# RECORD afc63bb9b7bf51c6364ea26f57de3de63b12d8bb # TITLE: Irish Property Price Estimation Using A Flexible Geo-spatial Smoothing Approach: What is the Impact of an Address? # YEAR: 2021.0 VENUE: Journal of real estate finance and economics DOI: 10.1007/s11146-022-09888-y # AUTHORS: Aoife K. Hurley; James Sweeney # PAGES: 39 CANDIDATE TABLE PAGES (1-based): [24] # ABSTRACT: Accurate and efficient valuation of property is of utmost importance in a variety of settings, such as when securing mortgage finance to purchase a property, or where residential property taxes are set as a percentage of a property’s resale value. Internationally, resale based property taxes are most common due to ease of implementation and the difficulty of establishing site values. In an Irish context, property valuations are currently based on comparison to recently sold neighbouring properties, however, this approach is limited by low property turnover. National property taxes based on property value, as opposed to site value, also act as a disincentive to improvement works due to the ensuing increased tax burden. In this article we develop a spatial hedonic regression model to separate the spatial and non-spatial contributions of property features to resale value. We mitigate the issue of low property turnover through geographic correlation, borrowing information across multiple property types and finishes. We investigate the impact of address mislabelling on predictive performance, where vendors erroneously supply a more affluent postcode, and evaluate the contribution of improvement works to increased values. Our flexible geo-spatial model outperforms all competitors across a number of different evaluation metrics, including the accuracy of both price prediction and associated uncertainty intervals. While our models are applied in an Irish context, the ability to accurately value properties in markets with low property turnover and to quantify the value contributions of specific property features has widespread application. The ability to separate spatial and non-spatial contributions to a property’s value also provides an avenue to site-value based property taxes. ===== PAGE 1 (score 5, numeric-paren 0, table-words 0, var-words 5) ===== / Published online: 5 May 2022 (2024) 68:355–393 https://doi.org/10.1007/s11146-022-09888-y Irish Property Price Estimation Using A Flexible Geo-spatial Smoothing Approach: What is the Impact of an Address? Aoife K. Hurley1 ·James Sweeney1 Accepted: 15 January 2022 © The Author(s) 2022 Abstract Accurate and efficient valuation of property is of utmost importance in a variety of settings, such as when securing mortgage finance to purchase a property, or where residential property taxes are set as a percentage of a property’s resale value. Internationally, resale based property taxes are most common due to ease of imple- mentation and the difficulty of establishing site values. In an Irish context, property valuations are currently based on comparison to recently sold neighbouring proper- ties, however, this approach is limited by low property turnover. National property taxes based on property value, as opposed to site value, also act as a disincentive to improvement works due to the ensuing increased tax burden. In this article we develop a spatial hedonic regression model to separate the spatial and non-spatial contributions of property features to resale value. We mitigate the issue of low prop- erty turnover through geographic correlation, borrowing information across multiple property types and finishes. We investigate the impact of address mislabelling on predictive performance, where vendors erroneously supply a more affluent post- code, and evaluate the contribution of improvement works to increased values. Our flexible geo-spatial model outperforms all competitors across a number of different evaluation metrics, including the accuracy of both price prediction and associated uncertainty intervals. While our models are applied in an Irish context, the ability to accurately value properties in markets with low property turnover and to quan- tify the value contributions of specific property features has widespread application. The ability to separate spatial and non-spatial contributions to a property’s value also provides an avenue to site-value based property taxes. Keywords House prices · Automatic valuation models · Spatial models · Generalised additive models · Hedonic regression · Non-linear smoothing  Aoife K. Hurley Aoife.Hurley@ul.ie 1 Department of Mathematics and Statistics, University of Limerick, Limerick, V94 T9PX, Ireland Journal of Real Estate Finance & Economics ===== PAGE 24 (score 20, numeric-paren 0, table-words 8, var-words 2) ===== A.K. Hurley, J. Sweeney Table 8 Results from decision tree and random forest approaches Model R2 RMSE Median Absolute Within 5% Within 10% Within 20% % Error of Price of Price of Price Tree 1 0.80 AC149,444 13.48% 20.22% 39.27% 67.49% Tree 2 0.81 AC148,977 13.32% 20.56% 39.13% 68.33% Forest 1 0.87 AC122,937 9.54% 28.01% 51.71% 80.07% Forest 2 0.88 AC117,328 8.82% 30.34% 55.38% 83.53% We trained both decision tree and random forest models as an alternative. Results from the implementation of two decision trees and two random forests are presented in Table 8. ‘Tree 1’ is a decision tree that includes all variables apart from ‘Longi- tude’ and ‘Latitude’, ‘Tree 2’ is a decision tree that includes all variables including ‘Longitude’ and ‘Latitude’ and ‘Forest 1’ and ‘Forest 2’ follow the same convention. From Table 8, we see that the random forests are marginally more accurate at pre- dicting the mean (R2 and RMSE) than our GAM models from Table 6, however are less accurate in prediction interval coverage. Although we can compare the results of the random forests and GAM models, in taking a GAM approach we reap the benefit of being able to interpret our models and can assess the impacts of specific variables. The Impact of Address Mislabelling? The initial analysis used the given postcodes on the property listings. However, some properties had given the neighbouring, typically more affluent, postcode rather than their actual one. We assess the impact of postcode misspecification by correcting the postcodes and re-running our analysis. 16% of properties in the data set (833 proper- ties) gave the incorrect postcode in their advertisement. We investigate 15 postcode changes (outlined in Table 9), based on a cross tabulation of the postcodes, shown in Fig. 8, and prior knowledge of the Dublin property market. In Fig. 8, the white lines for D20 and D22 indicate no changes in postcode labeling was required. In the models ‘GAM 5’ and ‘GAM 6’ in Table 10, the postcode changes are modelled by adding 15 dummy variables to previous models, one for each postcode change. GAM 5 and GAM 6 are the GAM 3 and GAM 4 models respectively, with the postcodes corrected. Figure 9 presents the estimates for the 15 postcode changes. The labels can be read as ‘Given Postcode.Corrected Postcode’, e.g. “D11.D9” is a property listed as D11 but with correct postcode D9. The majority of the estimated coefficients are positive, indicating a premium in sales price in each case where giving a neighbouring, more affluent postcode. It is clear that “D11.D9” and “D16.D14” are different to the other changes estimated, with scalings less than 1. These properties have listed postcodes with a lower median price (less affluent) than their correct postcode. For properties incorrectly listing a more affluent postcode, there is evidence of an additional pre- mium being added to property values - for 13 out of the 16 postcode changes the mean scaling estimates are greater than 1. The estimates have larger confidence intervals than those seen previously in Fig. 6, as there are fewer transactions to estimate from. 378 ===== PAGE 25 (score 4, numeric-paren 0, table-words 1, var-words 2) ===== What is the Impact of an Address? Table 9 15 Postcode Changes Investigated Given Postcode Corrected Postcode Occurrences Dublin 5 Dublin 13 3 Dublin 6 Dublin 6W 12 Dublin 6W Dublin 12 3 Dublin 9 Dublin 3 3 Dublin 9 Dublin 17 10 Dublin 11 Dublin 9 7 Dublin 14 Dublin 16 19 Dublin 16 Dublin 14 5 North County Dublin Dublin 15 9 South County Dublin Dublin 4 27 South County Dublin Dublin 18 101 South County Dublin Dublin 24 32 West County Dublin Dublin 15 38 West County Dublin Dublin 22 24 West County Dublin Dublin 24 45 In Fig. 10 we present the estimated scalings associated with the corrected post- codes. If the underlying spatial surface captured all spatial structure in the data, the postcodes should be labels alone, with no obvious patterns. However we note there is substantial evidence of a strong bias in postcode effects. The postcodes D7, D8 and D9 have extremely significant impacts on price with these representing popular, up-and- coming areas reflected in scaling estimates much greater than 1. D10, D22 and D24 all have scalings substantially less than 1 which would suggest these areas are under- valued relative to their spatial location. This is perhaps due to negative connotations Fig. 8 Heatmap of the changes that occur in postcodes 379