Estimating the Competitive Storage Model with Stochastic Trends in Commodity Prices

We propose a State-Space Model (SSM) for commodity prices that combines the competitive storage model with stochastic trend. This approach fits into economic rationality of decisions and adds to previous deterministic trend specifications model. For Bayesian posterior analysis SSM, which is nonlinear in latent states, we used Markov chain Monte Carlo algorithm based on particle marginal Metropolis–Hastings approach. An empirical application four markets showed SSM favored over specifications. The identifies structural parameters differ from those In particular, estimated price elasticities demand are typically larger under SSM.