Forecasting CPI inflation components with Hierarchical Recurrent Neural Networks

We present a hierarchical architecture based on recurrent neural networks for predicting disaggregated inflation components of the Consumer Price Index (CPI). While majority existing research is focused headline inflation, many economic and financial institutions are interested in its partial components. To this end, we developed novel Hierarchical Recurrent Neural Network (HRNN) model, which utilizes information from higher levels CPI hierarchy to improve predictions at more volatile lower levels. Based large dataset US CPI-U index, our evaluations indicate that HRNN model significantly outperforms vast array well-known prediction baselines. Our methodology results provide additional forecasting measures possibilities policy market makers sectoral component-specific price changes.