Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).
Room 209B, 2/F, Cheng Yu Tung Building, The Chinese University of Hong Kong
香港中文大学郑裕彤楼 2楼 209B 室
Presentations
Similar land price zone: a case study in Gwangjin-gu, Seoul, Korea
Seungwoo SHIN, Yeonjae Lee
konkuk university, Korea, Republic of (South Korea);
Discussant: Jeongseop SONG (Konkuk University);
Purpose
We estimate the zones of similar land prices using publicly announced land price data in Gwangjin-gu, Seoul. We estimate similar land price zones by zoning, current usage, and year.
The main research variables are the coordinates of the land parcel, officially announced land price, zone, and current usage status. In addition, the range of value influence of the subway station on the surrounding land price is estimated in the same way. Finally, the ranges of various similar price area maps are visualized.
Data:
The data for this study are the publicly announced land prices of Gwangjin-gu, Seoul. The observation period of this study is from January 2011 to January 2019.
Approach/Method
The methodology of this study is a generic t-test. First, estimate the borderline in which price gaps occur based on each individual parcel in Gwangjin-gu. A t-test is a method for confirming the statistical significance of the price gap between price zones.
Findings
The zone of similar land prices has widened over time based on the results of our studies. This implies that land prices in peripheral areas rose faster than in high-priced areas. A structural price break has occurred in the residential zones. Due to the large-scale residential development, there has been a structural break in land prices between the developed and surrounding areas. In the case of Gwangjin-gu, the commercial zone is tiny, so there has been no remarkable change in the years. Last, when considering the current usage situation, the range of similar price zones tends to expand in the case of parcels used for commercial purposes.
practical meaning
This study analyzes the officially announced land price from 2011 to 2019. Following our research findings, Gwangjin-gu is divided into various land price zones. Various zones are correctly classified, and appropriate urban regeneration plans are presented. In addition, our findings regarding similar price zones provide helpful information for officially announced land price calculation personnel, real estate brokers, and real estate investors.
Originality/ value
According to the central place theory, which is the basis of the Alonso-Muth-Mills model (AMM model), land prices decrease linearly as one moves away from the center. The findings of this study focus on the property development activities as factors that affect land prices. In addition, equal attention is paid to zoning regulations and deregulation. These factors interactively affect land prices, increasing the complexity of land price zone distribution. Visualizing a land price distribution map considering these various factors is crucial for a fair and efficient officially announced land price assessment.
Keywords: Land price, Similar price zone, t-test, Zoning, Land usage
THE PERFORMANCE OF HOTEL INVESTMENTS IN ASIA-PACIFIC
Graeme Newell1, Jufri Marzuki1, Martin Hoesli2
1Western Sydney University, Australia; 2University of Geneva, Switzerland, University of Aberdeen , UK;
Discussant: Hainan SHENG (Virginia Tech);
The hotel sector is an important real estate sector globally. This sees hotel investments figure prominently in the portfolios of many real estate investment managers in the Asia-Pacific (eg: CapitaLand, SC, IGIS, ESR). In Asia-Pacific, there were over $37 B in hotel transactions over 2019-2021, accounting for 7.4% of Asia-Pacific commercial real estate transaction volumes. This paper uses the MSCI Asia-Pacific hotel real estate index, which comprises 440 hotel properties across 64 portfolios valued at $32.0 B to assess the performance of Asia-Pacific hotels over 2005-2022. Analyses are also done for pan-Asia and for specific countries in Asia. The other major real estate sub-sectors and major asset classes are also used to assess the risk-adjusted performance and portfolio diversification benefits of Asia-Pacific hotels in an Asia-Pacific real estate portfolio and in a mixed-asset portfolio. The strategic real estate investment implications are also assessed, as well as the impact of COVID on hotel investment performance in the Asia-Pacific.
Mortgage default analysis: Lender-friendly or borrower-friendly states
Yeonjae LEE, Seungwoo SHIN
Konkuk University, Korea, Republic of (South Korea);
Discussant: Alla KOBLYAKOVA (Nottingham Trent University);
Purpose
We use the US Freddie Mac mortgage performance history and origination information to study the mortgage borrowers' default choices.
The main research dummy variables are a state dummy variable with a lender-friendly recourse loan system and another state dummy variable that recognizes a borrower-friendly judicial foreclosure and redemption right.
In addition, the monthly current DTI variable of individual mortgage borrowers is calculated and compared with the current unemployment rate variable by month and MSA. Lastly, the panel logit model, the multilevel perceptron (MLP), and the random forest (RF) are compared in terms of the prediction power of the models.
Data:
The data in this study is Prime FRM data that originated in 2009. The observation period of this study is from January 2010 to December 2020.
Approach/Method
The estimated models of this study are COX HAZARD MODEL, Panel Logit Model, Multilevel Perceptron, and Random Forest.
The other newly introduced variable in this study is the expected monthly income of individual mortgage borrowers. The income at origination is adjusted to the cumulative wage increase rate by MSA, and the unemployment rate for each MSA was considered as the unemployment probability of the individual borrowers.
Findings
Based on our Primary Findings, a lender-friendly recourse state dummy had a statistically significant negative influence on the borrowers' mortgage default choice. On the other hand, the borrower-friendly Judicial Foreclosure and Redemption dummies had little effect. The endogenous effect of effectively filtering risky borrowers through a more accurate underwriting in the loan process has occurred.
The monthly current DTI variable is significant and increases the model's goodness of fit, while the expected monthly income of individual mortgage borrowers contributes very little.
Regarding prediction power testing, unlike common sense, the F1 score of the panel logit model was not inferior compared to the machine learning methodology. This is due to the disadvantages of machine learning, which cannot use the panel data.
Practical implication
Based on our Primary Findings, the lender-friendly legal system could be linked to more flexible loan policies and, as a possibly unintended result, increase the utility of mortgage borrowers. In addition, in the case of panel analysis, the traditional econometric model may be superior to machine learning in terms of predictions, let alone explanatory power.
Originality/ Value
This study investigates the impact of lender- and borrower-friendly legal systems on the borrower's default decision. In addition, the expected monthly income and the current DTI are calculated for each borrower in an innovative fashion and used as a substitute variable for an area unemployment rate. Finally, the predictive power test results confirmed the advantages of the panel analysis.
