about
A model of the Dutch housing market. Built for owners.
House prices in Dutch cities move with interest rates, incomes, new building and policy. This model lets you set those conditions for a postcode area and shows the range of values they imply over ten years. It does not predict a number.
Most property forecasts give you a number. That number is usually wrong.
Open a property report and you get a number. Your flat will be worth Y in five years. The number looks certain. It is rarely right, and the error bars are never shown.
Property does not move in smooth trends. It moves in cycles, driven by decisions that feed on each other. When mortgage rates fall, buyers can borrow more, prices rise and more buyers rush in before they are priced out. When landlords face tighter rules they sell, and homes move from renting to owning. A forecast that only extends last year's growth misses all of this.
This model does not capture those cycles either, and it says so. It does something simpler. You set the conditions with six sliders: the ECB rate, income growth, supply, gentrification, policy risk and rental scarcity. The model turns them, together with CBS figures for the postcode area, into one annual growth rate and applies that rate for ten years. The result is a smooth path with a widening band around it, not a cycle. What you can see is how much each condition moves the outcome: what a rate cut adds over ten years, or how much extra supply takes off.
The model returns scenario ranges, not point forecasts. Anyone giving you a single number for a ten-year property value is selling something.
Five feedback loops, named but not yet modelled.
The thinking behind loelens comes from the study of feedback in housing markets. Five loops matter most in Dutch cities. Uitponden, landlords selling rented homes one by one as tenants leave. The construction freeze. Rental buffer collapse. Credit amplification. Herding.
None of these loops runs inside the trajectory today. In a real loop, this year's prices change next year's conditions. In the model, the growth rate is fixed once you set the sliders and does not respond to the prices it produces. Some sliders stand in for the conditions a loop would create: policy risk for uitponden, supply for the construction freeze, rental scarcity for the rental buffer and the ECB rate for credit. The model page marks which loops your settings point to, as a reading aid. Building the loops into the trajectory itself is planned.
Three things this model does.
| Scenario ranges, not single forecasts | Four presets anchor the spread: Baseline, Boom, Crisis and Policy. The chart shows the full envelope and you read the range, not a line. |
| Per-neighbourhood data | CBS figures per postcode area set the starting value and adjust the growth rate: the average WOZ value sets the start, the share of owner-occupied homes weights the credit term and a large share of new build adds a little. The profile also shows income, the share of residents on benefits, population and the split of rented homes. Those do not change the growth rate. Differences between postcode areas are small because the data did not support larger ones. How this was tested |
| Transparent about sources, assumptions and limits | Every data feed is named with its licence, and every number in the model is marked as either taken from data or chosen. The Sources and method page lists both. Where data does not exist at postcode level, the model uses city-level figures and says so. Where an event sits outside the normal range, the model does not extrapolate. It tells you the range it is built for, and where it stops. |
Built by an Amsterdam property owner.
The first version was a spreadsheet with three sliders, built to answer a personal question about a single Amsterdam apartment. The current version covers Dutch cities at PC4 level, draws on seven open data sources updated between weekly and annually and returns four scenario ranges rather than a single line.
Everything on this site is open about its sources and limits. The aim is not to predict your property's price. The aim is to give you a structured way to think about the conditions under which the price could move, and by how much.