Conference Agenda

It is well known that the large scale project of transportation improvement will significantly appreciate the housing price, especially in the neighborhoods nearby the stations. However, the housing price affected could variate according to the differences of urban construction environment, the city center area and rural area could be different, for example. In this study, we collected the housing price data within 2km centered at every train station of the Taichung Railway Elevation Project (TREP), which has been planned and constructed for more than 10 years and finished in 2020. A total number of 5,430 sample size of effective cases of housing transaction were selected from the Taiwan government open data sources. The hedonic housing price regression models with quadratic effect coefficients and the cubed effect coefficients were constructed and statistically tested. Important variables include housing and building attributes, urban land use regulations, train station accessibility, and annual riding (in and out) flow of train stations. Our empirical evidences show that, on the overall average of cases within 1.4km nearby train station, a 1% increase of distance away from the train station results in a 0.1% decrease of housing price. The higher category level of train station with higher riding flow will have a bigger significant effects on the housing price, as rational expected. And, when look into the housing price curve (the rent gradient) with respect to the distance from train station, the price curve illustrates a “volcano” shape, which means the housing price next to the train station is not the highest price due to some negative externality of train station, such as the congestion, environment and traffic noises and safety concern. It is empirically found in our case that the housing price reaches the peak around 400m away from the station, meaning the “best” location choice in the selected area of Taichung train stations. This results of housing price with “volcano” shape confirm the same findings of Chen and Lu（2019）in the study of Kaohsiung Light Rail case.

In this study, a grade classification model was constructed based on the building façade using AI's image recognition technology to establish a new grade indicator and discover the features of each grade. Specifically, the façade images were collected from the property portfolio of J-REIT. The collected facade images were then used to create a learning model by convolutional neural network (CNN) for image recognition technology. Its learning model was able to output more than 90% accuracy. Using this learning model, the features of each grade were identified by comparing the training sample with the predicted sample and visualizing the features., and the potential impact of this model on real estate values was considered using the fair value per net lettable area in each location. The results suggest that the façade grade has the potential to influence real estate values.

We study how an expected return shock affects households’ portfolio decisions, and the role of financial literacy in shaping those decisions. By using Shanghai's property tax trial in 2011, which altered the expected returns of investment properties compared to financial assets for a group of households, we examine how households adjust their portfolios in response to this shock. Following an expected return shock of 0.4 percentage points in favor of financial assets, households impacted by the property tax trial increased their investments in financial assets relative to housing assets; however, the extent to which they increased investments in risky assets varied considerably depending on their level of financial literacy. Households in the treatment group with high financial literacy increased their participation in the stock market and allocated a higher percentage of their portfolio to risky assets by 4.8 and 8.4 percentage points, respectively, while the low financial literacy group did not exhibit a significant increase in risky asset investments. We structurally estimate a quantitative life-cycle model with heterogeneity to match the group-specific moments and response functions, and find that households with low financial literacy on average bear a fixed entry cost of approximately $6,400 compared to $2,700 of households with high financial literacy, and that the expected excess return of stocks for the low financial literacy group is on average 2.8% compared to 5.4% for the high financial literacy group.

Treating the Russian invasion of Ukraine as a rare disaster event and defining proximity as both physical distance and political closeness, we analyze investors’ response to disaster risk by examining the performance of commercial real estate investments in countries of proximity to the event. We find that proximity to the war matters, but the impact of the disaster is not uniform across different property types. Firms with green and less obsolete properties are less likely to experience negative abnormal returns. Our findings highlight the differences in equity risk premia even within the same industry facing the same disaster, thereby contributing to the origins of internationalization “discount.” We also find support for the eminence of reducing reliance on brown fuel.