An Economic Analysis of Cotton Price Forecasting Using ANN in Andhra Pradesh, India

Abstract

Cotton is essentially produced for its fibre, which is universally used as a textile raw material. Cotton is an important commodity in the world economy. A remunerative price environment for the growers is very important for increasing production. In this context the study on area, production, export, import, supply and demand and their compound growth rates as well as influence on prices of cotton were analyzed using descriptive statistical tools and Artificial Neural Network model (ANN). The results showed that, compound growth rate of exports was negative and significant with -2.41 per cent whereas, imports showed a positive and significant growth rate with10.44per cent from 2006-07 to 2021-22. The seasonal indices of cotton arrivals in Andhra Pradesh were highest in the months of January (177.54), December (153.67) and November (146.10) because of holding of previous seasons crop by traders and farmers in anticipation of higher prices. The return on rupee investment was 0.596 which is concerned to tenant farmers and return on variable costs was 0.848 which is mostly related to owner farmers. The lower seasonal indices for cotton prices were observed in the months of December (97.23) and November (101.50). The results of ANN model revealed that, neural network 9-29-1(9 input nodes, 29 hidden nodes, and 1 output) outperformed all other neural networks with lower MAPE (2.904), RMSE (140.59), MAE (90.02), and MASE (0.114) values. It was expected that demand will persist in 2022-23 harvesting season also with a price around Rs. 8269/ q.