Computing impulse response functions from a copula-based vector autoregressive model: evidence from the italian agri-food value chain

We investigate the presence of asymmetric vertical price contagions in the Italian pork market, describing the multiple dependence structure along the supply chain and evaluating the degree of extreme value dependence at different market levels. We provide a nonlinear copula-based vector autoregressive model built under a three-stage estimation method. Our methodological approach is all-encompassing and flexible. We contribute to the literature by providing a specification of the copula-based impulse response functions, which yields an interpretation of the economic implications of the empirical analysis. From farm to retail channels, our application to the Italian pork value chain reveals asymmetric price transmission. Our findings suggest large price adjustments at the retail level to shocks at other market levels, but significantly less adjustment at the farm level. Across the marketing chain, we show that the dynamics of pricing changes to shocks vary significantly. While our application to Italy serves as a useful illustration, our analysis may be expanded in a number of ways.