” with the risk rating system is to separate “little good information” partners from “too much bad information” partners. As long as the system stands as it is, those partners that need the most benefit of the doubt will be unduly penalized by association with those who have “earned” their cautionary lot.
