Structural analysis of potato market behavior using neural network modelling in Peru

The complex behavior of crop markets is always difficult to characterize, especially under structural market failures that cause increasingly unstable and unprofitable prices. In this paper we propose an innovative analytical tool to identify potential potato market failures related to uncoordinated decisions in production which in time causes negative effects on farmers' wellbeing and persistently high rates of poverty in potato production areas. Based on a database including geographical origin, volume and price for potato production during the period 1997–2021, this work generates a market price prediction neural network model, using it to identify coordination problems in the functioning of the market and to test an alternative micro-scenario for a critical period of high volatility and price crisis. Using AI modelling and expert knowledge allows a better understanding of market coordination problems, to design more effective strategies and policy interventions towards reduction of poverty in potato producing rural areas in Peru.