Spatial Modelling of the Food Price Dynamics

The paper explores the impact of business indicators of the Bank of Russia’s monitoring on the monthly food prices growth rates, considering spatial effects. As a baseline, we used an adaptive Phillips curve with monitoring indicators as a gauge of economic activity (business climate indicator of the food industry or price expectations of the industry). Their advantages over traditional ones (e.g., the output gap): timeliness, high frequency, consideration of latent factors, independence on methodology, absence of revision risks. Calculations using three special cases of the spatiotemporal autoregressive Durbin model show a significant impact of food industry’s price expectations and their spatial lag on the food consumer price index (CPI) next month. A similar effect is confirmed for the spatial lag of CPI inertia. The model with both spatial lags of CPI (W?t) and CPI inertia (W?t – 1) didn’t yield satisfactory results. We proposed an asymmetric weights matrix that takes into account along with distance also direction of spatial interregional influence, allowing for more robust estimates. This indirectly confirms the hypothesis that the main factor of spatial correlation of food CPI is interregional transportation whose impact may manifest itself with a time lag. The study substantiates the potential of spatial lags for other models using the business monitoring data. The obtained results can be used to refine short-term forecast of regional CPI and, ultimately, to improve the quality of data for Bank of Russia’s monetary policy