Are Swedish House Prices Too High? Why the Price-to-Income Ratio Is a Misleading Indicator
According to ECB (2023) and European Systemic Risk Board (2022), Swedish owner-occupied housing (OOH) was overvalued by about 55% in 2021q2, the largest overvaluation in the EU and EEA; according to European Commission (2023c), by about 30% in 2022. These assessments affect warnings and recommendations issued for Swedish economic policy and the shocks in EBA stress tests of Swedish banks.
But these large overvaluation assessments are due to the use of misleading indicators: the deviations of price-to-income (PTI) and price-to-rent ratios from their historical averages. It is shown that according to the appropriate indicator, the user-cost-to-income (UCTI) ratio, Swedish owner-occupied houses have since 2010 instead become increasingly undervalued (not overvalued), by about 30% in 2019q4. Due to rising mortgage rates, they are less undervalued in 2023q2, but still about 20%.
For Sweden, the UCTI and PTI indicators are in fact strongly negatively correlated and have opposite signs. If the UCTI indicator is right, the PTI indicator is consistently wrong. Taking the average of the two indicators is not a good idea.
The PTI ratio disregards mortgage rates and other housing costs and lacks scientific support. According to a large housing literature, it is not the purchase price but the user cost that is the appropriate measure of the cost of living in OOH, the cost of the housing services that the OOH delivers. New improved estimates of the user cost are constructed, including an adjustment for a preference shift during the coronavirus crisis in favor of larger and better housing.
The valuation assessments of the ECB, the ESRB, the Commission, the OECD, and the IMF are scrutinized and compared. The problem of misleading indicators and overvaluation assessments—and resulting distorted warnings and recommendations—is not restricted to Sweden but concerns several other countries in the European Union.
I thank Ronald Albers, Johan Almenberg, Christophe André, Bo Becker, Robert Boije, Stephen Cecchetti, Alessandro Di Spirito, Peter Englund, Harry Flam, Lucia Fullin, Bernadette Galster, Isabel Grilo, Sten Hansen, Bengt Hansson, Jesper Hansson, Hana Hejlová, Matilda Kilström, Willem Kooi, John Muellbauer, Aurelio Nocera, Ida Peltonen, Tuomas Peltonen, Moreno Roma, Khaled Sakr, Antonio Sánchez Serrano, Rickard Sandberg, Frauke Skudelny, Agnieszka Szczypinska, Alessandro Turrini, Roine Vestman, Stefan Zeugner, and participants in seminars at Norges Bank, the Stockholm School of Economics, the ECB, the CEBRA, and the Bank of Italy for helpful recent or previous comments, discussions, and data. I am grateful to Vitor Martins and Stefan Zeugner for updates of the European Commission’s HouseLev database; to Alexander Leodolter and Klaus Grünberger for information and programs regarding the Commission’s Housing Taxation Database; to Bengt Hansson and Peter Englund for sharing data and calculations from their work; and to Peter Buvén, Tommy Lindkvist, Malin Sundberg, Maria Wiberg, and Michael Wolf, all of Statistics Sweden, for data and answering questions. Support from the Jan Wallander and Tom Hedelius research foundation and the Tore Browaldh research foundation is gratefully acknowledged. Any errors are my own. The views expressed herein are those of the author and do not necessarily reflect the views of the National Bureau of Economic Research.