Investigation of Criteria Used by Cypriot Real Estate Agents in the Decision-Making for the Valuation of Real Estate Market - Empirical Research Findings Dr. Odysseas E. Moschidis Assistant Professor, University of Macedonia Department of Marketing & Operations Management e-mail: fmos@uom.gr Dr. Efstratios S. Livanis University of Macedonia Department of Accounting & Finance e-mail: slivanis@uom.gr Abstract Dr. Ioannis T. Lazaridis Professor, University of Macedonia Department of Accounting & Finance 54006 Thessaloniki, Greece e-mail: lazarid@uom.gr This paper builds on data from a recent survey on the financial and organizational policy of real estate agents in Cyprus. For the data analysis we used the appropriate methods of classical statistics and multidimensional analysis, as Test Χ 2, Correspondence Analysis, Comparative Evaluation and Discriminant Factor Analysis. The results show that a high percentage of Cypriot real estate agents work for many years in the real estate market, have economic knowledge and evaluate real estate using and scientific methods. Moreover, the results of multidimensional analysis heighten the factors that affect intensely more or less the decision making about the value of real estate in Cyprus. Then, we investigate if there is diversification in the assessment of these factors in relationship with the degree of the realtors knowledge and in which factors there is diversification. JEL Classification: G30, C00 Keywords: real estate market, quantitative methods, Cyprus Α first draft of this paper was presented on the 2 nd International Conference in Accounting & Finance, Thessaloniki, 28 30 August, 2008. 1 Electronic copy available at: http://ssrn.com/abstract=1400666
Introduction The investment in real estate constitutes one of the oldest, but in the same time safest forms investment. Even if in the current period of time (October 2008) the international real estate market buckles under the international economical recession, the Cypriot real estate market during the period that we run the survey (spring of 2008) is still on a uphill course and it is possible that this will continue, even in slower rates compared to the previous months (Speech by the Cypriot Minister of Finance: The International Financial Cyprus Crisis - Effects on the Cypriot Economy, CIIM, 23/9/2008). The successful Cyprus recent integration to the European Union combined with growing expectations for the solution of the Cyprus Problem on behalf of the United Nation Organization contributes towards this direction. Another important factor for the degree of real estate development in Cyprus is the continuously growing activation of new multinational firms and bank organizations that raise the real estate demand, as well as the touristic development of the island. Of course, the international tourist industry downturn must be taken into account. Despite the real estate sector s importance for the Cypriot economy, there is not an official publication of real estate prices. However, in 2004 the BuySell Home Price Index was created and it is calculated by the company MAP S. Platis on behalf of BuySell Cyprus Real Estate Ltd. The Index is informed monthly and is supported from the sale and offer prices provided by customers of BuySell Cyprus Real Estate Ltd. The BuySell Home Price Index is published quarterly in the web page www.mapsplatis.com/research. The methodology of calculation of the Index is described in the article of S. Platis and M. Nerouppos titled Asking Price and Transaction-based Indices for the Cyprus Housing Market (Rebased) (available at www.mapsplatis.com/research). 2 Electronic copy available at: http://ssrn.com/abstract=1400666
Despite the international economic crisis, according to the last publication (third quarter of 2008,) of the BuySell Home Price Index, the real estate prices in Cyprus continued their ascendant course after a marginal decline that was observed in the second quarter of 2008. More specifically, the Index price on September 2008 was 145,23 units. This corresponds with 193.323 euro for the average price of residence in Cyprus. In general the real estate market in Cyprus this period presents a deceleration of building growth and a reduction in the transactions of real estate is recorded. However, Nicosia and Lemesos have not been influenced yet in the volume of sales to a large extent compared to the other Cypriot cities. The research team of Professor Lazaridis, constructed a questionnaire to examine the real estate market in Cyprus because we consider very interesting to examine the character of the financial and organizational policy followed by realtors in Cyprus. The questionnaires was distributed by the research team to a representative sample of 115 real estate offices in Cyprus with the help from students of the Department of Accounting and Finance, University of Macedonia, in spring 2008. The total number of real estate offices that answered our questionnaire was 70, fact that puts the correspondence rate to 61%, which in our mind is considered especially satisfactory. In this paper first we analyze the data produced from this survey. Moreover, we investigate the factors that according to the real estate agents affect dominantly the price or real estate in Cyprus. Furthermore, we examine if these answers differ between those realtors who have economic knowledge and training (as it is demonstrated by the composition of real estate appreciation report) and those don t, as well as which factors differ more essentially. 3
The data analysis is based on the methodology of Comparative Evaluation which was presented by Moschidis (2006). This methodology consists of one side of the construction of a special contingency table which is named comparative evaluation table and on the other hand of statistical analysis of this table according to the method of correspondence analysis. Methodology In this paper the analysis is not based on a particular mathematical model. We use descriptive statistics. Also, we use the known Χ 2 criterion from classical statistics to examine the dependence of two categorical variables. To answer correctly the questions brought forward by the data analysis we used exploratory methods of Multidimensional Data Analysis that do not require conditions such as normality etc. to be met and that generally are not in effect in such data. In the spirit of these methods belongs the holistic view of the phenomenon at hand. Using the data at hand we want to investigate the interdependence and interaction between all of the elements of the phenomenon aiming to heighten the more dominant and essential tendencies in its structure. The X 2 independence test of classical statistics that is implemented for the independence examination of two variables falls short of the potential of the Multidimensional Data Analysis of correspondence. Using the Multidimensional Data Analysis of correspondence one complicated piece of information is pictured/ portrayed in factorial diagrams. This visualization of the results which is a basic characteristic of this method is not a random visualization, but the best it could be. Using the Multidimensional Data Analysis we process data which are recorded to multidimensional tables, that is with many lines and mostly with many columns. This multidimensional character of the data is in the point of interest of 4
these methods and also their most prominent advantage over the classical statistics methods. More specific, the Correspodence Analysis (Analyse Factorielle des Correspodances, (AFC), in French) was created for the statistical process of contingency tables. The first foundational mathematical approach of this method is accredited to Hirschfeld (1935), Guttman (1941) and Hayashi (1956). Nishisato (1980) gave us a thorough historical presentation of the development of this method. Yet professor J.P. Benzecri in the early 60s was the one to add splendor to this method and also to name it AFC. In his monumental work, L analyse des donnes (1973), this method is presented geometrically. This geometrical approach is the prominent feature of the French School. Still, which is the aim of processing a contingency table using A.F.C.? Three are the main points in the analysis of a table like this: The closeness of the line points The closeness of the column points The relationship between the line and column points The results in these questions by using the AFC come in a visible fashion/ way, usually in a two-dimensional picture (in the first factorial level) best and known clearness (known rate of interpreted inertia). Many times other images are used supplementary (second, third, etc factorial levels). Often correspondence analysis seems to be a combination of the independence test X 2 and the Factor Analysis. There are fundamental differences in these two methods. The 5
statistical X 2 does not show the structure of dependence, especially in large tables where the variables have many categories while the correspondence analysis by the use of graphics depicts the relationship of variable categories. Furthermore, this method is applied to quality variables (in contrast with factor analysis which is applied to quantity variables) and it is not based on any assumption (about variable distribution etc). Moreover, it is applied in any number of data and it enables us to determine the number of dimension in which the categories of lines and columns will be showed. The basic idea of this method is the conversion of the distances between the distributions of the lines or the columns, which is expressed with the metric X 2, in Euclidean point distances in a space of few dimensions. Specifically, the Euclidean distance of the point from the origin of the axes approaches the X 2 distance between the distribution of each category and the distribution of an average category. The Euclidean distance of two points approaches the X 2 distance between the lines (or the columns). To conclude in this short presentation of the correspondence analysis we must highlight that by using this method we benefit not only in quality but also in quantity. Regarding the quality benefits we can say that the statistical data after the analysis has been raised to its structural characteristics and regarding the quantity benefits that the whole information is summarised (L. Lebart, 2002). The method of correspondence analysis is used on a special contingency table which we named comparative evaluation table and is ideal for the comparative evaluation of the views of two or more groups (Moschidis, 2006). Finally, we used the known discriminant analysis method to investigate if and to what extend the 6
questions part the estimators (real estate agents) into two distinct groups of the ones who use empirical knowledge and the experts. Empirical Results In the table 1 we observe that almost the half corresponding real estate agents in our research work in real estate for above 20 years while a percentage 20% is activated in the area less than 5 years. It is characteristic that a particularly high percentage (75,7%) of Cypriot real estate agents has certain economic knowledge (table 2). Table 1: Number of years working in the real estate market Years Percentage (%) <=5 20,0 <=10 18,6 <=20 17,1 >20 44,3 50,00% 40,00% Number of years working in the real estate market 44,30% 30,00% 20,00% 20,00% 18,60% 17,10% 10,00% 0,00% <=5 <=10 <=20 >20 Years Figure 1: Number of years working in the real estate market 7
Table 2: Economic knowledge Answer Percentage (%) Yes 75,7 No 24,3 Moreover the majority (tables 3 and 4) of Cypriot real estate agents think that it is necessary to establish in Cyprus graduate and postgraduate programs of study in Real Estate. Also most of them encourage the specialization on valuation. A high percentage (81,4%) considers that it is advisable to publish indicators on real estate prices in the Cypriot market by an official authority. Table 3: Need of establishing separate graduate programs of study in Real Estate Answer Percentage (%) Yes 88,6 No 11,4 Table 4: Need of establishing separate postgraduate programs of study in Real Estate Answer Percentage (%) Yes 80,0 No 20,0 Table 5: Encouragement of specialisation in Real Estate and valuation Answer Percentage (%) Yes 94,3 No 5,7 8
Table 6: Need of publication of indicators on the prices of real estates in the Cypriot market by an official authority Answer Yes No Percentage (%) 81,4 18,6 Do you think it is advisable to publish indicators on real estate prices in the Cypriot market by an official authority? 18,60% 81,40% Yes No Figure 2: Need of publication of indicators on the prices of real estates in the Cypriot market by an official authority Moreover the 69,6% of Cypriot real estate agents evaluates real estate. Almost half (58,2%) of them writes real estate appreciation report and about 50% uses certain standardised form for this real estate appreciation report. Table 7: Valuation of real estates Answer Yes No Percentage (%) 69,6 30,4 Table 8: Writing a real estate appreciation report Answer Yes No Percentage (%) 58,2 41,8 9
Table 9: Standardised form for the real estate appreciation report Answer Percentage (%) Yes 57,1 No 42,9 With regard to the methods of real estate valuation the majority of Cypriot real estate agents use the comparative method (71,83%). As second method they use the replacement method (43,66%) and the income approach, the replacement cost method and the residual method follow. Table 10: Methods of real estate valuation Method Percentage (%) Comparative method 71,83 Residual method 22,54 Replacement cost method 30,99 Replacement method 43,66 Income approach 35,21 80,00% 70,00% 60,00% 50,00% 40,00% 30,00% 20,00% 10,00% 0,00% Methods of real estate valuation 71,83% Comparative method 22,54% Residual method 30,99% Replacement cost method 43,66% Replacement method 35,21% Income approach Figure 3: Methods of real estate valuation 10
It is encouraging that a high percentage of Cypriot real estate agents keeps record of the prices of real estate that have been rented or bought. This creates a base for the official publication of indicators on the prices of real estate in Cypriot market. Table 11: Record of the prices of real estate that have been rented or bought Answer Percentage (%) Yes 87,1 No 12,9 Also, 78,6% of Cypriot real estate agents collaborate with banks for loans with more favourable terms for their customers. Moreover, a particularly high percentage knows leasing and suggests this form of financing to their customers. Table 12: Collaboration with bank Answer Percentage (%) Yes 78,6 No 21,4 Table 13: Knowledge of leasing Answer Percentage (%) Yes 84,1 No 15,9 Table 14: Suggest leasing to their customers Answer Percentage (%) Yes 91,4 No 8,6 Regarding the use of internet by Cypriot real estate agents to promote their services we observe that not very high the percentage (60%) has web page in the internet. 11
Half of them (45,8%) manage their web page and their electronic mail. However, remarkable is that the 76,8% knows the free service Google Earth with which anyone can place real estate information. Table 15: Existence of web page Answer Percentage (%) Yes 60,0 No 40,0 Table 16: Personal management of web page and e-mails Answer Percentage (%) Yes 45,8 No 54,2 Table 17: Knowledge of Google Earth Answer Percentage (%) Yes 76,8 No 23,2 Exploratory Data Analysis By the following analysis we try to investigate the effect of each one of the 13 factors, mentioned below, in the pricing of real estate according to the realtors in Cyprus. The 13 factors in the relative classification are as follows: Ε1: The development prospect of a city Ε2: The infrastructure networks and the service supply Ε3: The urban design of the city and the urban restrictions Ε4: The size and the shape of the city blocks 12
Ε5: The building regulations Ε6: The street and pavement width and layout Ε7: The allowed height of buildings Ε8: The integration of new areas in the urban structure in the near future Ε9: The demographics of the city s population Ε10: The occupations of the city s population Ε11: The size of the earthquake risk of the area Ε12: The form and the size of environment pollution Ε13: The tourist development of the city First Analysis (Evaluation of the 13 factors) In order to rate the 13 factors that affect the real estate prices we created a evaluation contingency table, setting the 13 factors as lines and the 5 degrees of the evaluation Likert scale (intensely less, less, mediocrely, more, intensely more) as columns. For example, the cell in line 3 and column 4 shows how many realtors from the total 70 give to the third question grade (Moschidis 2003, 2006). Firstly, we examined by using X 2 test if there is indeed dependence of the degrees of the scale and the 13 factors at 5% importance level. The results showed intense dependence. Next we want to note the intense diversification in the evaluation of the factors. The before mentioned table process with the factor analysis of correspondence method will give answers. 13
Correspodence analysis results a) Eigenvalues table Table 18: Eigenvalues table Total Inertia 0,14385 AXIS INERTIA %EXPLANATION TOTAL EIGENVALUES HISTOGRAM 1 0,1190865 82,78 82,78 ***************************************** 2 0,0163786 11,39 94,17 **** 3 0,0072717 5,05 99,22 ** 4 0,0011149 0,78 100,00 * The first factorial level interprets by 94.17% the phenomenon at hand. To be more specific, the first factorial axis interprets 82.78% and the second the 11.89%. The first factorial axis rates the factors according to their evaluation, and since its interpretation percentage is very high, the conclusions reached by using the first factorial axis are very important. b) First factorial axis interpretation In creating the first factorial axis mostly contribute the factors Ε1, Ε10, Ε11, Ε13 (CTR 123, 122, 325, 113 respectively). Among those factors Ε13 and Ε1 are the ones that differentiate more intense than the others regarding the positive influence on real estate prices, while factors Ε11 and Ε10 affect less the pricing of real estate. 14
Figure 4: The 1 st factorial axis c) Second factorial axis interpretation The second factorial axis matches the factors that were ranked with intensely less or intensely more (Ε1, Ε10, Ε11, Ε13) with the criteria that got the mediate values. The remaining factors had mostly mediate values. Figure 5: The 2 nd factorial axis d) First factorial level interpretation The first factorial level composes and explains at the high percentage of 91.17% all conclusions from the first and second factorial axis mentioned above. To sum up, we can add that by using the factor analysis of the evaluation table we conclude that the factors that affect mostly the real estate price are the E13 and E1 according to the realtors views, while factors E11 and E10 have less influence. 15
Figure 6: The 1 st factorial level Second Analysis (discriminant analysis the apriori and aposteriori groups) One of the questions in the questionnaire concerned if the realtors write a real estate appreciation report or not. This question divides in advance the sample of real estate agents in two (a priori) groups, in group A consisting of 29 realtors that do not write the before mentioned report and group B consisting of 41 realtors that do. By using the discriminant analysis method we examine if the real estate agents answers to the 13 factors constitute aposteriori groups that relate to the above A and B groups. The 13 factor by the way they were answered divide the realtors into two groups (a posteriori groups) GR1 and GR2 in such a way that the a priori group Α coincides with group GR1 by 72.41% and group Β with GR2 by 65.85%. In other words with 16
the help of the discriminant analysis we concluded that there is relative diversification between group A and B regarding the way they rate the 13 factors. Third Analysis (Comparative Evaluation of the a priori group A and B) Using the comparative evaluation (Moschidis, 2006) of groups A and B we will try to find the factors by which the above mentioned groups are differentiated. When investigating the differentiations between the groups A and B regarding the hierarchy of the 13 factors the given interpretation rises to the high percentage of 87.28%. The first factorial axis interprets 76.32% and the second 87.28%. Table 19: Eigenvalues table Total Inertia 0,15764 AXIS INERTIA %EXPLANATION TOTAL EIGENVALUES HISTOGRAM 1 0,1203038 76,32 76,32 ***************************************** 2 0,0172883 10,97 87,28 **** 3 0,0084992 5,39 92,67 ** 4 0,0050438 3,20 95,87 * 5 0,0022564 1,43 97,31 * 6 0,0018426 1,17 98,47 * 7 0,0016060 1,02 99,49 * 8 0,0007748 0,49 99,98 * 9 0,0000238 0,02 100,00 * 17
Observing and evaluating the indexes CTR, COR and the coordinates of the points in the first factorial level we reach the conclusion that the diversification of the two groups is noted regarding the rating of factors which influence less to medium the markets participants in real estate appreciation. In particular, the realtors that write appreciation report mark more negative question E11 than those who don t write a report. Furthermore, factors E10 and E12 are graded with less mostly by those writing a real estate appreciation report. Those who do not write the report give the same importance to the evaluation degree intensely less and less, which is not the case with those who write the report, since as mentioned above they distinguish the intensely less concerning the factor E11 from the less concerning the factors E10 and E12. Moreover, factor E9 is ranked medium mostly by the realtors that do not write an appreciation report, while the same implies for those who do write the report concerning factor E6. Regarding the evaluation degrees more there is consensus concerning factors E7 and E3, and also in the high degree ( intensely more ) there is a unanimity of view concerning factors E1 and E13. 18
Figure 7: The 1 st factorial level Conclusions A sector of particular importance for the Cypriot economy is the real estate. According to the results of our research in Cyprus, real estate agents work for many years in this sector and have certain basic economic knowledge. They consider essential the official publication of indicators on the prices of real estate and encourage the creation of graduate and postgraduate programs of study of Real Estate in Cyprus. From the methods of real estate valuation they use mainly the comparative method followed by the replacement method. It is encouraging that the majority of Cypriot real estate agents keep record of the prices of real estate that have been rented or bought. 19
Moreover, by using the three method of analysis presented before, at first we investigated the degree in which the 13 factors in our questionnaire affect the real estate appreciation in Cyprus. According to our results, from the above factors the one that plays important role in real estate evaluations are the development prospects of a city and especially the tourist development. On the other hand, less influential are the business activity of the city residents and the size of earthquake risk in the area. Furthermore, using the discriminant analysis we concluded that there is diversification in the factors in real estate evaluation process by realtors if they write real estate appreciation report or not. This diversification focuses mostly on the degree the real estate agents weight the factors that have medium or less affect on the real estate value. An interesting new angle over the matter that we will examine in another paper is to combine the answers of the Cypriot realtor s agents with the price of real estate in Cyprus. References Floropoulos J., Moschidis O., Corporate Environmental Disclosures in Annual Reports, Accounting and Finance in transition, Series in Accounting and Finance, (vol. 11. 2005) Greenacre M., Correspodence Analysis in Practise, Academic Press, London (1993) John P., Van de Geer., Multivariate Analysis of Categorical Data: Theory, Sage, California (1993). Lazaridis I., Livanis E.,, Real Estate in Cyprus: An empirical investigation, Working Paper (2008) 20
Lebart L., Morineau A., Piron M., Statistique Exploratoire Multidimensionnelle, Dunod, (Paris, 2000). Moschidis O., A proposition for coding scales of evaluation, 16th Panhellinic Congress in Statistic (2003) Moschidis O., Lazaridis I., Livanis E., Comparative Investigation of the Dividend Policy of Cypriot Firms traded in Cyprus Stock Exchange Using a New Methodology of Correspondence Analysis, Investment Research and Analysis Journal, (vol. 1, n 1, 2006) Moschidis O., Contribution to comparative survey of multidimensional scales with the methods of multivariate analysis, PhD Disertation, (University of Macedonia, 2003) Moschidis O., Proposal Of Comparative Evaluation With The Correspodence Analysis: Comparative Evaluation Of Degree Of Satisfaction Of Patients And Doctors, «SPOUDAI», (Vol. 56, No 3, 2006) Platis S., & Nerouppos M., Asking Price and Transaction based Indices for the Cyprus Housing Market (Rebased), www.mapsplatis.com/research, (2005) Appendix Questionnaire 1. Years of working in real estate: <=5, <=10, <=20, >20 21
2. Do you have any economic knowledge? ; YES NO 3. Do you think that in Cyprus there should be established separate graduate programs of study in real estate? YES NO 4. Do you think that in Cyprus there should be established separate postgraduate programs of study in real estate? YES NO 5. Would you encourage someone to specialize on real estate and real estate valuation? YES NO 6. Do you think it is advisable to publish indicators on real estate prices in the Cypriot market by an official authority? YES NO 7. Do you evaluate real estate? YES NO 8. Do you write a real estate appreciation report? YES NO 8a. If YES do you use a standardised form? YES NO 9. Which method(s) of real estate valuation do you use? 1. Comparative method 2. Residual method 3. Replacement cost method 4. Replacement method 5. Income approach 22
10. Which of the below factors in your experience affect real estate prices? intensely less less mediocrely more intensely more The development prospect of a city The infrastructure networks and the service supply The urban design of the city and the urban restrictions The size and the shape of the city blocks The building regulations The street and pavement width and layout The allowed height of buildings The integration of new areas in the urban structure in the near future The demographics of the city s population 23
The occupations of the city s population The size of the earthquake risk of the area The form and the size of environment pollution The tourist development of the city 11. Do you keep record of the prices of the real estate that have been rented og bought? YES NO 12. Do you collaborate with (a) bank(s) to grant loans for your customers on more favourable terms? YES NO 13. Have you got any knowledge of the real estate leasing? YES NO 13a. If YES, do you suggest leasing to your customers? YES NO 14. Do you have a web page in the internet? YES NO 15. Do you manage on your own your web page and your electronic mail in general? YES NO 16. Do you know that by using free Google service you can place on the map information about your client s real estate? YES NO 24