The housing market no-arbitrage relation states the asset market equilibrium relationship between housing prices and rents. However, the theory gives neither the adjustment speed towards the equilibrium relationship nor the reaction magnitudes of housing prices and rents to changes in the user cost for housing capital. This study investigates empirically the adjustment of housing prices towards the no-arbitrage relation using data for the period 1975-2010 for the Helsinki metropolitan area. The empirical analysis is based on a three-dimensional near-VAR model that is augmented by error-correction mechanisms in each of the three equations and that is estimated by Seemingly Unrelated Regression (SUR). In line with the theory, housing prices adjust significantly towards the housing market equilibrium. The adjustment is highly sluggish, however. In contrast with the common assumption that housing prices are sticky especially downwards, the empirical analysis implies that real housing prices adjust more rapidly towards the no-arbitrage relation when prices are above the relation. It is argued that a likely reason for the sticky upwards adjustment is the liquidity constraints faced by households. In line with this argument, the results indicate that liquidity constraints have a significant influence on the adjustment speed of housing prices, i.e., looser liquidity constraints appear to induce faster adjustment of housing prices towards the equilibrium. The estimated model is also used to explore the impacts of user cost changes on housing prices, rents, and supply.