Working Paper: CEPR ID: DP15624
Authors: Ricardo Reis
Abstract: Expected long-run inflation is sometimes inferred using market prices, other times using surveys. The discrepancy between the two measures has large business-cycle fluctuations, is systematically correlated with monetary policies, and is mostly driven by disagreement, both between households and traders, and between different traders. A parsimonious model that captures both the dispersed expectations in surveys, and the trading of inflation risk in financial markets, can fit the data, and it provides estimates of the underlying expected inflation anchor. Applied to US data, the estimates suggest that inflation became gradually, but steadily, unanchored from 2014 onwards. The model detects this from the fall in cross-person expectations skewness, first across traders, then across people. In general equilibrium, when inflation and the discrepancy are jointly determined, monetary policy faces a trade-off in how strongly to respond to the discrepancy.
Keywords: No keywords provided
JEL Codes: E31; E52; D84
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| monetary policy actions (E52) | discrepancy between market prices and survey measures of inflation expectations (E31) |
| tighter monetary policy (E52) | discrepancy (D80) |
| discrepancy (D80) | compensation for inflation risk (E31) |
| discrepancy (D80) | disagreement between public expectations and market traders (D84) |
| discrepancy (D80) | disagreement between marginal and average traders (F12) |
| changes in trader expectations (D84) | unanchored inflation expectations since 2014 (E31) |
| market shocks (G17) | unanchored inflation expectations since 2014 (E31) |