21Aug 2023 A Proposal for a Residential Housing Price Index in Cyprus Through Analysis of Transaction-Based Data and Comparison With Existing Indices The process of constructing a housing price index based on transactional data in Cyprus has yet to be studied by the scientific community. However, it must be noted that three articles related to this work were identified, but all are based on non-actual data. Hence, it was deemed appropriate to attempt researching the topic in order to provide up-todate information and summarize the conclusions drawn from such a study when different statistical models with actual data were put to test. The closest studies regarding the construction of property price indices in Cyprus were published by Karagiannakis et al. (2016) and by Platis and Nerouppos (2004), both using hedonic regression. However, in the first study their model was run with asking prices collected from the adverts section in newspapers, while in the second study they used asking prices and actual sale prices (although “actual sale price” is not further defined) collected by a real estate agency. Moreover, an article by Pashardes and Savva (2009) also utilized asking prices from newspaper advertisements with a hedonic regression model and they are introduced additional macroeconomic indices in the model such as GDP per capita, construction cost, population growth, and so on to explain the effect on housing prices. In addition, an article was found by Panayiotou et al. (1999), for which they designed a regression model to appraise properties (flats and houses) using an empirical study, conducted only in a specific Municipality of Nicosia, Strovolos, supplementing the study with property attributes by interviewing Department of Lands and Surveys (DLS) officers, owners, and so on. Other studies, such as an article from Kyriakidis (2015), analyzed housing submarkets of Cyprus on a theoretical level by reviewing relevant literature, and Michail and Thucydides (2019) utilized published macroeconomic indices to identify the impact of foreign buyers on house prices. Meanwhile, in Cyprus there are three residential property price indices publicly available from credible institutions, which derive from the Central Bank of Cyprus (CBC), the Cyprus Statistical Service (CYSTAT), and the Royal Institute of Chartered Surveyors (RICS), currently in association with KPMG. All three indices use different sources and types of data and methodologies or variations of methodologies. The CBC index and RICS index use historic appraisal-based data and survey data from valuers’ opinions, respectively. The CYSTAT index uses DLS data (the same data source as in this study); however, the database size and database cleaning procedures are not specified. Nevertheless, most of the existing research and the published indices by CBC, CYSTAT, and RICS appear to give emphasis to the construction of the models, rather than the database. Therefore, it was established that there was an opportunity to utilize transaction- based data. Furthermore, it was unknown what the results of an index using actual/ official transactions would be and how the index would behave with a database structure utilizing DLS available property parameters. Therefore, the aim of this study was to exploit the maximum volume of actual transactions by testing several methods and models and comparing results with existing indices. Ultimately, with this study, authors endeavor to contribute to the residential price index methodology that covers all districts of Cyprus. You can read the article here Tags: Publications, house price index, cyprus residential real estate market, standard and rolling dummy, hedonic regression methods, databases, transaction-based indices
