How to unlock agricultural loans and serve rural populations

The starting point of my latest research adventure was our last week’s credit committee. On the agenda: a loan to a new MFI client in El Salvador, focussing on lending to rural households. After summarizing the major opportunities and principal risks of this institution, my colleague justified the above what is usual interest rate spread for this credit by pointing out the MFI’s high portfolio concentration in agriculture. After the loan proposal was approved, I approached him, asking for an explanation of the risks inherent to agricultural loans. Essentially, this kind of loan is exposed to significant covariant risk, that is to say the probabilities of payment defaults are highly correlated across the agricultural sector. This lack of diversification combined with adverse weather conditions can eventually compromise the economic survival of the MFI: farmers facing crop losses due to supernatural forces (e.g. El Niño) are unable to pay back their loans, the MFI accumulates more and more defaults and unless it has subsidized alternative financing sources, it risks bankruptcy. In short, MFIs seeking profitability do not have an incentive to work in the agriculture sector due to the increased risks versus non-agriculture loans. This is a quite alarming because the poorest of the poor tend to live in rural areas.

But not only the supply side is hesitant to enter the market, even “farmers themselves limit their relationship with formal intermediaries because of the high cost of credit, relatively low profitability of agriculture, and fear that an unexpected problem will prevent them from repaying their loan” (see BASIS Brief # 46). Until today, for lack of developed credit markets, poor people exposed to hazardous weather rely mostly on informal risk management tools, including mutual aid among family members, neighbours and within social networks. But since these mechanisms fail in cases of regional natural disasters, farmers are likely to choose low-risk, low-return activities, hence reducing their chances to get out of the poverty trap. And with the prospect of climate change and more frequent extreme weather conditions, the situation will likely not improve, making farmers even more risk-averse.

In light of these circumstances, recent research in the microfinance sector has focused on microinsurance, considered as a promising tool to mitigate revenue and expenditure risk. After microcredit in the 80s, and microsavings in the 90s, microinsurance seems to be the new panacea to aid people, but the actual revolution will be a long time in coming. In 2006, only 3% of the world’s poor were policyholders, and the main product was life insurance which doesn’t really help people in smoothing out their short frequency consumption patterns. Microinsurance also faces several challenges, including high transaction costs and information asymmetries.

These are in particular the primary obstacles to crop insurance: in-the-field assessments of crop damages are costly (high transaction costs); insurance companies automatically attract “bad” risks, meaning exclusively farmers exposed to high risks are willing to sign an insurance policy (adverse selection); and insured clients have no incentive to mitigate potential risks by their own as they are covered by the policy (moral hazard). How can this business be profitable? Researchers are struggling to find new methods, enabling farmers to buy insurance policies at reasonable prices and offering insurers a profitable market at the same time.

An innovative approach deals with index-based insurance: instead of measuring crop damages directly, pay-outs are linked to a transparent index, optimally, one that is highly correlated to the farmer’s income. Different indexes have been proposed, ranging from rainfall to world market prices, to average regional yields. Since these are exogenous factors, beyond the farmer’s sphere of influence, insurers do not have to be afraid of asymmetric information problems. And transaction costs can be minimized. Today, several pilot projects have been launched and if the main challenges can be overcome, the development of an entire new market seems inevitable.

Market participants, including commercial funders like BlueOrchard , could benefit from this market development in several ways: in order to alleviate risks in agricultural loans, MFIs could link their credits’ interest rates to the amount of rainfall measured via weather stations. As such, farmers would be less reluctant to take loans (thereby getting the chance to pull themselves out of poverty) and demand will rise. At the same time, since weather risks sometimes occur across a whole region, MFIs with a limited regional outreach need to reinsure against these natural disasters, thereby transferring the risk to a reinsurance company holding a more diversified portfolio. Faced with MFIs with less risky portfolios, commercial funders would be keener to invest in these institutions.

Once theoretical and practical questions on how to design insurance products have been solved, there remains one key to the commercialization of microinsurance: ‘massification’. Since margins per unit are thin at the bottom of the pyramid and institutions operating only regionally are exposed to the dangers of covariant risks, an efficient network is essential to scale up microinsurance and make it profitable. The chain from the poor farmer in El Salvador to the international operating reinsurance company is made out of many links, and somebody has to join them. Who will assume the role of such a broker?