Vol. 4 No (Mar. 2011) of location. We then measure the risk in real estate prices which are caused by it. Moreover, we apply our model t

Size: px

Start display at page:

Download "Vol. 4 No (Mar. 2011) of location. We then measure the risk in real estate prices which are caused by it. Moreover, we apply our model t"

まともあると
4 years ago
Views:

1 Vol. 4 No (Mar. 2011) of location. We then measure the risk in real estate prices which are caused by it. Moreover, we apply our model to develop a prototype system called Real Estate Valuation Maps. This system effectively use the real estate data on the Web. By clicking the pin at which the target real estate is located on the well-developed globe or map applications such as Google Earth and Maps, its price and risk are displayed Web Google Earth Maps Valuation and Risk Measurement of Real Estate Prices: Model and Its Application Hiroshi Ishima, 1 Akira Maeda 2 and Tomohiko Taniyama 3 This paper presents a model to evaluate the price and risk of real estate rigorously based on the financial theory and engineering. We accommodate the model to considering the characteristics of real estate which are crucial for the evaluation. As a background for this study, although real estate is no doubt the largest asset for individuals and firms, there are few data or tools available when they make a decision to buy or sell real estate. The situation is quite different from the one concerning financial assets such as stocks. Hence we are to present one solution in this study. Also we implement an empirical analysis on the Japanese condominium market to show the advantages of our model when compared to the existing ones. We obtained two findings. Firstly, there exist distortions in the Japanese condominium prices which are larger than we expected from the theory. Secondly, the prices vary according to its individuality 1. 2/ NEEDS Bloomberg Thomson Reuters CRSP The Center for Research in Security Prices Web Web API ROE PER PBR DCF 1 Graduate School of International Accounting, Chuo University 2 Graduate School of Energy Science, Kyoto University 3 Nomura Research Institute, Ltd. 1 c 2011 Information Processing Society of Japan

2 2 CAPM 1 HP 2 (1) (2) Web Google Earth Maps ROE = Return On Equity PER = Price Earnings Ratio PBR = Price Book-value Ratio DCF = Discounted Cash Flow CAPM = Capital Asset Pricing Model (1) (2) (3) 1 1 5) = ( k ) ( k ) (1) k 5) hedonic model Lancaster 6) Rosen 10) 2 (1) N H N i i i =1,...,N j j =1,...,n i

3 3 i n i n i i 1 N N H i j H K x (k) {H ; x (k) (k =1,...,K)} K =3 x (1) x (2) x (3) 2.1 (1) 1 (1) ε α Jensen α N x (l) l =1,...,N α := N l=1 x(l) β(l) x (l) i j l i = l 1 0 β (l) l N K H = x (l) β(l) + x (k) β(k) + ε l=1 k=1 (i =1,...,N; j =1,...,n i) (2) β (k) x (k) ε 0 N H 2.2 (1) Luenberger 8) 2 1 H Box-Cox Box-Cox 1) H (λ) = { H λ 1 λ log H (λ 0 ) (λ =0 ) (2) H (λ) = N l=1 x (l) β(l) + K k=1 x (k) β(k) + ε (i =1,...,N; j =1,...,n i) (4) λ λ =1 λ =0 (3)

4 4 λ Box-Cox fixed effects model 2 1 k β (k) ν (k) i (4) H (λ) = N l=1 ν i := (ν (1) i x (l) β(l) + K k=1 ( x (k) β (k) + ν (k) i ) + ε (i =1,...,N; j =1,...,n i) (5)...ν (k) i...ν (K) i ) 0 K G Box-Cox mixed effects model (5) mixed effects model random effects model 1 longitudinal data panel data Hsiao 4) Fitzmaurice 2) McCulloch 9) SAS MIXED Littell 7) REML; Restricted Maximum Likelihood BLUP Best Linear Unbiased Prediction Box-Cox λ Gurka 3) (5) G 3. (5) (4) N =8 1 (5) Box-Cox (4) Box-Cox 2 3 AIC Akaike Information Criterion AIC 2 1 Table 1 Definition of real estate classes by location and number of observations in each class. 1 SAS Statistical Analysis System SAS Institute Inc.

5 5 2 P 3 P Table 2 Estimation results by mixed effects model. Figures in parentheses are P values. Table 3 Estimation results by fixed effects model. Figures in parentheses are P values. 3 1 λ 1 λ 1 λ 0 λ 0.2 λ =

6 % (5) Box-Cox (4) Box-Cox 3 (5) (4) BLUP (5) (4) 99% SAS 99% VaR Value at Risk VaR 99% 99% VaR 99% 99% 99% % Fig. 1 Real estate price index based on mixed effects model (solid line). Dashed lines show upper and lower bounds of 99% confidence interval. 1 2 AIC

7 7 4 99% Table 4 Real estate price index based on mixed effects model. Figures in parentheses are the percentage differences between the real estate price indices and their upper and lower bounds with 99% confidence interval, respectively. 2 99% Fig. 2 Real estate price index based on fixed effects model (solid line). Dashed lines show upper and lower bounds of 99% confidence interval % 3 (5) Web Google Earth Maps GIS

8 8 5 99% Table 5 Real estate price index based on fixed effects model. Figures in parentheses are the percentage differences between the real estate price indices and their upper and lower bounds with 99% confidence interval, respectively. 3 Fig. 3 Flow chart of real estate valuation maps (1) XML Web API XML 2 (3) SAS Filename Web API (2) XML 1 (3) SAS MAP XML SAS (3) 2 (5) SAS (4) XML (5) XML + XML XML Google Earth Maps

Google Earth GE /Google Maps GM XML GE/GM KML XML GE/GM XML 4.

9 9 XML KML Keyhole Markup Language 4.2 XML XML Extensible Markup Language IT XML EDINET XML XBRL extensible Business Reporting Language XML XML Web API XML Google Earth GE /Google Maps GM XML GE/GM KML XML GE/GM XML 4.3 Google Earth/Google Maps KML XML GE/GM 4 5 Web API (a) GE Windows PC Mac iphone/ipod Touch/iPad (b) Web Windows PC Mac 3 4 Google Earth Fig. 4 Example on Google Earth: different colors are assigned according to its real estate standard score at pinned locations. (1) 4 5 (2)

10 VaR 99% VaR 95% 99% 95% (5) (6) standard score

standard scores and risks are displayed.

10 10 VaR 99% VaR 95% 99% 95% (5) (6) standard score T-score blue chip = Google Earth Fig. 5 Example on Google Earth: real estate information such as basic attributes, ask prices, theoretical prices, standard scores and risks are displayed. (3) 2 Box-Cox (5) BLUP (4) 99% 95% Google Chart Tools 2 AIC Web Google Earth Maps

11 11 3 (A) (B) (C) 1) Box, G.E.P. and Cox, D.R.: An Analysis of Transformations (with Discussion), Journal of the Royal Statistical Society: Series B, Vol.26, pp (1964). 2) Fitzmaurice, G.M., Laird, N.M. and Ware, J.H.: Applied Longitudinal Analysis, John Wiley & Sons, Inc. (2004). 3) Gurka, M.J., Edwards, L.J., Muller, K.E. and Kupper, L.L.: Extending the Box- Cox Transformation to the Linear Mixed Model, Journal of Royal Statistical Society: Series A, Vol.169, No.2, pp (2006). 4) Hsiao, C.: Analysis of Panel Data: 2nd Edition, Cambridge University Press (2003). 5) JAFEE 2009 pp (2009). 6) Lancaster, K.: A New Approach to Consumer Theory, Journal of Political Economy, Vol.74, pp (1966). 7) Littell, R.C., Milliken, G.A., Stroup, W.W., Wolfinger, R.D. and Schabenberber, O.: SAS for Mixed Models: 2nd Edition, SAS Publishing (2006). 8) Luenberger, D.G.: Investment Science, Oxford University Press (1997). (2002). 9) McCulloch, C.E., Searle, S.R. and Neuhaus, J.M.: Generalized, Linear, and Mixed Models: 2nd Edition, John Wiley & Sons (2008). 10) Rosen, S.: Hedonic Prices and Implicit Markets: Product Differentiation in Pure Competition, Journal of Political Economy, Vol.82, pp (1974). (1) 5) (5) 3 (i) (ii) (iii) 1 5) (37) [ ] H i,t = E t δ τ L i,t+τ b i,t+τ Mt+τ Z (i =1,...,N H ) (6) τ=0 H i,t i t E t t δ L i,t i t 1 b i,t R 1 K t i K M Z t+τ := ( u(c t+τ, Z t+τ ) / Z t+τ ) / ( u(ct, Z t) / C t ) R K 1 u C t t Z t R K 1 t N H u C t Z t 1 b i,t = b i i, t (7) (6) [ ] H i,t = b ie t δ τ L i,t+τ Mt+τ Z (8) τ=0 1 L i,t+τ i (8) H i,t = b i ˆπ t (9)

12 [ ˆπ t := E t τ=0 δτ L t+τ Mt+τ] Z i (9) t i H i,t i b i ˆπ t 1 L i,t t (8) H i,t = L

1 2 1 L i,t i t (9) (10) (1) 5) 2 1 2 linear pricing (1) 2 (1) (2) (4) (5) ( 22 8 30 ) (

12 12 [ ˆπ t := E t τ=0 δτ L t+τ Mt+τ] Z i (9) t i H i,t i b i ˆπ t 1 L i,t t (8) H i,t = L ib i ˆπ t (10) [ ˆπ t := E t τ=0 δτ Mt+τ] Z i (10) t i H i,t i L ib i ˆπ t (9) (10) (1) L i,t i t (9) (10) (1) 5) linear pricing (1) 2 (1) (2) (4) (5) ( ) ( ) ( ) FP SAS / & 2010 JAFEE FP MS Ph.D. Engineering-Economic Systems and Operations Research, Minor: Economics FP

Vol.8 No (July 2015) 2/ [3] stratification / 1 2 J-REIT 2 1 2 J-REIT % J-REIT J-REIT 6 J-REIT J-REIT 10 J-REIT *3 J-

Vol.8 No (July 2015) 2/ [3] stratification / *1 2 J-REIT *2 *1 *2 J-REIT % J-REIT J-REIT 6 J-REIT J-REIT 10 J-REIT *3 J- Vol.8 No.2 1 9 (July 2015) 1,a) 2 3 2012 1 5 2012 3 24, 2013 12 12 2 1 2 A Factor Model for Measuring Market Risk in Real Estate Investment Hiroshi Ishijima 1,a) Akira Maeda 2 Tomohiko Taniyama 3 Received: