Determinants of residential property valuation

Transcription

1 Determinants of residential property valuation Author: Ioana Cocos Coordinator: Prof. Univ. Dr. Ana-Maria Ciobanu Abstract: The aim of this thesis is to understand and know in depth the factors that cause changes in residential real estate prices. The paper is divided in two sections: a first introductory part, which includes a brief description of the specific elements of real estate market, its evolution over the analysed period , and the state of knowledge on the topic. The second part includes the case study itself, also structured in two stages. These are the empirical research realized first at national level using a simple regression, and subsequently analyzed six main counties for which we used a panel data.after we obtained the regressions we presented the results and their interpretation based on the outputs from E-views. In the end, we added final conclusions drawn from the study, after which we presented a set of recommendations and proposals for future studies in this area. Keywords: Residential Real Estate; Romania; Panel Data; mortgage; JEL Classifications: G3, C23, C33 1. Introduction The subject of this paper is a very important one and of high interest for national, European or international level. We proposed to develop some approaches previously used in the economic literature and to study fundamental issues such as: - Housing and mortgage market development in Romania;

2 - Development of a model fit to decide one of the most important indicators based on the information gathered from relevant real estate ads and offers, namely the sale price per square meter. These aspects allow us to achieve the main objective of this paper, namely to build a picture of the real estate market in Romania. Taking into consideration the importance of the topic and the extremely rapid developments in recent years, we tried using the most recent data available for the analysis performed. Thus, besides the use of classical references in paper, we used the electronic publications and databases (jstor, Science Direct, ideas RePEc, SSRN papers) which also represented very important sources. One of the most studied areas of research regarding the economic and financial market of real estate is about residential real estate pricing. Empirical research mainly focuses on identifying characteristics with the greatest influence on the sale price. In valuation and study of the real estate market, residential sales prices and rental value are generally analyzed by a regression model based on microeconomic theory. Based on previous studies we built a model, which in a reasonable way can show in what proportion the selling price per square meter is influenced by factors such as: population average income, mortgages, the interest on these loans, the exchange rate or inflation rate. Through this model we proposed to compare the results of the Romanian market with results in the literature related. 2. The theoretical (literature review) for the research field The real estate market in Romania had a very interesting trend in recent years, making some investors to move from a state of ecstasy (which result from the safety of spectacular returns) to one of agony as a result of state bankruptcy. After a decade of relative calm on the Romanian real estate market in the 90s, since 2002 and until 2008 was recorded a significant increase in prices. The value of residential buildings increased from 363 million euros in 2000 to 4.2 billion in 2008, and from this value detach individual buildings with a total value of 2.4 billion euros and the apartments built about 1.8 billion euros, according to the chart below.

3 Source: After 2008, however, the housing market was on a negative trend, and given that the Romanian economy will continue to record modest growth, it is hard to believe that in the next 6-12 months, this trend will undergo significant changes. Sursa: Nationally, in the T2.12, apartment prices fell by 0.1% compared to T1.12, and by 9% to T2.11. In Bucharest decline was 0.1% and 7.9% and in the rest of the country 0.1% and 10%. In recent quarters, since the INS began collecting data in Bucharest declared notaries prices fell by 31.5% and 23.5% in the rest of the country. For comparison, the datas from

4 imobiliare.ro show a 29% decrease in the price per square meter requested by apartment owners between T1.09 and T Previous research Sursa: Berglund (2007) conducted an empirical study with the objective of determining what changes affect housing prices. He used data from the Stockholm and Sydney to create a model that predicts changes in housing prices in the two cities. The findings suggest that the main determinants are nominal interest rates, income and housing offer us. It also investigated whether the use of financial indicators, such as capital market development has an impact on housing prices. These are explanatory variables in one of the models presented in this paper. It further presented a more complicated model with exogenous variables, but it failed to create a better overall prognosis despite the inclusion of explanatory variables. The simple model has exceeded its counterpart so advanced in terms of forecast housing prices in Stockholm. Sydney market, however, the result was more favorable for complex model, although the model just was not much weaker. In conclusion, the addition of new financial indicators regressions has some positive effects in this study. Egert and Mihaljek in 2007 studied the determinants of housing prices in eight transition economies of Central and Eastern Europe (CEE) and 19 OECD countries components (Organisation for Economic Cooperation and Development). The main question is whether the determinants of house price fundamentals such as GDP per capita, real

5 interest rates, the mortgage and demographic factors have played a role in the dynamics of house prices in CEE. The model used is based on standard variables used in the empirical literature, besides the main factors listed above, consider factors such as transforming specific transitional housing market and housing finance institutions in CEE, newly built better housing and increased demand more housing in this part of Europe by residents of other parts of Europe. In the empirical analysis we used a comparative approach, studying the determinants of price changes for the countries considered. The main econometric techniques used in this paper is the method of least squares - OLS. The existence of long-term relationships that connect housing prices to a set of explanatory variables is verified using termenide error correction specification derived from the error correction model. GDP per capita is very important and has a positive sign, which is expected in almost all regressions, indicating that changes in income are closely linked to positive change in housing prices. Coefficients of the real interest rate in most cases have the expected negative sign and are statistically significant, indicating that the decline in real interest rates is associated with increased house prices. The authors examined the role played in the dynamics of housing prices traditional factors listed above and the importance of factors specific transition. It was also analyzed how these factors affect the dynamics of housing prices in different groups of OECD and transition economies. The main result is that, overall, per capita GDP, interest rates, credit growth, demographic factors and indicators of institutional development of housing markets and housing finance are important determinants of house prices in CEE. 4.Empirical evidence 4.1.The database In the present study the analysis period spans on a period of approximately five years, from April 2008 to December 2012, the datas are monthly and totals 57 observations. Based on previous studies, we built a model which in a reasonable way can show to what extent the price per square meter depends on factors such as mortgage loans, gross income of the population, interest on mortgage loans, the annual percentage rate or the

6 exchange rate. Cercetarea exchange is carried out in Romania, both nationally monthly statistics, as well as at the 6 counties main (Bucharest, Timis, Brasov, Cluj, Constanta, Iasi), we chose to include in the study mainly because they share a number of features that are among the largest, most developed, and most productive by GDP per capita, the same but different enough to fiind an interesting comparison. Top counties from Romania by GDP per capita Count Gdp per capita in (eur) București Ilfov Timiș Argeș Brașov Cluj Constanța Sibiu Prahova Iași Source: We have collected the data regarding the price per square meter through online real estate market in Romania, represented by portal: imobiliare.ro. The numbers represent an overall average of the county, based on asking price by the site from which we removed duplicate offers.

7 Next, we collected information related to household average income, we found either on the official website of the National Institute of Statistics (insse.ro) or we have been given by representatives of county departments from each county statistics analyzed. Inflation rate, exchange rate, mortgages and interest on foreign currency mortgage loans we took from monthly reports published by bnro.ro Methodology The model used is relatively new, since we introduced in our regressions new variables or we combined determinants from different studies, the most relevant being: Egert and Mihaljek (2007), Berglund (2007 ) and Marco (2008). For all data series used, except inflation and interest rates, we calculated variations from month to month, which used in the models. In our empirical analysis we studied initially nationally correlation between the price per square meter and independent variables such as foreign currency mortgage, average gross monthly interest rate on mortgage loans in foreign currency over 10 years, the annual percentage rate mortgages and foreign exchange, based on the equation below: P = β 0 + β 1 * credit ip + β 2 castig brut + β 3 * r d10 + β 4 * DAE + β 5 * curs sc + Ԑ t, where: P sale price per square meter, β 0 constant, β 1 β 5 - independent variables coefficients, credit ip foreign mortgage variation, castig brut average income variation, r d10 interest rate for mortgages loaned for more than 10 years, DAE annual interest rate, curs sc exchange rate variation, Ԑ t - the error term.

8 In the second part, at the county level, we used a panel data set that used as independent variables: inflation, average gross and total mortgages. Regression equation has the following form: P = β 0 + β 1 credit ip + β 2 castig br + β 3 r i + Ԑ t, unde: P price per square meter, credit ip - foreign mortgage variation, câstig br - average income variation, r i inflation rate Results We noticed that in the studied articles, after the regressions were realized, it was established a negative relationship between real interest rates and housing prices, which is also present in our research. This indicates that lowering interest rates are associated with increasing selling prices. In addition to this we found that between mortgage and price are strong positive correlation, so as the number of these loans will be higher, and the price will increase. These results were somewhat predictable because these three variables are closely related, and as long as the mortgage interest rate decreases, the immediate effect will be an increase in demand for these loans, and as a result will increase the price of housing. From the output obtained we can see that the most significant factor affecting price is the exchange rate, which increased by one percentage point will influence the price increase by 0.48 percent. Devaluation against the euro negatively affects the population, because the vast majority receive wages in USD, wages and salaries that are not related to or influenced by the euro. Instead, the price utilities and many other goods, products and services increase as the euro. This is the case of real estate prices, which since the introduction of the euro, prices for housing and land was and is still expressed in euro. So even without any significant movement in the housing market in terms of supply and demand for housing, only in the currency in domestic housing prices rose. Also other studies have found a positive relationship between income and housing prices, suggesting that an increase in income encourages property owners to

9 take advantage of increased purchasing power of tenants and therefore, increase both housing prices and the rents. This however, is not apparent from our model, the coefficient of average gross earnings are negative. 5.Conclusions In this paper, "Determinants in residential property valuation" we intended to highlight the main factors that will influence the selling price per square meter for residential buildings in Romania. In preparing this paper we have considered a diverse literature including studies conducted in both Romania and other countries in Europe and internationally. Structure proposed in this thesis work was to show the influence of factors such as population average gross rate mortgage, the interest on these loans, exchange rate or inflation rate on the selling price. The model used is relatively new, since our regressions we introduced new variables to what I studied in the literature or have combined key determinants of different studies in the same model. In presenting information on the sale price and foreign currency mortgages, interest rates, exchange rates, inflation or average gross population, we wanted to give an overview of the residential real estate sector in Romania. Econometric analysis of the determinants of the sale price was the use of a linear regression model proved to be valid, the coefficients significant in statistical terms. The most relevant independent variables were found to be foreign exchange and mortgage loans. Following regression, we establish a negative relationship between real interest rates and housing prices, which indicates that lowering interest rates is associated with increased selling prices, which are inversely proportional. In addition to this we found that between mortgage and price are strong positive correlation, so as the number of these loans will be higher, and the price will increase. These results were somehow predictable because these three variables are closely related, and as long as

10 the mortgage interest rate decreases, the immediate effect will be an increase in demand for these loans, and as a result will increase the price of housing Proposals for future studies: There are many aspects of this project that could be developed further in future studies. Research is carried out in order to identify factors that influence residential real estate prices, we hope that further research will be conducted more in-depth to explore this area. An interesting study would be to analyze the data on rental market to see if the factors relevant to the selling prices used in the present model are relevantly similar for rental prices. Any discrepancies between the two sets of prices could be explored in an attempt to model the difference between selling and renting.