Agribusiness foresight tools: GARCH models in relation to the price conjuncture

This study examines a methodology for forecasting the volatility of price indices associated with agricultural production and consumption. First, we investigated the time series of the price index growth rates for the plant and animal husbandry sectors. In addition, the dynamics of the volatility of the consumer price index for food products was considered. The main goal of this study was to develop a set of generalized autoregressive conditional heteroscedasticity (GARCH) models based on the available data. For analysis, growth rates of indices were used, and information criteria, hyperparameters, and specifications of models were selected. A preliminary exploratory analysis of the time series was performed, and an intermediate analysis of mean values was performed using the SARIMAX algorithm. Formed models were used to develop a forecast for the next period, which includes twelve months of 2023. The volatility of production indices is much higher than that of consumer indices, which is probably related to the properties of the underlying data. According to the obtained data, particular attention should be paid to plant husbandry, which is characterized by increased volatility. This is because of the specific features of production (seasonality) and the characteristics of interaction with other industries and agribusiness, as well as consumer activity for this type of product. It is also predicted that there will be a small trend toward convergence in the dynamics of all indicators. An interesting area for further research is to study the influence of digital transformation in agriculture on indices. Digital transformation, which influences all business process chains, can become the main factor that mitigates external conjunctural fluctuations.