Working Paper: CEPR ID: DP10419
Authors: Sumru G. Altug; Cem Cakmakli
Abstract: In this paper, we formulate a statistical model of inflation that combines data on survey expectations and the inflation target set by central banks.. Our model produces inflation forecasts that are aligned with survey expectations, thereby integrating the predictive power of the survey expectations together with the baseline model. We further incorporate the inflation target set by the monetary authority to examine the effectiveness of monetary policy in forming inflation expectations and therefore, predicting inflation accurately. Results indicate superior predictive power of the proposed framework compared to the model without survey expectations as well as several popular benchmarks such as the backward and forward looking Phillips curves and naive forecasting rule
Keywords: inflation forecasting; inflation targeting; state space models; survey-based expectation; term structure of inflation expectations
JEL Codes: C32; C51; E31; E37
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| survey expectations (C83) | inflation forecasting model (E37) |
| target inflation (E31) | inflation forecasting model (E37) |
| survey expectations + target inflation (E31) | predictive performance of the model (C52) |
| discrepancies between model-based inflation expectations and target inflation (E31) | insights into monetary policy effectiveness (E52) |
| inflation expectations closely aligned with target inflation (E31) | predictive performance of the model (C52) |
| inflation process in Turkey is more volatile than in Brazil (E31) | effectiveness of survey-based expectations in predicting inflation (D84) |