Residential real estate, not equity, has been the best long-run investment over the course of modern history.
–“The Rate of Return on Everything, 1870–2015” by Jorda, Knoll, Kuvshinov, Schularick, and Taylor
That’s one of several surprising findings in a new paper from the crew that’s been studying historical economic data on everything from the growth of mortgage lending to the relationship between loose monetary policy and financial excess.
Using a new set of data they constructed by hand from news articles, advertisements, statistical abstracts, and more, they estimated the investment performance of deposits, government bonds, residential real estate, and stocks across 16 rich countries. (Alphaville will look into some of their other big findings in subsequent posts.)
Housing beat stocks mostly because the returns were less than half as volatile. Thanks to the magic of compounding, this created a performance gap of more than 2 percentage points each year, on average, since 1950, with an even bigger gap if you start the clock in 1870:
This is surprising. The ratio between the average yearly return above the short-term risk free rate and the annual standard deviation of those returns — the Sharpe Ratio — should be roughly equivalent across asset classes over long stretches time. There might be short periods when an asset class’s Sharpe ratio looks unusually high, especially in individual countries, but things tend to revert to their long-term average sooner or later.
As you can calculate from the table above, rich-country government bonds have historically had a Sharpe ratio of about 0.2, while stocks have had a Sharpe ratio of around 0.3. A diversified portfolio of housing, by contrast, has had a Sharpe ratio greater than 0.7 since the late 1800s and greater than 0.8 since 1950!
It makes sense that bonds have a slightly lower Sharpe ratio than stocks because their value goes up when the economy goes down. By contrast, stocks and other risky assets often lose more than half their value right when people are most likely to lose their jobs and need the money. Savers should therefore be willing to pay more to own bonds than stocks, even after adjusting for simplistic measures of risk, which would explain the modest difference in Sharpe ratios.
There is no comparable explanation for why housing has historically been such a better investment than stocks. Yes, it’s much harder for the typical retail investors to construct a diversified portfolio of residential real estate than a broad basket of equities, but most wealth isn’t held directly by regular people.
The institutions working on behalf of regular people, and the hyper-rich, shouldn’t have much trouble building diversified portfolios of residential real estate if they really wanted to. It’s therefore tough to explain the chart below simply by appealing to differences in the degree of difficulty:
Spanish housing has had the worst risk/return ratio within the sample, but it still looks outstanding compared to the best of global equities:
One possible explanation is that houses are hard to sell when you really need the money, except at a steep discount. The volume of home sales tends to collapse during recessions and rise when the economy recovers, for example, presumably because many people who might want to sell are withholding supply because they don’t like the price they think they would get. Public price indices may not fully capture this volatility in the same way that stated returns from private equity funds understate the actual riskiness of their investments.
None of this means you should increase your exposure to housing in your personal portfolio.
Jorda et al provided a valuable service but they didn’t calculate the returns most homeowners actually experience. Most people borrow to buy housing and most people live in their properties without renting them out. This makes a big difference.
Housing returns are only partly determined by changes in house prices. Jorda et al created their total return index based on both capital gains and income from renting. Net rental income has historically accounted for half of the total returns from owning housing. It’s also far less volatile, dramatically boosting the Sharpe ratio compared to what you would get just by looking at changes in house prices.
Housing has beaten stocks since 1950 because rental income has been better than dividend income, not because house prices have grown more than stock prices:
The importance of rental income fits with data from earlier papers showing minimal increases in real house prices in Amsterdam and Beijing over long periods. American readers should note that capital gains historically explain less than a third of the total returns to US housing, with the rest coming from rent.
Then there is debt. If you put 10 or 20 per cent down to get a mortgage for a house that you live in without tenants, your absolute returns — and the volatility of those returns — will look very different than what Jorda and his colleagues measured. You will experience much larger gains in your net worth when house prices rise and much bigger drops when they fall. Instead of earning stable rental income to offset changes in house prices, you’re stuck paying interest you have to pay out of your wages or savings.
It’s possible to imagine a world where most housing is owned by large diversified investment trusts that anyone can invest in, but until then, “housing for the long run” is not a practical investment strategy.
Americans are making more money from renting out homes than ever before, sort of — FT Alphaville
Alan Taylor on financialisation, business cycles, and crises; and a mid-year review — Alphachat
Myths and facts about “risk parity” — FT Alphaville
No, stocks aren’t a good inflation hedge. Try bonds (really). — FT Alphaville
Easy money, housing bubbles, and financial crises — FT Alphaville
Fix housing finance, fix the economy? — FT Alphaville