Studies of equilibrium conditions in housing markets

This dissertation studies the market equilibrium conditions for housing markets from three related perspectives, macro inter-city, macro intro-city, and micro intra-city. At macro inter-city level, Dividend Pricing Hypothesis requires that for arbitrage equilibrium to hold, rental returns and appreciation returns in different cities have to trade off each other. At macro intra-city level, Standard Urban Model predicts that for areas with decreasing transport cost, increasing income, and decreasing densities, appreciation returns should be higher in suburbs than in center cities. Since arbitrage equilibrium requires that total returns are the same for suburbs and center cities, rental returns have to be higher in center cities than in suburbs. At micro intra-city level, a textbook assumption is that rental returns are the same for all housing units within a single housing market regardless the size of the unit. A model with arbitrage and segmentation predicts that rental returns are constant at moderate segment of the housing market. At wealthy segment of the housing market, however, rental returns seem to decrease with the value of the housing unit.;Tests with carefully formulated panel data show considerable arbitrage opportunities between cities. Investment allocation within cities seems to be explained well by Dividend Pricing Hypothesis and fits the arbitrage equilibrium at macro level. Significant market segmentation is confirmed by a micro data set collected for Washington, D.C. metropolitan area.