# RECORD 59a1872eca8f8697a739527375ba1a22843bf5bd # TITLE: The housing price dynamics and macroeconomic factors: Evidence of Taiwan housing market # YEAR: 2025.0 VENUE: International Journal of Strategic Property Management DOI: 10.3846/ijspm.2025.24610 # AUTHORS: Chih-Hsing Hung; Hui-Chuan Liu; Pei-Ran Su; Shyh-Weir Tzang # PAGES: 21 CANDIDATE TABLE PAGES (1-based): [8, 9, 10, 11, 15, 16, 17, 18] # ABSTRACT: The housing price in Taiwan has seen a sharp increase since 2020 after a long period of stability from 2011 to 2019. To provide a more granular analysis on this substantial rise of the housing prices since 2020 in Taiwan, we applied the hedonic pricing model to analyze the completed and presale residential housing across four types of houses: luxury, first-class, second-class and suite. By incorporating relevant macroeconomic variables from 2014 to 2022, we try to capture the demand of homebuyers across the three biggest cities in Taiwan. In the completed housing market, we find that the suite prices in Taipei and Taichung cities are more significantly affected by Taiwan’s continuing economic growth, whereas the suite prices in Kaohsiung city are hindered by its historically stagnant employment market. Notably, suite prices in Kaohsiung display greater sensitivity to construction cost changes and broader economic cycles compared to the suite prices in Taipei and Taichung. These findings highlight the unique dynamics of real estate markets in northern, central, and southern Taiwan, providing valuable insights for housing investors, prospective homebuyers in different regions and policymakers in their decision making. ===== PAGE 1 (score 10, numeric-paren 0, table-words 0, var-words 10) ===== This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Copyright © 2025 The Author(s). Published by Vilnius Gediminas Technical University ISSN: 1648-715X / eISSN: 1648-9179 INTERNATIONAL JOURNAL OF STRATEGIC PROPERTY MANAGEMENT 2025 Volume 29 Issue 4 Pages 245–265 https://doi.org/10.3846/ijspm.2025.24610 THE HOUSING PRICE DYNAMICS AND MACROECONOMIC FACTORS: EVIDENCE OF TAIWAN HOUSING MARKET Chih-Hsing HUNG1, Hui-Chuan LIU2, Pei-Jia SU1, Shyh-Weir TZANG   3* 1 Department of Money and Banking, National Kaohsiung University of Science and Technology, Kaohsiung, Taiwan, R.O.C. 2 College of Management, National Kaohsiung University of Science and Technology, Kaohsiung, Taiwan, R.O.C. 3 Department of Finance, Asia University, Taichung, Taiwan, R.O.C. Article History: Abstract. The housing price in Taiwan has seen a sharp increase since 2020 after a long period of stability from 2011 to 2019. To provide a more granular analysis on this substantial rise of the housing prices since 2020 in Taiwan, we applied the hedonic pricing model to analyze the completed and presale residential hous- ing across four types of houses: luxury, first-class, second-class and suite. By incorporating relevant macro- economic variables from 2014 to 2022, we try to capture the demand of homebuyers across the three biggest cities in Taiwan. In the completed housing market, we find that the suite prices in Taipei and Taichung cities are more significantly affected by Taiwan’s continuing economic growth, whereas the suite prices in Kaohsiung city are hindered by its historically stagnant employment market. Notably, suite prices in Kaohsiung display greater sensitivity to construction cost changes and broader economic cycles compared to the suite prices in Taipei and Taichung. These findings highlight the unique dynamics of real estate markets in northern, central, and southern Taiwan, providing valuable insights for housing investors, prospective homebuyers in different regions and policymakers in their decision making. ■received 13 January 2025 ■accepted 19 June 2025 Keywords: hedonic price model, housing price index, macroeconomic variables, completed residential property, presale residential property. * Corresponding author. E-mail: swtzang@gmail.com 1. Introduction According to the latest international statistics, Taiwan has one of the lowest fertility rates in the world. In 2024, Tai- wan’s total fertility rate was estimated at 1.09 children per woman–well below the replacement level of 2.1 needed to maintain a stable population (worldpopulationreview. com). This low fertility rate has led to population decline and aging, significantly affecting education, the labor mar- ket, and economic development. Due to Taiwan’s declining population growth rate, people will increasingly move to and concentrate in several major cities–an almost certain trend for the future. As a result, to measure its impact on housing market, we focused on the three major cities in Taiwan to examine the changes in the housing market, as this can reflect shifts in real estate supply and demand and their impact on the value of different tiers of housing. Based on the hedonic pricing models, this study also aims to highlight the recent sharp increase in 2020 in Taiwan housing prices by developing separate indices for existing and presale homes across Taiwan’s three largest cities–Taipei, Taichung, and Kaohsiung. By considering the macroeconom- ic variables such as consumer price indices (CPI), construc- tion cost indices (CCI), national income (NI), money supply, mortgage interest rates, and Taiwan Stock Exchange Market (TAIEX) performance, this research seeks to uncover distinct relationships between these factors and housing prices in different market segments. To better account for the slight changes in the period prior to 2020, we collected the data from 2014Q2 to 2022Q1. The findings will contribute to a deeper understanding of the dynamics of Taiwan’s real es- tate market and provide timely information for policy mak- ers to constrain the overheating housing market. Real estate in Taiwan holds both economic and cul- tural significance. Beyond serving as a hedge against infla- tion for investors, property ownership symbolizes wealth and social status in Taiwanese society. According to Tai- wan 2020 Population and Housing Census, the average rate of homeownership was 78.6%, of which Taipei city and Changhua county were 72% and 87%, the lowest and highest in Taiwan island, respectively. In this study, we se- lect three cities, Taipei, Taichung, and Kaohsiung, as the focus of analysis by highlighting their geographical and economic differences, which can be better reflected in their respective real estate markets. Taipei is the traditional political center; Taichung is a hub for traditional, emerging, and electronics industries, and Kaohsiung is a center for traditional industries and tourism. ===== PAGE 8 (score 37, numeric-paren 0, table-words 5, var-words 8) ===== 252 C.-H. Hung et al. The housing price dynamics and macroeconomic factors: Evidence of Taiwan housing market 4.2. Descriptive statistics By Chang et al. (2008), we screened the data by removing the transactions deviating from normal market prices, and special real estate registration data, followed by the consoli- dation of standardized housing information for each cat- egory. Tables 6 and 7 are the feature statistics according to the four categories of standard housing, i.e., luxury homes, the first-class housing, the second-class housing, and suite in Taipei city. Note that the data of 2016Q4 and 2019Q3 in Table 7 are not shown due to the incomplete data. Table 8 shows the basic statistics of Taipei city existing home price indices across four different categories of hous- ing. We observed that the standard deviation of the luxury home and the second-class home price indices in Taipei are relatively high. This suggests greater fluctuation over the eight-year period. Moreover, the maximum and minimum values of the luxury home index exhibit the largest differ- ence, indicating the greatest magnitude of variation. Table 9 shows the basic statistics of Taichung city ex- isting home price indices across four different categories of housing. We noted that the standard deviation of the suite index exhibits the highest variability, followed by the second-class housing index. The difference between the maximum and minimum values of the luxury home index is the smallest, indicating the least fluctuation. Table 10 shows the basic statistics of Kaohsiung city existing home price indices across four different catego- ries of housing. We noted that the standard deviation and maximum value of the suite index are the highest com- pared to other housing categories. Conversely, the stand- ard deviation of the first-class index is the smallest, indi- cating less fluctuation. Table 6. The existing luxury and the first-class housing in Taipei city Cat- egory Feature Stats 2014Q2 2014Q3 2014Q4 2015Q1 2015Q2 2015Q3 2015Q4 2016Q1 2016Q2 2016Q3 2016Q4 Luxury homes total number of floors mean 14.24 15.16 16.22 16.89 18.15 18.35 17.36 15.31 15.88 15.88 16.61 sampled floor mean 8.14 8.69 8.70 9.14 9.37 9.57 9.79 9.15 9.48 8.71 9.73 registered total area median 77.01 79.54 90.46 75.15 79.85 83.27 97.31 87.78 87.29 82.03 93.73 housing age (year) median 2.85 3.80 2.60 4.30 4.60 5.60 3.05 3.15 3.90 1.45 4.30 parking lot mean 1.61 1.71 1.99 1.35 1.68 1.60 1.88 1.58 1.62 1.55 1.88 location mean 0.47 0.63 0.47 0.74 0.66 0.67 0.65 0.59 0.64 0.73 0.72 number of quarterly data # obs. 108 249 463 328 326 263 348 130 146 56 190 Feature Stats 2017Q1 2017Q2 2017Q3 2017Q4 2018Q1 2018Q2 2018Q3 2018Q4 2019Q1 2019Q2 2019Q3 total number of floors mean 16.41 16.85 18.50 17.23 16.86 18.94 17.43 19.81 18.97 17.79 17.97 sampled floor mean 9.56 9.78 10.59 10.02 9.12 12.54 9.17 10.66 10.40 9.71 10.58 registered total area median 92.39 94.48 93.26 92.29 87.15 104.99 84.51 91.92 95.80 86.22 75.85 housing age (year) median 4.45 5.00 3.60 4.00 2.35 3.50 3.80 4.00 5.30 6.10 4.00 parking lot mean 1.83 1.87 1.87 1.89 1.58 2.01 1.78 1.80 1.83 1.72 1.58 location mean 0.80 0.48 0.45 0.50 0.59 0.68 0.67 0.69 0.55 0.63 0.83 number of quarterly data # obs. 176 211 205 218 90 99 147 134 108 155 101 Feature Stats 2019Q4 2020Q1 2020Q2 2020Q3 2020Q4 2021Q1 2021Q2 2021Q3 2021Q4 2022Q1 total number of floors mean 16.26 18.62 17.48 17.75 17.80 20.49 18.27 16.83 21.63 16.83 sampled floor mean 9.65 10.32 10.25 10.26 9.65 11.46 10.49 9.56 13.45 9.04 registered total area median 83.33 74.25 85.57 80.44 89.37 95.29 76.36 74.52 129.56 69.13 housing age (year) median 7.95 5.30 7.00 7.10 7.80 6.65 7.90 7.60 6.15 8.30 parking lot mean 1.46 1.71 1.61 1.47 1.69 1.95 1.48 1.36 2.64 1.23 location mean 0.33 0.74 0.53 0.58 0.56 0.60 0.73 0.66 0.78 0.74 number of quarterly data # obs. 72 91 181 204 203 312 208 249 64 297 ===== PAGE 9 (score 42, numeric-paren 0, table-words 2, var-words 8) ===== International Journal of Strategic Property Management, 2025, 29(4), 245–265 253 Cat- egory Feature Stats 2014Q2 2014Q3 2014Q4 2015Q1 2015Q2 2015Q3 2015Q4 2016Q1 2016Q2 2016Q3 2016Q4 The first- class housing total number of floors mean 12.69 13.96 14.08 14.15 14.17 14.21 14.11 14.26 17.22 18.49 18.14 sampled floor mean 9.22 8.26 8.18 8.07 8.17 8.11 8.41 8.58 9.94 10.27 10.23 registered total area median 39.59 47.11 45.20 45.51 41.71 41.37 42.92 44.03 41.35 39.25 38.55 housing age (year) median 6.40 6.40 6.10 6.40 6.70 8.30 8.10 8.20 5.20 3.30 3.40 parking lot mean 0.63 0.85 0.78 0.78 0.66 0.59 0.60 0.62 0.76 0.82 0.77 location mean 0.56 0.42 0.53 0.54 0.66 0.64 0.65 0.66 0.71 0.73 0.73 number of quarterly data # obs. 197 515 967 931 769 490 607 553 736 532 624 Feature Stats 2017Q1 2017Q2 2017Q3 2017Q4 2018Q1 2018Q2 2018Q3 2018Q4 2019Q1 2019Q2 2019Q3 total number of floors mean 15.66 15.58 15.58 15.57 15.73 15.78 14.20 14.35 14.53 14.64 14.57 sampled floor mean 8.99 8.91 8.80 9.12 9.40 9.85 8.35 8.34 8.53 8.79 8.90 registered total area median 42.08 42.32 43.80 42.79 42.78 41.51 42.67 41.61 41.37 40.86 40.21 housing age (year) median 7.40 8.00 8.10 8.85 9.00 9.20 8.75 9.90 8.90 9.40 9.80 parking lot mean 0.71 0.70 0.73 0.67 0.65 0.62 0.70 0.67 0.64 0.60 0.59 location mean 0.66 0.58 0.51 0.51 0.53 0.60 0.40 0.44 0.46 0.49 0.46 number of quarterly data # obs. 496 724 723 752 486 281 176 275 455 529 620 Feature Stats 2019Q4 2020Q1 2020Q2 2020Q3 2020Q4 2021Q1 2021Q2 2021Q3 2021Q4 2022Q1 total number of floors mean 14.83 15.00 15.08 14.91 14.76 14.78 14.74 15.12 14.84 14.71 sampled floor mean 9.18 9.15 8.85 8.59 8.52 8.67 8.53 8.61 12.22 11.75 registered total area median 40.10 40.28 40.74 42.02 41.56 40.67 39.68 39.42 40.94 40.63 housing age (year) median 8.90 8.90 8.80 9.80 10.15 10.80 11.30 13.80 13.70 13.80 parking lot mean 0.64 0.66 0.70 0.67 0.68 0.65 0.65 0.61 0.67 0.66 location mean 0.45 0.44 0.46 0.49 0.50 0.51 0.49 0.51 0.57 0.57 number of quarterly data # obs. 664 622 704 729 770 662 559 519 1008 1099 Table 7. The existing second-class housing and suite in Taipei city Category Feature Stats 2014Q2 2014Q3 2014Q4 2015Q1 2015Q2 2015Q3 2015Q4 2016Q1 2016Q2 2016Q3 The sec- ond-class housing total number of floors mean 13.51 14.00 14.44 14.57 14.97 14.77 13.99 14.07 14.25 14.07 sampled floor mean 7.70 7.67 8.07 8.16 8.04 7.79 7.60 7.68 7.58 7.71 registered total area median 55.83 50.03 48.36 49.23 48.83 48.19 47.65 48.25 47.79 48.16 housing age (year) median 5.10 5.80 5.10 5.80 6.50 6.40 7.00 8.35 7.70 6.60 parking lot mean 1.00 0.95 0.91 0.96 1.02 0.94 0.89 0.95 0.93 0.95 location mean 0.35 0.38 0.33 0.38 0.41 0.38 0.40 0.28 0.26 0.26 number of quarterly data # obs. 524 1285 1875 1925 1713 1590 991 462 411 271 Feature Stats 2017Q1 2017Q2 2017Q3 2017Q4 2018Q1 2018Q2 2018Q3 2018Q4 2019Q1 2019Q2 total number of floors mean 14.38 14.25 14.17 14.23 14.56 14.46 14.31 14.53 14.35 14.06 sampled floor mean 7.53 7.72 7.58 7.53 7.80 7.90 7.69 7.87 7.76 7.90 registered total area median 49.48 49.55 51.49 50.87 49.68 49.28 48.32 48.89 48.84 49.81 housing age (year) median 8.95 9.10 9.10 9.80 10.00 10.50 11.00 10.80 11.50 12.10 parking lot mean 0.93 0.87 0.93 0.92 0.90 0.88 0.84 0.86 0.86 0.87 End of Table 6 ===== PAGE 10 (score 41, numeric-paren 0, table-words 2, var-words 9) ===== 254 C.-H. Hung et al. The housing price dynamics and macroeconomic factors: Evidence of Taiwan housing market Table 8. Statistics of Taipei city existing home price indices across four different categories of housing* Categories Mean Median Max Min Std. dev. Skewness Kurtosis J-B Probability # obs. Luxury home 101.21 100.99 108.62 97.79 2.10 1.29 6.12 21.84 0.0000 32 First-class 98.27 98.23 100.00 96.57 0.90 0.04 2.25 0.75 0.6867 32 Second-class 99.98 99.37 105.63 96.65 2.43 0.91 2.79 4.44 0.1088 32 Suite 100.42 100.50 103.67 97.07 1.99 –0.07 1.79 1.98 0.3724 32 Note: *The base period to compute the index in this study is the second quarter of 2014 (2014Q2). Category Feature Stats 2014Q2 2014Q3 2014Q4 2015Q1 2015Q2 2015Q3 2015Q4 2016Q1 2016Q2 2016Q3 location mean 0.32 0.34 0.32 0.35 0.38 0.39 0.34 0.35 0.33 0.32 number of quarterly data # obs. 480 477 548 639 547 428 467 417 626 486 Feature Stats 2019Q4 2020Q1 2020Q2 2020Q3 2020Q4 2021Q1 2021Q2 2021Q3 2021Q4 2022Q1 total number of floors mean 14.31 14.29 14.20 14.37 14.30 14.35 14.48 14.55 14.65 14.99 sampled floor mean 7.93 7.69 7.59 7.77 7.59 7.75 7.94 7.91 8.12 8.29 registered total area median 47.29 45.77 45.66 47.48 48.74 48.21 44.63 43.33 43.76 44.09 housing age (year) median 11.45 12.20 12.70 12.70 13.25 15.00 14.55 14.00 13.85 14.80 parking lot mean 0.88 0.82 0.81 0.85 0.85 0.83 0.82 0.81 0.81 0.72 location mean 0.34 0.40 0.39 0.35 0.39 0.43 0.42 0.38 0.36 0.45 number of quarterly data # obs. 464 436 477 575 522 437 430 513 288 241 Category Feature Stats 2014Q2 2014Q3 2014Q4 2015Q1 2015Q2 2015Q3 2015Q4 2016Q1 2016Q2 2016Q3 Suite total number of floors mean 12.82 13.25 13.43 13.26 13.42 13.60 13.67 13.78 16.17 17.21 sampled floor mean 8.30 7.88 7.62 7.86 8.58 8.45 8.51 8.58 8.87 9.18 registered total area median 17.43 18.04 18.76 19.51 19.51 19.41 20.20 20.07 19.43 19.92 housing age (year) median 18.45 17.45 14.45 16.50 18.70 18.70 18.95 18.80 8.70 7.80 parking lot mean 0.07 0.09 0.14 0.13 0.07 0.06 0.08 0.09 0.08 0.07 location mean 0.78 0.75 0.74 0.69 0.62 0.62 0.62 0.62 0.74 0.80 number of quarterly data # obs. 196 304 258 245 227 205 190 175 203 147 Feature Stats 2017Q1 2017Q2 2017Q3 2017Q4 2018Q1 2018Q2 2018Q3 2018Q4 2019Q1 2019Q2 total number of floors mean 14.43 13.89 13.90 16.10 16.69 14.94 14.29 13.99 14.13 14.13 sampled floor mean 8.70 8.55 8.42 9.26 9.79 9.18 8.58 8.11 8.64 8.52 registered total area median 19.55 19.43 19.43 19.07 19.58 19.85 19.27 19.63 19.21 19.47 housing age (year) median 19.30 13.60 20.50 10.30 9.30 11.50 14.55 20.85 20.50 20.40 parking lot mean 0.08 0.09 0.06 0.04 0.04 0.04 0.04 0.05 0.09 0.08 location mean 0.78 0.79 0.72 0.78 0.81 0.75 0.73 0.70 0.73 0.74 number of quarterly data # obs. 169 163 173 279 264 187 180 122 127 220 Feature Stats 2019Q4 2020Q1 2020Q2 2020Q3 2020Q4 2021Q1 2021Q2 2021Q3 2021Q4 2022Q1 total number of floors mean 15.10 15.08 14.12 14.31 14.38 14.28 14.24 14.25 14.13 14.15 sampled floor mean 9.45 9.54 9.10 9.33 9.17 8.82 8.60 8.64 8.59 8.62 registered total area median 19.70 19.56 18.20 16.27 15.50 15.50 15.35 15.36 15.36 15.25 housing age (year) median 14.90 22.25 23.30 22.60 23.80 24.40 24.90 25.20 25.10 26.20 parking lot mean 0.09 0.09 0.12 0.10 0.07 0.06 0.08 0.08 0.07 0.08 location mean 0.70 0.70 0.77 0.73 0.71 0.75 0.75 0.76 0.80 0.78 number of quarterly data # obs. 185 182 244 381 406 358 328 397 569 579 End of Table 7 ===== PAGE 11 (score 28, numeric-paren 0, table-words 4, var-words 5) ===== International Journal of Strategic Property Management, 2025, 29(4), 245–265 255 Table 9. Statistics of Taichung city existing home price indices across four different categories of housing* Categories Mean Median Max Min Std. dev. Skewness Kurtosis J-B Probability # obs. Luxury home 100.28 100.35 102.23 96.45 1.28 –0.93 4.22 6.56 0.0376 32 First-class 101.49 101.25 104.90 98.26 1.55 0.28 2.52 0.73 0.6935 32 Second-class 102.69 102.71 109.96 98.25 2.42 0.54 4.08 3.12 0.2101 32 Suite 108.72 108.04 163.73 59.38 18.03 0.65 6.18 15.70 0.0004 32 Note: *The base period to compute the index in this study is the second quarter of 2014 (2014Q2). Table 10. Statistics of Kaohsiung city existing home price indices across four different categories of housing* Categories Mean Median Max Min Std. dev. Skewness Kurtosis J-B Probability # obs. Luxury home 102.48 102.48 106.79 99.14 1.73 0.11 2.83 0.11 0.9469 32 First-class 100.65 100.54 103.81 98.28 1.17 0.63 3.62 2.63 0.2691 32 Second-class 101.99 100.87 108.14 99.84 2.29 1.22 3.48 8.24 0.0162 32 Suite 111.82 114.17 120.35 97.44 5.88 –0.76 2.82 3.12 0.2105 32 Note: *The base period to compute the index in this study is the second quarter of 2014 (2014Q2). From Table 11, it is observed that the standard devia- tion of the construction cost index in Taichung city is rela- tively higher compared to other cities, indicating a larger fluctuation range. Conversely, Kaohsiung city exhibits the smallest fluctuation range in the construction cost index. However, Kaohsiung city has the highest standard devia- tion in consumer price levels, suggesting the greatest de- gree of index variation. 4.3. Analysis of housing index Based on the results of the calculated residential price in- dices for various tiers over eight years in Taipei, Taichung, and Kaohsiung using the house price index model (Chang et al., 2008), Figures 3a to 5d present the existing housing index charts for the three biggest cities in Taiwan. Figures 3a to 3d contains four subplots of the four tiers of housing index for the existing housing property. Figure 3a shows the existing luxury residential index in Table 11. Summary statistics of macro and construction factors across three cities* Variables Mean median Max Min Std. dev. Skewness Kurtosis J-B Probability # obs. Taipei CPI 97.52 97.56 101.31 94.11 1.82 0.20 2.42 0.66 0.7195 32 Taichung CPI 97.24 97.51 101.45 93.48 2.01 0.18 2.32 0.78 0.6759 32 Kaohsiung CPI 97.12 97.37 101.25 92.94 2.20 0.02 2.22 0.80 0.6694 32 Taipei CCI 88.13 86.32 105.40 81.54 6.45 1.41 4.13 12.34 0.0021 32 Taichung CCI 86.85 84.65 105.99 80.07 7.02 1.49 4.27 13.98 0.0009 32 Kaohsiung CCI 88.31 86.45 104.25 81.32 6.25 1.29 3.79 9.77 0.0076 32 M1B 18021118 16972986 25389297 13716369 3397258 0.80 2.52 3.75 0.1536 32 M2 44103120 43445430 55006759 36443453 5033015 0.55 2.45 2.03 0.3616 32 NI 4057689 3990863 5066643 3469291 395478 0.86 2.95 3.91 0.1414 32 AMR 0.0165 0.0163 0.0199 0.0135 0.0021 0.12 2.04 1.29 0.5255 32 TAIEX 11333 10518 18219 8181 2986 1.30 3.36 9.15 0.0103 32 Note: *The base period to compute the index in this study is the second quarter of 2014 (2014Q2). Taipei, which exhibits significant fluctuations but generally remains within a certain range over the long term. This indicates periodic implementation of government policies aimed at controlling luxury home prices. Figure 3d illus- trates the suite index in Taipei, which has shown a recent upward trend. This suggests an increasing demand for rental suites driven by the influx of population, particularly from other regions. Consequently, there is also an increase in demand from investors purchasing suites, leading to a rise in the apartment price index. In view of the sensitiv- ity of the housing price index, the existing suite housing index seems to be the most sensitive to the changes in housing market, which showed a strong rebound as early as 2020Q1, and then the first- and second-class housing indices (Figures 3b and 3c) began to rise accordingly. The existing suite housing index, therefore, should become an important signal for the government housing policy mak- ers to monitor closely so as to make timely reaction curb the speculation in housing market. ===== PAGE 12 (score 3, numeric-paren 0, table-words 0, var-words 3) ===== 256 C.-H. Hung et al. The housing price dynamics and macroeconomic factors: Evidence of Taiwan housing market 90 95 100 105 110 103Q2 103Q4 104Q2 104Q4 105Q2 105Q4 106Q2 106Q4 107Q2 107Q4 108Q2 108Q4 109Q2 109Q4 110Q2 110Q4 a) Taipei existing luxury housing index 94 95 96 97 98 99 100 101 103Q2 103Q4 104Q2 104Q4 105Q2 105Q4 106Q2 106Q4 107Q2 107Q4 108Q2 108Q4 109Q2 109Q4 110Q2 110Q4 b) Taipei existing first-class housing index 92 94 96 98 100 102 104 106 108 103Q2 103Q4 104Q2 104Q4 105Q2 105Q4 106Q2 106Q4 107Q2 107Q4 108Q2 108Q4 109Q2 109Q4 110Q2 110Q4 c) Taipei existing second-class housing index 92 94 96 98 100 102 104 106 103Q2 103Q4 104Q2 104Q4 105Q2 105Q4 106Q2 106Q4 107Q2 107Q4 108Q2 108Q4 109Q2 109Q4 110Q2 110Q4 d) Taipei existing suite housing index Note: The X axis is using the Republic of China (ROC) calendar which is the method of numbering years currently used in ROC (including the islands of Taiwan, Kinmen and Matsu). To find out the ROC year equivalent to any Gregorian calendar year, we can add 1911 to the ROC year. For example: 103 + 1911 = 2014 year of the Gregorian calendar. Figure 3. Taipei existing housing indices across four tiers of housing from 2014Q2 to 2021Q4 Figures 4a to 4d depict the luxury, the first-class, the second-class and suite housing indices of the existing property in Taichung city. Over the past eight years, the luxury, the first-class and the second-class housing indices have all shown a trend of moderate growth. This indicates that in recent years, an increasing number of people have been moving to Taichung. According to statistics from the Ministry of the Interior, despite the serious issue of de- clining birth rates in the past seven years, the population of Taichung has increased by more than 100,000 people. Consequently, there has been an increased demand for housing, leading to a growth in price indices. In view of the sensitivity of the housing price index, the existing luxury and the first-class housing price indices seem to be more sensitive to the housing market than the second- and suite housing indices which are more stable in completed hous- ing market. Figures 5a to 5d depict the luxury, the first-class, the second-class and suite housing indices of the exist- ing property in Kaohsiung city. By the Figure 5d of the suite housing index, it has consistently remained around the mean value in recent years. This suggests that there may be relatively fewer job opportunities in Kaohsiung compared to Taipei. As a result, the increase in demand for studio apartments or suites among office workers has been relatively small, and there is also less demand from investors in the suite market. Therefore, this price index has shown a stable trend in recent years. On the other hand, the existing second-class housing price has risen much earlier than other indices and cities since 2019Q2, indicating possible potential speculators from the market. This can also be an important signal for local government to monitor in the future. Due to insufficient data on pre-sale housing, the cal- culation of the eight-year pre-sale housing price index is not feasible. We only calculated and showed the results of the pre-sale housing price indices for Taipei, Taichung and Kaohsiung for a three-year period in Figures 6a–8d. Figure 6a shows the luxury pre-sale housing index in Taipei, reaching its peak in the Q1 of 2019, and entering a consolidation phase after a decline in the Q2 of 2019. Figure 6b illustrates the first-class pre-sale housing index in Taipei, showing an upward trend reaching its peak in the Q1 of 2022. Figure 6c represents the second-class presale housing index in Taipei, reaching its highest point in the Q2 of 2021 with the index consolidating before reaching the peak. Figure 6d depicts the suite presale housing in- dex in Taipei that remains in a consolidation phase over three years. Figure 7a depicts the luxury pre-sale housing index in Taichung, reaching its peak in the Q1 of 2020. Following a decline in the subsequent year, it then demonstrates an upward trend. Figure 7b illustrates the first-class presale housing index in Taichung, showing an upward trend over ===== PAGE 15 (score 31, numeric-paren 16, table-words 4, var-words 4) ===== International Journal of Strategic Property Management, 2025, 29(4), 245–265 259 three years. Figures 7c and 7d represent the second-class and suite housing indices in Taichung respectively, both maintaining a consolidation phase over three years. Figure 8a shows the luxury pre-sale housing index in Kaohsiung, reaching its lowest point in the Q2 of 2019, followed by an upward trend. Figure 8b illustrates the first- class presale housing index in Kaohsiung, peaking in the Q3 of 2019. After a decline the following year, it shows an upward trend. Figure 8c represents the second-class presale housing index in Kaohsiung, exhibiting an upward trend over three years. Figure 8d depicts the suite presale housing index in Kaohsiung, peaking in the Q2 of 2019, declining until the Q4 of 2019, and then entering a con- solidation phase. 4.4. Unit root test In the context of unit root tests for variables such as price indices and money supply, the test with a single intercept term performs better than the test with both intercept and trend terms (Elliott et al., 1996; Culver & Papell, 1997; Ley- bourne & Newbold, 1999; Ng & Perron, 2001). Therefore, the study adopted the Augmented Dickey-Fuller (ADF) test with a single intercept term to confirm whether the mac- roeconomic variables are stationary data. Table 12 above presents the tests before and after differencing. The results reveal that the first-order differencing CPI, money supply (M2), average mortgage interest rates of the five major banks, TAIEX, as well as the second-order differencing of the construction cost index and national income in Taipei, Taichung, and Kaohsiung, which reject the presence of a unit root. Therefore, this study employs the differenced macroeconomic variables for regression analysis. 4.5. Regression results The hses in Taiwan are categorized into the primary four types: luxury homes, the first-class, the second-class and suites as shown in the following: 94 96 98 100 102 104 106 108 103Q2 103Q4 104Q2 104Q4 105Q2 105Q4 106Q2 106Q4 107Q2 107Q4 108Q2 108Q4 109Q2 109Q4 110Q2 110Q4 a) Kaohsiung presale luxury housing index 94 96 98 100 102 104 106 103Q2 103Q4 104Q2 104Q4 105Q2 105Q4 106Q2 106Q4 107Q2 107Q4 108Q2 108Q4 109Q2 109Q4 110Q2 110Q4 b) Kaohsiung presale first-class housing index 94 96 98 100 102 104 106 108 110 103Q2 103Q4 104Q2 104Q4 105Q2 105Q4 106Q2 106Q4 107Q2 107Q4 108Q2 108Q4 109Q2 109Q4 110Q2 110Q4 c) Kaohsiung presale second-class housing index 0 20 40 60 80 100 120 140 103Q2 103Q4 104Q2 104Q4 105Q2 105Q4 106Q2 106Q4 107Q2 107Q4 108Q2 108Q4 109Q2 109Q4 110Q2 110Q4 d) Kaohsiung presale suite housing index Figure 8. Kaohsiung presale housing indices across four tiers of housing from 2019Q2 to 2022Q1 Table 12. Unit root test of macroeconomic variables using ADF test Macroeconomic variables Level First difference t-stat Second difference t-stat Taipei CPI 0.5349 –7.2165*** –7.2361*** Taichung CPI 0.1165 –4.9485*** –6.5187*** Kaohsiung CPI –0.4335 –5.8720*** –6.5726*** Taipei CCI 0.7294 –1.9685 –6.2356*** Taichung CCI 1.0966 –1.9904 –6.2546*** Kaohsiung CCI 3.1752 –2.4651 –6.3955*** M2 3.2084 –3.2987** –6.3022*** NI 3.3628 –2.2260 –7.3865*** AMR –0.7222 –4.5898*** –8.1799*** TAIEX 0.4984 –6.0249*** –11.6901*** Note: ***and ** denote the unit root test with 1% and 5% significance level, respectively. ===== PAGE 16 (score 38, numeric-paren 13, table-words 7, var-words 4) ===== 260 C.-H. Hung et al. The housing price dynamics and macroeconomic factors: Evidence of Taiwan housing market ■Luxury homes: high-income groups; ■First-class: middle- and high-income groups; ■Second-class: young couple or people earning the average-income groups; ■Suite: owner-occupiers and investors. After calculating the house price indices for each cat- egory using the index model, the regression analysis is conducted with macroeconomic variables to examine the positive or negative correlations and the extent of their impact. Tables 13, 15 and 17 show the correlation coef- ficients between the house price indices of various levels and macroeconomic variables for Taipei, Taichung, and Kaohsiung over eight years. Tables 14, 16 and 18 present the regression results of the house price indices of various levels and macroeconomic indicators for Taipei, Taichung and Kaohsiung over eight years. Existing property in Taipei city ■Taipei city luxury housing index for completed homes Model (1) in Table 14 shows the regression results of the Taipei city luxury housing index for completed homes and macroeconomic variables. There is no significant im- pact between the luxury housing index and macroeco- nomic variables, indicating that high-income groups are not influenced by any macroeconomic variables when purchasing homes. ■Taipei city first-class housing index for completed homes Model (2) in Table 14 shows the regression results of the Taipei city first-class housing index for completed homes and macroeconomic variables. The first-class hous- ing index is positively and significantly influenced by the Taipei city CPI and the money supply, while it is negatively and significantly influenced by mortgage rates. This indi- cates that when the market is active and interest rates are lower, middle- and high-income groups are likely to use the benefits from their investments to purchase homes due to reduced interest costs. ■Taipei city second-class housing index for completed homes Model (3) in Table 14 shows the regression results of the Taipei city second-class housing index for completed homes and macroeconomic variables. There is no signifi- cant impact between the second-class housing index and macroeconomic variables, indicating that young people and average-income earners have an urgent need for housing, and therefore changes in macroeconomic indi- cators are less likely to affect housing prices. Table 13. Correlation between the Taipei housing indices of existing property and macroeconomic variables Luxury First-class Second-class Suite Taipei CPI Taipei CCI M2 NI AMR TAIEX Luxury First-class 0.0908 Second-class 0.2575 –0.0925 Suite 0.0523 0.5278*** –0.2921 Taipei CPI –0.1314 0.2850 –0.0870 0.0550*** Taipei CCI –0.2514 0.1988 –0.0493 0.3334* 0.0217 M2 0.2645 0.5050*** –0.0187 0.4651*** –0.2217 0.0205 NI 0.0175 0.0224 –0.0696 –0.0688 0.4962*** –0.2277 –0.1964 AMR –0.0033 –0.0984 0.0838 –0.0915 0.0609 0.0581 0.1640 –0.1827 TAIEX 0.1444 0.1869 –0.0903 0.1968 –0.0323 0.0103 0.3438* 0.0817 –0.2294 Note: *** and * denote 1% and 10% significance level, respectively. Table 14. Regression of Taipei housing index and macroeconomic variables across four housing categories Luxury First-class Second-class Suite (1) (2) (3) (4) Intercept 100.5442*** 97.2415*** 100.3078*** 98.4118*** Taipei CPI –0.3353 0.6406*** –0.3470 0.6251 Taipei CCI –0.5667 0.1502 –0.1541 0.7264 M2 1.29E-06 1.37E-06*** –2.15E-07 2.81E-06*** NI 3.76E-07 –2.62E-07 –2.21E-07 –1.34E-07 AMR 2.2027 –467.0859* 398.6749 –921.4761 TAIEX 0.0001 –7.48E-05 –0.0001 –8.33E-05 adj R2 –0.077 0.396 –0.229 0.229 # obs. 32 32 32 32 Note: *** and * denote 1% and 10% significance level, respectively. ===== PAGE 17 (score 40, numeric-paren 13, table-words 9, var-words 2) ===== International Journal of Strategic Property Management, 2025, 29(4), 245–265 261 ■Taipei city suite index for completed homes Model (4) in Table 14 shows the regression results of the Taipei city suite index for completed homes and mac- roeconomic variables. The suite index is positively and sig- nificantly influenced by the Taipei city construction price index and the money supply. This indicates that when construction costs increase, housing prices also rise, lead- ing to an increased demand for renting suite apartments. Consequently, investors are likely to use their investment returns to purchase real estate. Existing property in Taichung city ■Taichung city luxury housing index for completed homes Model (1) in Table 16 shows the regression results of the Taichung city luxury housing index for completed homes and macroeconomic variables. There is a significant and positive impact between the luxury housing index and both the Taichung city CPI and the money supply. This in- dicates that when the market is more active, capital more easily flows into the luxury housing market, reflecting that there is both residential and investment demand for luxury homes in Taichung. ■Taichung city first-class housing index for completed homes Model (2) in Table 16 shows the regression results of the Taichung city first-class housing index for completed homes and macroeconomic variables. There is a significant and positive impact between the first-class housing index and the money supply. This suggests that middle- and high-income groups are likely to use the benefits from their investments to purchase higher-quality homes. ■Taichung city second-class housing index for com- pleted homes Model (3) in Table 16 shows the regression results of the Taichung city second-class housing index for com- pleted homes and macroeconomic variables. There is a significant and positive relationship between the second- class housing index and the Taichung CPI, and a significant and negative relationship with national income. This indi- cates that when prices rise and national income decreases, young people and average-income earners tend to buy houses below the average price. ■Taichung city suite index for completed homes Model (4) in Table 16 shows the regression results of the Taichung city suite index for completed homes and Table 15. Correlation between the Taichung housing indices of existing property and macroeconomic variables Luxury First-class Second- class Suite Taichung CPI Taichung CCI M2 NI AMR TAIEX Luxury First-class 0.5777*** Second-class 0.4216*** 0.4652** Suite 0.2178 0.3188* 0.2422 Taichung CPI 0.2944 0.1463 0.2817 0.3115 Taichung CCI 0.2175 –0.0056 –0.2026 –0.1562 0.1287 M2 0.3991*** 0.5750*** 0.1897 0.0754 –0.1367 –0.1563 NI –0.1106 –0.0981 –0.3936** 0.1678 0.2120 –0.0173 –0.1931 AMR –0.0309 0.1023 0.1046 0.0669 0.1636 –0.2589 0.1516 –0.1803 TAIEX 0.1530 0.2493 –0.1226 –0.0429 –0.1014 0.1773 0.3476 0.0821 –0.2311 Note: ***, ** and * denote 1%, 5% and 10% significance level, respectively. Table 16. Regression of Taichung housing index and macroeconomic variables across four housing categories Luxury First-class Second-class Suite (1) (2) (3) (4) Intercept 99.1064*** 100.1698*** 101.8984*** 102.2586*** Taichung CPI 0.8018* 0.5782 1.7055** 8.2996 Taichung CCI 0.2711 0.0593 –0.5949 –2.6463 M2 1.66E-06** 2.22E-06*** 9.96E-07 5.39E-06 NI –4.84E-07 –2.05E-07 –3.64E-06** 5.95E-06 AMR –375.5697 –21.2252 –803.0944 –1354.89 TAIEX –6.20E-05 1.03E-04 –0.0002 –7.40E-04 adj R2 0.195 0.224 0.241 –0.069 # obs. 32 32 32 32 Note: ***, ** and * denote 1%, 5% and 10% significance level, respectively. ===== PAGE 18 (score 46, numeric-paren 19, table-words 8, var-words 4) ===== 262 C.-H. Hung et al. The housing price dynamics and macroeconomic factors: Evidence of Taiwan housing market macroeconomic variables. There is no significant rela- tionship between the suite index and macroeconomic variables. This suggests that Taichung has attracted many residents from other counties and cities in recent years, and these new residents have a stable demand for suite apartments. Therefore, their demand for purchasing suite apartments is less affected by macroeconomic changes. Existing property in Kaohsiung city ■Kaohsiung city luxury housing index for completed homes Model (1) in Table 18 shows the regression results of the Kaohsiung city luxury housing index for completed homes and macroeconomic variables. There is no signifi- cant impact between the luxury housing index and macro- economic variables, indicating that the high-income group in Kaohsiung city views real estate has a stable demand. Therefore, changes in macroeconomic indicators do not affect luxury housing prices. ■Kaohsiung city first-class housing index for complet- ed homes Model (2) in Table 18 shows the regression results of the Kaohsiung city first-class housing index for completed homes and macroeconomic variables. There is a significant and positive relationship between the first-class housing index and both the Kaohsiung CPI and the money supply. This suggests that when the economy is active, middle- and high-income groups are likely to use the benefits from their investments to purchase higher-quality homes. ■Kaohsiung city second-class housing index for com- pleted homes Model (3) in Table 18 shows the regression results of the Kaohsiung city second-class housing index for com- pleted homes and macroeconomic variables. There is a significant and positive relationship between the second- class housing index and both the Kaohsiung CPI and the money supply. This indicates that when young people and average-income earners purchase homes, their limited funds lead them to use the profits from their investments to buy houses. Compared to the home-changing, middle- and high-income groups, the average-income group is more affected by the money supply. ■Kaohsiung city suite index for completed homes Model (4) in Table 18 shows the regression results of the Kaohsiung city suite index for completed homes and macroeconomic variables. There is a significant and nega- Table 17. Correlation between the Kaohsiung housing indices of existing property and macroeconomic variables Luxury First-class Second-class Suite Kaohsiung CPI Kaohsiung CCI M2 NI AMR TAIEX Luxury First-class 0.4792*** Second-class 0.6340*** 0.7284*** Suite 0.3965** 0.2822 0.5046*** Kaohsiung CPI 0.0019 0.2390 0.1233 0.0080 Kaohsiung CCI 0.0775 0.1414 –0.0227 –0.1532 –0.0367 M2 0.3442* 0.4158** 0.6245*** 0.2408 –0.2326 0.0790 NI –0.2588 –0.1378 –0.0602 –0.2085 0.3981** –0.4152** –0.1964 AMR 0.1927 0.0040 0.0937 –0.0066 0.1503 0.0757 0.1640 –0.1827 TAIEX 0.1337 0.1763 0.2267 0.3827** –0.1037 0.0828 0.3438* 0.0817 –0.2294 Note: ***, ** and * denote 1%, 5% and 10% significance level, respectively. Table 18. Regression of Kaohsiung housing index and macroeconomic variables across four housing categories Luxury First-class Second-class Suite (1) (2) (3) (4) Intercept 101.8043*** 99.6302*** 99.3691*** 110.9012*** Kaohsiung CPI 0.4161 0.7513** 1.0824* 1.7546 Kaohsiung CCI –0.1250 0.0229 –0.2435 –1.8221** M2 1.18E-06 1.39E-06** 3.99E-06*** 1.01E-06 NI –1.67E-06 –9.76E-07 –9.46E-07 –8.11E-06** AMR 323.9451 –434.592 –369.9859 –171.7181 TAIEX 1.96E-04 3.73E-05 6.12E-05 2.36E-03** adj R2 –0.007 0.205 0.349 0.192 # obs. 32 32 32 32 Note: ***, ** and * denote 1%, 5% and 10% significance level, respectively. ===== PAGE 19 (score 6, numeric-paren 0, table-words 0, var-words 6) ===== International Journal of Strategic Property Management, 2025, 29(4), 245–265 263 tive relationship between the suite index and both the Kaohsiung city CCI and national income, and a significant and positive relationship with the stock price index. In a city facing a prolonged economic depression, a decline in national income is likely to shift homebuyers’ demand from second-class housing to suites as well as studio apartments, where prices will see an increase. However, the inflexibility of rising rental prices of suites in Kaohsi- ung, coupled with the escalation in CCI, will in turn reduce the demand of investors for the apartment studios. On the other hand, the increase in stock market will also boost the demand of investors for apartment studios, which is a popular investment that can generate minimum required return for investors in Kaohsiung. 5. Conclusions This study compiles price indices for completed and pre- sale homes across four categories in the three biggest cities in Taiwan, i.e., Taipei, Taichung, and Kaohsiung, and examines the extent to which macroeconomic variables impact the price indices of completed homes in each cat- egory. The results are summarized as follows. ■Analysis of the completed home price index The real estate market price indices in Taiwan have shown different trends. Although the luxury home price index in Taipei fluctuates significantly, its long-term trends indicate that prices remain within a certain range, suggest- ing that the government periodically implements policies to control housing market prices. Due to the increasing population of people moving to Taipei, the demand for suites has also increased, which in turn drove up the de- mand from investors, resulting in a rising trend in the suite price index in recent years. On the other hand, Taichung’s growing population has led to increased housing demand, resulting in a moderate growth trend in its housing price index. In contrast, Kaohsiung has seen fewer people mov- ing there for work compared with Taipei and Taichung, leading to lower demand for suites and subsequently a more stable suite price index that stays around the aver- age value. ■Analysis of the pre-sale home price index Between 2015 and 2018, the government implemented policies to regulate housing prices, causing a 10% drop in prices and transaction volumes. In 2020, the pandemic prompted the government to lower the home purchase interest rate to 1.35%. At the end of 2020, the central bank implemented selective credit control policies to curb housing prices, increasing the cost of home purchases for the public and reducing the leverage used in real estate investments to curb speculative behavior. After the regula- tory period from 2015 to 2018, the housing market began to recover in 2019. The pre-sale luxury home indices in Taipei and Taichung reached their peak in the first quarter of 2020. In March 2020, the pandemic led to a drop in the home purchase interest rate to 1.35%, combined with rising construction costs, resulting in an upward trend in Taipei’s second-class and suite pre-sale home indices in the second quarter of 2020. The first-class pre-sale home indices in Taipei and Taichung started to rise after the third quarter of 2020. Due to the Kaohsiung government’s pro- motion of the Qiaotou Science Park and TSMC’s invest- ment announcement of a new Kaohsiung plant, Kaohsi- ung’s first-class pre-sale home index began to rise early in the second quarter of 2020. To suppress housing prices, the government introduced four rounds of credit control and the Housing and Land Tax 2.0 policies starting to take effect from the end of 2020. Taichung’s luxury home index began to decline in the first quarter of 2021, while Tai- pei’s first and second-class home indices saw a significant downward trend in the third quarter of 2021. Kaohsiung’s pre-sale home market however remained largely unaf- fected. ■Analysis of the impact of macroeconomic variables on completed home indices by category In Taipei, high income homebuyers for the luxury homes are not affected by economic changes. Middle to high income homebuyers are more interested in buying higher quality housing when the market interest rate is low. Young people and average wage earners have an ur- gent demand for renting suites irrespective of economic conditions. Despite the rising construction costs driving up suite and home prices in recent years, the strong de- mand for suites has attracted more investors in buying suite apartments. When the market is active, capital flows into Taichung’s luxury home market, indicating that high-income groups have both residential and investment demand for luxury homes. Middle- and high-income groups also use invest- ment gains to purchase higher-quality homes in Taichung. When prices rise and national income decreases, young people and average-income earners tend to buy houses below the average price. In recent years, Taichung has attracted many residents from other counties and cities, leading to a strong demand for suites from the migrating population, making them less affected by macroeconomic indicators. Compared to central and northern Taiwan, high-income groups in Kaohsiung view real estate as a rigid demand, making luxury home prices less affected by macroeco- nomic indicators. During active economic periods, young people and average-income earners, who have limited funds, use investment gains to purchase homes, making them more affected by the money supply compared to middle- and high-income groups who buy higher-quality homes with their investment profits. When the stock mar- ket is active and national income decreases, the high cost of entering the stock market drives investors to allocate funds to suite apartments. As the construction price in- dex increases, suite investors seek to raise rental income. However, southern homebuyers tend to be conservative, preferring to purchase second-class homes rather than renting suites or studio apartments, leading to decreased investor demand for suites and studio apartments.