Analysis of Agricultural Commodities Prices with New Bayesian Model Combination Schemes

摘要: In the described research three agricultural commodities (i.e., wheat, corn and soybean) spot prices were analyzed. particular, one-month ahead forecasts built with techniques like dynamic model averaging (DMA), median probability Bayesian averaging. The common features of these methods are time-varying parameters approach toward estimation regression coefficients dealing uncertainty. other words, starting multiple potentially important explanatory variables, various component linear models can be constructed. Then, from an averaged forecast Moreover, mentioned easily modified into a selection approach. Considering as benchmark models, all considered potential price drivers, historical average, ARIMA (Auto-Regressive Integrated Moving Average) naive Diebold–Mariano test suggested that DMA is interesting alternative model, if accuracy aim. Secondly, interpretation weights ascribed to containing given variable economic development emerging BRIC economies (Brazil, Russia, India China) recently one most drivers prices. analysis was made on monthly data between 1976 2016. initial fundamental, macroeconomic financial factors.