The price information is one of the prime objectives of marketing strategies, and the farmers are unable to determine the marketing strategies without knowing the price movements of the agricultural commodities. In this context, the study has examined the spatial price integration among four major onion markets using the threshold vector error correction model (TVECM) that takes into account transaction costs in the price adjustment process. Augmented Dickey–Fuller and Phillips–Perron tests for unit root suggest that the time series is I(1). The application of the Johansen cointegration technique supports the presence of long-run price association and equilibrium in all pairs of onion markets. The Granger Causality test unveils that Bengaluru Granger causes all the markets except Kolkata. The Hansen and Seo supreme Lagrange Multiplier ( SupLM) test of linearity suggests that non-linear TVECM with one threshold and two regimes is best fit for the underlying data for three pairs of markets. While the rest of the three pairs, the SupLM test rejects the null of linearity, therefore, linear vector error correction model (VECM) is estimated. Finally, VECM and TVECM results reveal that Mumbai and Bengaluru are dominant markets in price formation in rest of the markets. Against these findings, it is suggested that the prices should be stabilised in the dominant markets so that the price shocks are not transferred to other markets. The threshold parameter, which is analogous to transaction cost, reveals the high transaction costs between the selected markets pairs, especially Mumbai and Delhi. One of the reasons for the high transaction costs may be the inefficiencies in infrastructure and communication. While a more correct explanation for this difference can be attributed to the differences in marketing fees, taxes, commission charges, license fees, etc., across the spatially separated agricultural markets. JEL Codes: Q1, Q13, D4, C22, D23
Read full abstract