A very long view on house prices

The Herengracht house index

One of my favourite pieces of academic research is a study of house prices along the Herengracht canal in Amsterdam made by real estate finance professor Piet Eichholz of Maastricht University. Thanks to outstanding dutch record-keeping,  Eichholz managed to construct the Herengracht house index covering house prices from the contruction of the Herengracht in the 1620s all the way to 1975 at first, and later extended to 2008.

Herengracht

The thing with Herengracht is that it never fell out of favour during all these years, it was always (and still is) part of Amsterdam’s most attractive real estate. It was a place where wealthy spice and slave merchants based themselves in the 17th century, and where big banks and consulting firms base themselves today. This makes the Herengracht index a great way of getting an insight in how inflation adjusted real estate prices change over time, and what one can realistically expect in the future.

So what does the Heregracht index show?

Timeline

1620s The Dutch Republic is in its first stage of its rise to global power. With money pouring in and the population doubling in 20 years, the city leaders finally breake an old taboo against building outside the medieval city walls and authorize one of the earliest and most far-reaching urban-planning schemes in history. The Herengracht and two other canals that arc around the city are built.

1628 – 1633 The economy soars. Herengracht house prices doubles. The index begins at 100 in 1628 and reaches 212.7 in 1632

1635 – 1636 Plague hits Amsterdam, kills about 20% of the city’s population.

1637 Tulip mania and subsequent crash. House prices almost drop back to the level of 1628 .

1640s – 1660s Dutch golden age continues. House prices surpass previous highs and reach 314.3 in 1664.

1672 France and England declare war on the Dutch Republic. England hinders Dutch trade on the sea. French king Louis XIV invades over land.  Herengracht house prices drop 58% from 268.6 in 1670 to 151.8 in 1678.

1701 – 1714 The war of Spanish Succession devastates the Netherlands. Herengracht house prices drop 16% to 169.5 in 1714.

1720s – 1730s Recovery. Herengracht index reaches a peak of 345.9 in 1736.

1791 United East India Company goes out of business. Herengracht index at 221.9 reflecting an 18th century post golden age decline, and Amsterdam becoming a second tier European city.

1794 France lead by Napoleon  invades Holland and captures Amsterdam. Herengracht index drops 74% from 1792 to 1814 where it hits a bottom of 68.1, the lowest value during the entire 380 year period.

1815 Napoleonic wars end. The Dutch reclaim Amsterdam and inaugurate a new monarch, William of Orange. Index begins rise to a peak of 131.4 in 1824.

1839 Amsterdam’s first railway was opens. Index at 125.7

1840s -1890s Economy soars with the industrial revolution. Herengracht index reaches peak of 267.2 in 1886.

1914 – 1918 World War I. Netherlands stays non-aligned, but Herengracht index reaches a war time low of 182.2 in 1916.

1920s Roaring 20s. House prices recover fast after the war reaching 231.7 in 1920 and climbing to 248.8 in 1928.

1930s Worldwide depression. Herengracht index down to 146.6 in 1938.

1940 – 1945 Hitler’s armies reach Amsterdam the 10th of May 1940. Occupation lasts until the Netherlands is liberated by allied forces in 1945. House prices stay at depression lows during the occupation.

1960s Post war boom with index peaks of 245.4 in 1966 and 263.3 in 1970.

1979-1984 Oil crisis Economic downturn. Herengracht index goes below 150.

2000s Global property boom. In 2007 Dutch newspaper Handelsblad reported:

The average house on the Herengracht now costs 2.6 million euros. That is, on an inflation-adjusted basis, just a bit less than in 1736, when house prices along the Herengracht were at their historical high. If house prices keep rising at their current tempo, the 271-year-old record will be tied in 2008.

2008-2010 Property crash. I haven’t seen any Herengracht index data beyond 2008, but I’d guess prices have gone down below 200 again.

Herengracht index in dutch newspaper. Euros on y-axis

Lessons from 380 year of house price data

As the text on the last graph says, over a 380 year period the real (ie inflation adjusted) house prices have only doubled, which corresponds to an annual average price increase of something like 0.1%.  And that’s with a starting point just at the start of the Dutch golden age. With a different starting point one could just as well get a 0% or even negative annual price change.

So the conclusion is that  there are ups and downs, but that over time prices roughly follow inflation. To expect house prices to rise much faster than inflation every year over a 10 or 20 year period without reverting down again does not make much sense. Guess it’s such expectations that create subprime loan debacles or Irish ghost estates pictured below:

All of this doesn’t mean that real estate is a bad investment over time. A buy to let investor that bought a house on the Herengracht in 1628 would have bought an asset that followed inflation minus costs of maintenance and upgrades, plus getting rental returns of maybe 4% – 8% per year, which well matches good performing stock market returns over long time periods.

Implications for real estate investments in the Ivory Coast

Now there are cases where areas jump up in price and then continue to follow inflation from a higher point. One example would be a transformation from working class low rent area to a sophisticated enclave such as the 20th century transformation of Södermalm in Stockholm, Sweden or Greenwich Village in the US. These types of transformations appear to occur when overall living standards increase and cities expand.

I’d say Herengracht is analogous to the Plateau area in Abidjan. It is attractive and will stay attractive, and prices are likely to have ups and downs along with Ivory Coast’s fortunes, but follow inflation over time.

So the areas to aim for would be those that have a chance of transforming themselves in, say, the next say 30 years. (Unfortunately, human life spans aren’t 380 years long.)  In Abidjan, which is a fast growing city with potential for good economic growth, there should be many areas with a chance for a positive transformation. My guess would be at centrally located areas that are below middle class but close to already established wealthy areas. So many parts of Cocody, around Zone 4, and maybe the best parts of Yopougon.

Milwaukee is the exception, not Abidjan

In an excellent article about the Herengracht index from 2006 in the New York Times Magazine, there is a discussion about Amsterdam being a turbulent place and whether it’s representative:

On the face of it, Amsterdam doesn’t seem the most apt model to apply to other cities, particularly American cities. For one thing, the Dutch have a turbulent history. The Low Countries were invaded or occupied many times by foreign powers, with subsequent collapses in the housing market. Also, notes Shiller, “Amsterdam — with tulip mania and the birth of the stock market — is the home of speculation. If you look at, say, Milwaukee, you don’t see the same degree of volatility. In a place like that, price is driven by land availability and construction costs, which is the tradition.”

But then, volatility is more prevalent in world history than stability, which, as far as Eichholtz is concerned, makes the Herengracht data more, rather than less, widely applicable. “The financial literature has been dominated by America,” he said. “And most models are created using post-1950 U.S. data, which give a biased picture of reality. There has been no other country like America, and there has been no other period like that in terms of stability. So I would say that in global terms, Milwaukee is the exception, not Amsterdam.”

From the perspective of Abidjan, it’s easy to agree with the conclusion. The post independence period in the Ivory Coast with booms, busts, coups, civil war, foreign military interventions feels a lot more like historical Amsterdam than Milwaukee.