1# RECORD 78763be000cac6c28c0629b4c2e5fffb2f7f7fde2# TITLE: Factors Affecting Spatial Autocorrelation in Residential Property Prices3# YEAR: 2022.0 VENUE: Land DOI: 10.3390/land110609314# AUTHORS: D. Lo; Wing-Chau Kwong; S. Wong; M. McCord; M. Haran5# PAGES: 17 CANDIDATE TABLE PAGES (1-based): [13]6# ABSTRACT: Within housing literature, the presence of spatial autocorrelation (S.A.) in housing prices is typically examined horizontally in a two-dimensional setting. However, in the context of apartment buildings, there is also a vertical component of S.A. for housing units located on different floor levels. This paper therefore explores the determinants of both horizontal and vertical S.A. within residential property prices. First, we posit that S.A. in housing prices is a consequence of the price discovery process of real estate, in which property traders acquire price information from recent market transactions (i.e., comparables) to value a subject property. Furthermore, we contend that the extent to which property traders rely on comparables to determine housing prices is governed by the liquidity and volatility conditions of the market, which in turn affects the magnitude of the S.A. By developing and testing several spatial autoregressive hedonic models using open market transaction data for the Hong Kong residential property market, we find that market liquidity tends to increase both vertical and horizontal S.A., whilst market volatility is more prone to increase vertical S.A. but depress horizontal S.A.789===== PAGE 1 (score 2, numeric-paren 0, table-words 0, var-words 2) =====1011 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33Factors Affecting Spatial Autocorrelation in Residential Property Prices34Lo, D., Chau, K. W., Wong, S. K., McCord, M., & Haran, M. (2022). Factors Affecting Spatial Autocorrelation in35Residential Property Prices. Land, 11(6), 1-16. Article 931. https://doi.org/10.3390/land1106093136Link to publication record in Ulster University Research Portal37Published in:38Land39Publication Status:40Published (in print/issue): 17/06/202241DOI:4210.3390/land1106093143Document Version44Publisher's PDF, also known as Version of record45Document Licence:46CC BY47For Author Accepted Manuscripts (AAM) published under Ulster University's Rights Retention Policy for Scholarly Works48(RRPSW)49When citing an AAM published under Ulster University's RRPSW please use the following citation structure:50Author, A. A. (Year). Title of article. Journal Name, [Accepted Author Manuscript]. PURE Portal URL. Licensed under CC BY 4.0.51General rights52The copyright and moral rights to the output are retained by the output author(s), unless otherwise stated by the document licence.53Unless otherwise stated, users are permitted to download a copy of the output for personal study or non-commercial research and are54permitted to freely distribute the URL of the output. They are not permitted to alter, reproduce, distribute or make any commercial use of the55output without obtaining the permission of the author(s).56If the document is licenced under Creative Commons, the rights of users of the documents can be found at57https://creativecommons.org/share-your-work/cclicenses/.58Take down policy59The Research Portal is Ulster University's institutional repository that provides access to Ulster's research outputs. Every effort has been60made to ensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in61the Research Portal that you believe breaches copyright or violates any law, please contact pure-support@ulster.ac.uk62Download date: 20/10/2026636465===== PAGE 13 (score 79, numeric-paren 66, table-words 4, var-words 5) =====6667Land 2022, 11, 9316812 of 1669prices. In comparison with the results of Equation (3), we observed no material difference70for the coefficients of the other hedonic variables.71Table 3. Regression Results.72Variables73Equation (2)74Equation (3)75Equation (7)76Equation (8)77Equation (9)78WVPt−m79-800.016781(0.0000)820.009183(0.0004)840.010485(0.0000)860.011287(0.0014)88WHPt−m89-900.018491(0.0000)920.014193(0.0000)940.026495(0.0000)960.023097(0.0000)98L ∗WVPt−m99-100-1011.74 × 10−5102(0.0001)103-1042.18 × 10−5105(0.0000)106L ∗WHPt−m107-108-1091.76 × 10−5110(0.0000)111-1121.54 × 10−5113(0.0000)114V ∗WVPt−m115-116-117-1180.3883119(0.0000)1200.4192121(0.0000)122V ∗WHPt−m123-124-125-126−0.1839127(0.0000)128−0.1877129(0.0000)130Constant1310.2045132(0.0000)1330.1677134(0.0000)1350.1704136(0.0000)1370.1675138(0.0000)1390.1698140(0.0000)141CBD142−6.57 × 10−5143(0.0000)144−5.98 × 10−5145(0.0000)146−5.80 × 10−5147(0.0000)148−5.96 × 10−5149(0.0000)150−5.78 × 10−5151(0.0000)152Age153−0.0305154(0.0000)155−0.0307156(0.0000)157−0.0307158(0.0000)159−0.0306160(0.0000)161−0.0306162(0.0000)163Age21640.0003165(0.0000)1660.0003167(0.0000)1680.0003169(0.0000)1700.0003171(0.0000)1720.0003173(0.0000)174Area1750.0027176(0.0000)1770.0027178(0.0000)1790.0027180(0.0000)1810.0027182(0.0000)1830.0027184(0.0000)185Area2186−4.88 × 10−7187(0.0000)188−4.85 × 10−7189(0.0000)190−4.86 × 10−7191(0.0000)192−4.86 × 10−7193(0.0000)194−4.86 × 10−7195(0.0000)196Floor1970.0066198(0.0000)1990.0069200(0.0000)2010.0068202(0.0000)2030.0070204(0.0000)2050.0069206(0.0000)207Floor22084.49 × 10−6209(0.0000)2101.67 × 10−6211(0.0000)2126.20 × 10−6213(0.0000)2143.05 × 10−6215(0.0000)2163.02 × 10−6217(0.0000)218Scale of Dev.2190.0014220(0.0000)2210.0014222(0.0000)2230.0014224(0.0000)2250.0014226(0.0000)2270.0014228(0.0000)229Scale of Dev.2230−1.59 × 10−6231(0.0000)232−1.59 × 10−6233(0.0000)234−1.60 × 10−6235(0.0000)236−1.59 × 10−6237(0.0000)238−1.60 × 10−6239(0.0000)240Included Obs241167,418242167,418243167,418244167,418245167,418246R22470.8963312480.8965942490.8966622500.8966382510.896711252Adjusted R22530.8961692540.8964172550.8964832560.8964592570.896531258Prob(F-Stat)2590.00002600.00002610.00002620.00002630.0000264Note: p-value in brackets.265Equation (7) expands the hedonic models further by including the liquidity factor (L)266in the spatial processes to test H(1) and H(2). As established above, a higher market267trading volume will increase vertical S.A., but the way in which it affects horizontal S.A.268is less apparent, depending on how traders make use of comparables. Our results show269that both θL,V and θL,H were positively signed and failed to reject H(2), but for H(1), the270results were significant at the 1% level. This finding infers that traders seem to employ271recently sold comparables found both within and outside the subject buildings to ascertain272new prices when the market is more liquid. This may also be a manifestation of the273relatively efficient nature of Hong Kong’s real estate market, where price information about274properties is easily accessible in terms of time and cost. The results for the coefficients275of other hedonic variables, as well as the overall explanatory power of the model, were276statistically satisfactory.277Equation (8), constructed to test H(3), differed from (6) in that the liquidity variable278was substituted by the volatility variable, V. The findings show that θV,V was positive279while θV,H was negative at the 1% level, in agreement with our expectation that when280281282===== PAGE 14 (score 4, numeric-paren 0, table-words 0, var-words 4) =====283284Land 2022, 11, 93128513 of 16286the market is volatile, traders are inclined to put heavier weight on comparables found287in the same building, and less on other comparables; hence, H(3) was not rejected. The288results not only confirm the role volatility plays in the formation of S.A., but also are289suggestive that housing market participants react to varying levels of market uncertainty290when undertaking property valuation differently. The results for other hedonic variables291were statistically satisfactory in terms of the saliency and statistical significance of their292respective coefficients.293As a robustness test, Equation (9) takes into consideration the effects of both liquidity294and volatility in explaining the variation of S.A. in housing prices. The results confirm our295aforementioned information search conjectures in relation to the formation of S.A., which296is evidenced by all coefficients displaying the expected signs and significant at the 1% level.297Last but not least, we contend that since spatial autocorrelation in housing prices is not298time-varying, our empirical results should remain statistically intact even if a different299dataset of the Hong Kong housing market with a different time period of investigation300was examined.3015. Concluding Remarks302This paper attempts to explore the determinants of spatial autocorrelation in housing303prices, which is a much under-researched topic within the existing housing literature.304Throughout the paper, we argue that S.A. in housing prices is a direct consequence of the305price determination process undertaken by housing market participants. To verify this,306we proposed three hypotheses, linking (1) market liquidity and (2) market volatility with307S.A. in housing prices in a three-dimensional setting. Vertical S.A. is defined as spatial308autocorrelation in housing prices for housing units located in the same building, whereas309horizontal S.A. is defined as the counterpart for housing units located in different buildings.310With a huge number of open market transaction data for Hong Kong, we established and311statistically confirmed with the use of several spatial autoregressive hedonic models that312market liquidity tends to increase both vertical and horizontal S.A. We also found that313market volatility tends to increase vertical S.A. and depress horizontal S.A. Hence, we314can confirm the abovementioned role that price discovery plays in the formation of, and315changes in, the level of S.A. Moreover, our observation that S.A. is partially dependent316on liquidity and volatility may signal that the constant S.A. assumption, which is widely317accepted in the literature, should warrant closer scrutiny.318The findings of this paper reveal several implications beyond the theoretical housing319literature, particularly in the areas of property valuation and the economic analysis of the320housing market. Traditional hedonic models usually consist of a long list of locational321variables to account for the geographical factors of real estate. If inappropriately designed,322the models could likely produce spurious results. Furthermore, as elucidated by studies323such as McMillen and Redfearn, 2010 [64], McCord et al., 2014 [65], and Bidanset et al.,3242019 [66], geographic information systems allied with spatial–statistical approaches can325better model the spatial nature of property price data, which take into consideration spatial326dependence and spatial heterogeneity and help account for vertical and horizontal inequity.327Against this backdrop, a more thorough conceptualization on the causes of S.A. in housing328prices can prompt the development of more parsimonious yet more econometrically ap-329pealing spatial hedonic models, which can make the mass appraisal of properties much330less computationally intensive and more reliable.331Further research could be undertaken to examine whether the liquidity and volatility332arguments put forward in this study are empirically valid using data from other interna-333tional property markets. For instance, regional housing markets within the U.K. have, over334the course of the past few years, been extremely volatile, both in terms of price and return,335due to political uncertainty arising from political shocks. It would therefore be worth336investigating whether and how spatial autocorrelation in housing prices have changed337with respect to political events in the context of price determination. In addition, the recent338outbreak of COVID-19 has had a tremendous impact on almost every housing market across339