This doctoral thesis consists of five self-contained essayspresented to the Faculty Board of the Royal Institute ofTechnology. Property valuation is a central issue that forms acommon thread in the analysis in these essays. In the thesisproperty is considered in a mixed asset context in an attemptto build a bridge between valuation, property investment andfinancial theory. The object of the thesis is to value propertyfor finance, sales and purchases and investment. Theinvestigation of the thesis extends traditional valuation withan integrated approach using econometric technology.
Essay I estimates the market value of townhouses underdevelopment in a North American city. The traditional hedonicregression model is used to predict the value of the complex asa whole, as well as of the individual unit. The role of theproperty tax in explaining property valuation is indicated andthe valuation errors of the predicted values estimated in theessay relative to the prices realized in the market suggest thefeasibility of regression analysis for preconstructionappraisal.
Essay II investigates the implicit prices of property valuesin the Beijing residential market. An uncertainattribute"perceived construction risk" enters the modelas a proxy for a consumer's subjective probability ofconstruction quality. Public facilities are found to reduce thevalue of residences and consumers would be willing to pay ahighly substantial amount of money to protect themselves fromthe risk of poor construction quality.
Essay III studies the long-term relationship between housingprices and property stock prices under the Swedish rent controlsystem from 1980 to 1998. The Vector Autoregressive (VAR) modelwith a subsystem approach is used to test cointegration and theError Correction Model (ECM) and Granger Causality are alsotested. The tests provide evidence of co-movement between thehousing market and property stock market and suggest the roleof rentals in raising the speed of movement towards thelong-term equilibrium of asset prices.
Essay IV models the volatility of property stock returns inthe Swedish market from 1990 to 1999. The GeneralAutoregressive Conditional Heteroskedasticity (GARCH) model isapplied to capture time-varying volatility and the GARCH-Mmodel is used to price this volatility. The vacancy rate isfound to help explain persistent volatility and risk spillovereffects from the bond market and the direct real estate marketare expected.
Essay V analyzes the inflation hedging ability of Swedishproperty stocks from 1980 to 1999. Two expected inflationratesUND1x inflation and GARCH inflationandcointegration technology are used for this study. For theperiod as a whole, no inflation hedging behavior is found, butfor the period of 19861993 when the vacancy rate was low,short-run inflation hedging is indicated.
Institutionen för fastigheter och byggande , 2000. , 18 p.