Working Paper: CEPR ID: DP16381
Authors: Mikhail Chernov; Lars Lochstoer; Dongho Song
Abstract: We present evidence that the mix of transitory and permanent shocks to consumption is changing over time. We identify three regimes: two highly persistent regimes where either permanent or transitory shocks are relatively more dominant, and a disaster regime that is largely transitory. We study implications of this finding for asset prices. The transition from the second to the first regime in the mid-1990s makes the correlation between equities and bonds switch sign from positive to negative as in the data. The real bond and equity yield curves are approximately flat. The nominal bond curve is upward sloping. These results are achieved without relying on the nominal channel too much. That is, as in the data, the variation of inflation in the model is under 40% as a fraction of variation in nominal yields.
Keywords: Permanent and transitory components of consumption; Bond-stock comovement; Bond yield curve; Equity yield curve
JEL Codes: E21; E31; E43; E44; G12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| changing mix of permanent and transitory shocks to consumption (D15) | different correlations between stock and bond returns (G12) |
| permanent component dominates (D15) | positive covariance between realized and expected consumption growth (F62) |
| transitory component predominates (E32) | negative covariance between realized and expected consumption growth (E21) |
| negative shocks to consumption in the permanent regime (E21) | lower-than-average growth (O49) |
| negative shocks to consumption in the transitory regime (E21) | higher-than-average growth (O49) |
| disaster regimes (H84) | transitions from risky to hedging consumption shocks (D15) |
| transitions from risky to hedging consumption shocks (D15) | affecting bond pricing and correlation between consumption claims and real bond returns (G19) |
| model can generate signswitching correlations between consumption claims and asset returns (C10) | consistent with observed data (Y10) |