Empirical Application of the Housing Market No-Arbitrage Condition: Problems, Solutions and a Finnish Case Study

Abstract

The often used housing price-to-income and housing price-to-rent ratios are problematic in housing market analysis and may result in misleading conclusions. Instead, the no-arbitrage condition of housing market is a theoretically sound basis to evaluate if housing prices are misaligned. Unfortunately, empirical application of the no-arbitrage condition has notable complications. This article reviews these complications and suggests some solutions to them. The use of implied expected appreciation derived from the no-arbitrage condition is recommended. It is also claimed that the real appreciation is better to use than the nominal one in the no-arbitrage computations. Furthermore, the paper shows that the maintenance costs as a fraction of housing price vary substantially in time and location, which may significantly affect the equilibrium housing price level relative to rental prices. An empirical application of the no-arbitrage relation using data from ten Finnish cities shows that housing price level in 2007 was not based on high expected appreciation. This lowers the fears for a price bubble.

Keywords: Housing, no-arbitrage condition, overpricing, user cost

JEL Classification: R21