abstract: After a relative period of growth (2000-06), the U.S. economy experienced a sharp decline (2007-09) from which it is yet to recover. One of the primary factors that contributed to this decline was the sub-prime mortgage crisis, which triggered a significant increase in residential foreclosures and a slump in housing values nationwide. Most studies examining this crisis have explained the high rate of foreclosures by associating it with socio-economic characteristics of the people affected and their financial decisions with respect to home mortgages. Though these studies were successful in identifying the section of the population facing foreclosures, they were mostly silent about region-wide factors that contributed to the crisis. This resulted in the absence of studies that could identify indicators of resiliency and robustness in urban areas that are affected by economic perturbations but had different outcomes. This study addresses this shortcoming by incorporating three concepts. First, it situates the foreclosure crisis in the broader regional economy by considering the concept of regional economic resiliency. Second, it includes the concept of housing submarkets, capturing the role of housing market dynamics in contributing to market performance. Third, the notion of urban growth pattern is included in an urban sprawl index to examine whether factors related to sprawl could partly explain the variation in foreclosures. These, along with other important socio-economic and housing characteristics, are used in this study to better understand the variation in impacts of the current foreclosure crisis. This study is carried out for all urban counties in the U.S. between 2000 and 2009. The associations between foreclosure rates and different variables are established using spatial regression models. Based on these models, this dissertation argues that counties with higher degree of employment diversity, encouragement for small business enterprises, and with less dependence on housing related industries, experienced fewer foreclosures. In addition, this thesis concludes that the spatial location of foreclosed properties is a function of location of origination of sub-prime mortgages and not the spatial location of the properties per se. Also importantly, the study found that the counties with high number of dissimilar housing submarkets experienced more foreclosures. / Dissertation/Thesis / Ph.D. Environmental Design and Planning 2012

資產報酬的波動程度隱含風險與不確定性，不同的投資者存在不同之風險偏好與風險承受能力，因此瞭解報酬波動之特性有其必要性；又鑑於過去不動產市場對於商用與住宅不動產兩次市場之相關研究較欠缺。因此本研究擬分別探討商用與住宅不動產市場報酬波動特性與差異，並檢視其風險與報酬間之關係。此外，總體經濟環境之變動會影響不動產市場供需關係，進而影響其價格與報酬之波動，因此本研究最後再進一步討論影響其市場報酬之總體經濟因素。 為捕捉不動產市場報酬之波動特性，本研究擬透過GARCH模型分別檢驗商用與住宅不動產市場報酬波動特性與差異；進而應用GARCH-M模型，探討商用與住宅不動產市場報酬與風險之關係；最後透過落遲分配模型實證比較分析顯著影響商用與住宅不動產市場報酬之總體經濟因素。樣本取自台北地區，資料期間為1997年2月至2009年3月之月資料。 實證結果顯示，商用不動產市場中投資人較容易透過自身過去的報酬波動推測未來的波動，反觀住宅不動產市場部分，投資人則傾向注意整體市場消息的散佈，因為其較容易受到外在因素影響而導致報酬波動；由GARCH-M模型實證結果顯示，住宅與商用不動產市場報酬與風險間均呈現顯著正相關，顯示其市場波動風險增加時期，會提供更高之報酬以均衡投資者所面對之較高市場波動風險；由落遲分配模型實證結果顯示，商用與住宅不動產市場報酬確實和總經變數之間有著程度不同的關聯性，所有當期總經變數與不動產報酬間均不存在顯著影響關係，顯示各總經變數對不動產報酬的影響存在時間落差。此外，總經變數對商用報酬的影響程度相對大於對住宅報酬的影響，且總體經濟環境變動對於商用不動產市場報酬之衝擊相對較為迅速。 / This research plans to study the relative volatility characteristic of commercial and residential property returns. In addition, the changing real estate environment can be linked to the macro economy, so we further discusses the relationship between property returns and the macro economy. In order to catch the volatility characteristic of real estate returns, we use GARCH model to examine the volatile behavior of real estate returns of commercial and residential property in the Taipei area during the period of February 1997 to March 2009, and because risk is time-varying in the market, we continue to employ GARCH-M model to observe whether can explain the change in expected returns of commercial and residential property. Furthermore, we use distributed-lag model to explore the relationship between macroeconomic factors and real estate returns. The major findings of this article can be summarized as follows. First, it is easier for investors to infer the future fluctuation through oneself returns in the past in the commercial real estate market, but part on the residential real estate market, the volatility of residential property returns is influenced by external factor more easily. Second, our empirical applications in both commercial and residential real estate markets show that the risk is positively correlated with both property returns and high risk can bring high return. Third, there are different relations of intensity between real estate returns and macroeconomic factors and the impact of macroeconomic factors on real estate returns exist time-lag. In addition, macroeconomic factors’ impact on commercial returns is relatively great, and the environmental change takes place to the impact of the commercial property returns comparatively fast.