The agricultural season naturally induces seasonal variation to prices for things like maize. Is it possible to make money by timing this market?

Download

Abstract

Consumption expenditures in rural areas of low-income countries are highly variable across seasons, yet the literature still lacks a standard framework for asking whether seasonal poverty reflects local credit market failure or poor integration with the broader economy. We develop an intertemporal model of farmers’ portfolio choices under seasonal price risk and borrowing constraints, and derive a sign diagnostic: The amount a farmer is willing to pay for a small risk-free bond (call this price q) must rise in response to a positive income shock if credit constraints bind, but fall if precautionary motives dominate. We apply this framework to a randomized post-harvest loan program in Gombe, Nigeria, and supplement the experiment by also collecting high-frequency data on prices, stocks, and expenditures. The loan sharply reduces the marginal utility of expenditure around delivery, but q never rises over the full follow-up. Precautionary savings, not credit constraints, govern the intertemporal allocation. Receipt of the loan leads to a portfolio rebalancing, as farmers adjust their grain stores, and increase investment. But maize prices increase little after harvest and season-average consumption expenditure effects are small, though the MUE—a more sensitive welfare measure—detects a large and significant improvement around delivery that expenditures miss. We fail to reject the null of well-functioning local financial markets. The binding constraint is poor spatial integration rather than inefficient local allocation—promoting market integration may improve lean-season welfare more than would the local provision of credit.

BibTeX

@Unpublished{ligon-silver26,
  author          = {Ethan Ligon and Jedidiah Silver},
  title           = {Credit and Welfare Across the Lean Season},
  year = 2026,
  url          = {https://escholarship.org/uc/item/81k9j2q0}}