Asset value, rent, and expected capital appreciation of real estate market within a metropolitan area

Much of research on real estate markets has focused either on space market or capital market. In fact, these two markets are highly interrelated: fundamental connections exist between the space market which determines rent and the capital market which determines asset price. However, there has been little research about this linkage, especially within a single metropolitan area. The purpose of this research is to identify the relationship and to explain the spatial rent and price variation within a metropolitan area, with using observable rent data unlike most previous studies that suffer from limitation of using rent proxies. This research consists of two essays.; Essay One explores the existence of the linkage, and the variation of rents and asset prices. The essay recognizes the distinction between use decisions made in the space market and investment decisions made in the capital market. This essay clarifies which factors affect rents and asset prices separately, and estimates the rents and prices simultaneously. The two-stage least squares method is applied to avoid a simultaneous-equations bias problem. The empirical results strongly support the hypothesis for the existence of the linkage. The results also show that rents are substantially determined by use decisions, whereas asset prices are determined by investment decisions. These results suggest that an understanding of underlying market structure contributes to conclusive findings.; Essay Two focuses on the implication of the relationship. Some studies suggest a forward-looking hypothesis that the relationship has predictive power for future capital appreciation. Empirical results, however, offer little evidence for this suggestion. This essay infers that the inefficiency findings might be attributed to inaccurate definition of expected capital appreciation, and argues that capital appreciation should be defined as rent growth based on the present value model for estimating model to be dynamically stable. Essay Two applies a random coefficient model to testing the hypothesis, recognizing that the parameters in the estimating model are not fixed, but rather vary, because the data used in this analysis have a hierarchical structure. Essay Two offers the strong support for the forward-looking hypothesis by empirical results, if a model is correctly specified.