Constructing an Early Warning Indicator under Seasonality and Asymmetry in Agricultural Product Price Distribution: The Case of Onion

The purpose of this paper was to construct an early warning system for major Korean agricultural products given a clear understanding of the patterns of agricultural product prices. An introduction of an early warning system in Korean agriculture is expected to help producers, consumers and policy makers as well in that it will increase the efficiency of a supply-demand stabilization policy of agricultural product prices. The literature on early warning system usually focuses on the mean of probability distribution and its deviation from the mean in order to identify critical points for constructing relevant stages of early warning: normal, attention, care, warning, and seriousness. In contrast, this paper utilizes the “median” of probability distribution and associated probability density in constructing such a stage. This turns out to be quite appropriate given the asymmetric nature of probability distribution of agricultural product prices. For a case study, this paper uses daily price data for onion for identifying critical points. Specifically, daily prices of 5 recent years are considered. The analysis is done in a monthly base to incorporate seasonality issues in green onion prices.