Updated Risk Premium Estimates — March 2026
Pipeline rewritten in Julia. CAY constructed from publicly available FRED data, removing the dependency on Lettau's website.
DOLS coefficients (estimation sample: 1951Q4–2019Q3, pre-COVID)
| Coefficient | |
|---|---|
| β_a (wealth) | 0.195 |
| β_y (income) | 0.863 |
CAY validated against Lettau's published series (correlation = 0.986).
Predictive regression (285 obs, 1947Q4–2021Q4)
| Coefficient | SE | |
|---|---|---|
| D/P ratio | 2.779*** | (1.021) |
| cay | 0.398 | (0.360) |
| T-bill | −1.232*** | (0.306) |
| R² | 0.234 |
Data sources
- CRSP MSI through 2024Q4 (from WRDS
crsp.msi) - T-bill (
TB3MS) through 2026Q1 (from FRED) - CAY components through 2025Q4 (from FRED)
Changes from v1.0 (paper version)
- Pipeline rewritten from R to Julia
- CAY constructed from FRED data instead of downloaded from Lettau's website
- DOLS estimated on pre-COVID sample (1951Q4–2019Q3) to avoid distortion from pandemic-era swings in transfers, consumption, and asset values
- CRSP data extended from 2017Q4 to 2024Q4
For the exact replication of the published paper, use v1.0.