#REAL ESTATE
Real estate hasn't beaten the stock market. It tied — with half the swings.
Published on August 7, 2026
Òscar Jordà (San Francisco Fed), Katharina Knoll, Dmitry Kuvshinov, Moritz Schularick and Alan M. Taylor reconstructed total real returns for four asset classes — treasury bills, bonds, stocks, residential real estate — across 16 advanced economies, from 1870 to 2015. That's 1,739 annual observations. Result: 7.05% a year for real estate, 6.89% for stocks, 2.50% for bonds, 0.98% for treasury bills. Published in the Quarterly Journal of Economics in 2019.
What about risk, in all this?
The annual standard deviation of real returns comes out at 9.98% for real estate, versus 21.94% for stocks. More than half the volatility, for an equivalent average return.
In other words: over a century and a half, housing returned about as much as stocks while causing half the fright. That's exactly what classic financial theory says shouldn't happen: more risk should pay more.
Three caveats, because they matter
- These returns are gross. Neither notary fees, nor renovation costs, nor vacancy, nor property tax are deducted. On real estate, these items weigh heavily.
- After 1950, the ranking flips: stocks return 8.28% a year in real terms, real estate 7.44%. Housing's lead comes mostly from before World War II.
- The indices are national and aggregated. An individual holds two or three properties in one city, not a whole country's real estate: diversification is far harder than with an ETF.
Why it still matters: the “bricks versus paper” debate is framed wrong. Over the long run, the two have paid roughly the same. What separates them is volatility, liquidity, taxation and leverage — not the gross return.
For you
- When you compare your property to a World ETF, are you comparing net or gross returns?
- How many of your real estate decisions rest on the return, and how many on not seeing the price every day?
Source: Jordà, Knoll, Kuvshinov, Schularick & Taylor, “The Rate of Return on Everything, 1870-2015”, Quarterly Journal of Economics, vol. 134, no. 3, 2019, table 3.
View the source →