Working Paper: NBER ID: w31760
Authors: Vernica Bckerperal; Jonathon Hazell; Atif R Mian
Abstract: Every month, a fraction of UK property leases are extended for another 90 years or more. We build a new dataset of thousands of these natural experiments from 2000 onwards to estimate the expected long term housing yield, y*. Starting from a level of 5.3%, y* starts to fall during the Great Recession, reaching a low of 2.8% in 2023. Real time data shows y* has not risen since 2021, despite rising shorter term yields. Cross-sectional estimates show that y* is higher in areas with more housing risk, and falls by more in areas with more inelastic housing supply.
Keywords: Natural Rate of Return; Leasehold Extensions; Natural Experiments; UK Housing Market
JEL Codes: E4; E40; E5; G5
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| leasehold extensions (R21) | natural rate of return on capital (rk) (D33) |
| price changes of leasehold properties before and after lease extensions (K25) | natural rate of return on capital (rk) (D33) |
| price before extension (D49) | short-run value (D46) |
| price after extension (D49) | short and long-run values (D46) |
| natural rate of return on capital (rk) remains unaffected by short-run shocks (G31) | rk (Y60) |
| absence of pre-trends in prices (D41) | parallel trends assumption (C32) |
| balance in hedonic characteristics between treatment and control groups (C90) | parallel trends assumption (C32) |
| similar evolution of market rents (R31) | parallel trends assumption (C32) |
| substantial decline in rk from approximately 4.8% to 2.3% (R11) | long-term downward trend (E32) |
| robust and precise estimates of rk (C13) | standard errors significantly lower than traditional time series estimates (C22) |