An analysis of housing credit program for urban hoausehold case study in HCMC Housing development bank(HDBANK)

INSTITUTE OF SOCIAL STUDIES THE HAGUE THE NETHERLANDS UNIVERSITY OF ECONOMICS HO CHI MINH CITY VIETNAM =======oOo======= VIETNAM- NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS AN ANALYSIS OF HOUSING CREDIT PROGRAM FOR · URBAN HOUSEHOLD CASE STUDY IN HCMC HOUSING DEVELOPMENT BANK (HDBANK) BQ GIAO D~C E>AO Tl;\0 TRUONG Dl;\1 H9C KINH TE TP.HCM THUVIEN • A THESIS PRESENTED BY DOHONGNGOC IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE DEGREE OF MASTER OF ARTS IN DEVELOPMENT ECONOMICS SUPERVISOR DR TRAN TIEN KHAI Ho Chi Minh City, February 2009 CERTIFICATION "I certify that the substance of this study has not already been submitted for any degree and is not being currently submitted for any other degree I certify that to the best of my knowledge any help received in preparing this thesis, and all sources used, have been acknowledged in this dissertation." Ho Chi Minh City, February 2009 Do HongNgoc i ACKNOWLEDGEMENT i would like to express deeply my appreciation and many thanks to the following individuals and organization helped me to finish this thesis Doctor Tran Tien Khai, the academic supervisor who has spent much time to give me his guides, comments, assistance in this research All lecturer and staff of Viet Nam-Netherland master program which is helpful program for me to obtain the knowledge and the method of science studying with international standard All my classmates, especiaily my group have encouragement, cooperation and help during my stUdying and doing the thesis All ofmy family members have help and encouragement for me to try my best to finish this research ii ABSTRACT VietNam is going on the way of modernization, industrialization and especially of globalization with the purpose of economic development and enhancing the life standard of the resident This implies· to· increase further with a rise in the urbanization levels and in the population along with increasing the demand for housing in urban of VietNam To solve the d~mand of urban household, the government has many policies to support the resident such as housing finance system from bank or other financial institution, housing program for low inconie : This study only concentrates to analysis housing credit program from bank for urban household on two major aspects The first is the determinants of probability to get a housing loan And the second is the determinants of housing loan amount SO the household should improve their character and capacity to be able to borrow housing loan from bank Conversely, the bank also should have-suitable credit policy and condition for the customer to speed up effective and sustainable credit development iii TABLES OF CONTENTS ·fi C ertl 1cat1on ··················~····· Acknowledgement : ii Abstract iii Contents :······· iv List of tables and figures viii ··Abbreviation ix CHAPTER I·: INTRODUCTION • · ~ ! 1.1 Prob-em statement ~········································································································· I 1.2 Objectives ofihe stlldy ~ l 1.2.1 General objective ~ 1.2.2 Specific objective · 1.3 Research questions ~·~··············································•···················································2 ·1.4 SuJi:J.macy on rese8.rch· methQdology- and data • 1.5 ·The organization of the thCsis ···~····················································································3 CHAPTER II: LITERATURE REVIEW TheOry ba~kground 2.1.1 Major concepts Household Credit Borrowing Household credit market 2.1.2 Banking theory · Credit management Credit analysis and loan decision Issues in credit market 2.2' Experiences of housing development in some Asian countries •• • .•• ; 2.2.1 In Singapore ; , 2.2.2 In China , 2.2.3 In Thailand 10 iv 2.2.4 In Korea 11 2.3 Theoretical models and empirical studies for house demand 12 2.3.1 General observations 12 2.3.2 Determinants ofhousingloan.- 13 2.3.3.Determinants of loan amount 15 2.3.4 Theoretical models 16 The basic models 16 :The extended models ; 16 2.3 Empirical models applied in previous studies 17 CHAPTER III: RESEARCH METHODOLOGV 20 ·3·.1 AnalyticiJJ· framework ········~······················•·····································································20 3.1.1 Hypotheses , 20 3.1.2Detenninants ofthe probability 20 3, 1.3 Determinants of loan amount • 21 3.1·.4 Specific empirical models • 22 Model1 _ 22 Description of the model1 22 Definition and explanation of variables of the model 22 Expected signs of the variables' coefficients 23 Model2 24 Description of the model , • ; 24 ·Definition and explanation of variables of the model2 24 Expected signs of the variables' coefficients 25 3.2 Data source· aild sa~pling 26 3.2.1 Data source 26 Where is data source? : 26 Its va•I d.Ity? _ 26 What-Is When are they collected? 26 Population : : ; 26 3.2.2 Sampling • 27 v .sample size 27 Sampling method 27 3.3 Analysi-s methods ······~····································································································29 3.3 !Statistical tests for descriptive analysis 29 3.3 Correlation analysis 29 3.3.3 Statistical tests for validity of specific models 30 3.4 Analysis· too1 _ ~ 30 CHAPTER IV: ~SU~TS AND DISCUSSION , 31 4.1 Situation ofhousingcreditprogram for urban household in Ho Chi Minh City 31 4.1.1 Housing demand of urban household : 31 4.1.2 National strategy on housing up to the year 2010 31 4.1.3 The overview of urban finaricial system, borrowing by urban households and housing finance project 32 Financial system and source of credit to urban households in VietNam 32 Overview ofborrowing by urban households 32 · 4.1.4 HCMC Housing development program 35 4.2 IiCMC Housing D-evelopment Bank ·························~······-············································35 ·4~2.l.Overview ofHDBank 35 4.2.2 Housing credit program ofHDBank 38 4.3 HCMC ·aousing D.eveloprri.ent Bank ·····································································~·····40 4J~1Analysis ofborrower' characteristics .40 4.3.2 Correlation.analysis 44 4.3.3 Statistical tests for validity of specific models 47 4.3 Description collecting data and choosing suitable method .48 ·4.4 Results ·of ~odels ~esting and explanation 49 4.4.1 ·:Empirical result- Model I ; 49 4.4.1.1 Enter method ; 49 4.4.1.2 Backward LR method : ; ;· 50 vi 4.4.1.3 Analysis the factor affect to the probability to get housing loan for household : 50 4.4~2 Empiricalresult- Mod.el2 : 54 4.4.2.1 Enter method 54 4.4.2.2 Backward LR method 54 4.4.2.3 Analysis the factor affectto the probability to get housing loan for household , 54 CHAPTER V: CONCLUSION AND IMPLICATION 58 5.1 Conclusion on the applied methodology and limitation • • •.••.•.• • .•• 58 5.2Conclusi0n On the studie.d results 58 5.3 Po.licy impli~ation in macro levCI • 59 5.4 Policy implication for HDBank as well as housing credit program • • •• .•• 59 References Annex vii LIST OF TABLES Table 1.1: Compare mean of variable Table 1.2: Correlation Matrix Table L3: Durbin:_ Watson in model! Tahie 1.4: Durbin- Watson in model Table 1.5: Regression result of model from step to step by back ward method Table 1.6: Regression result ofmodel2 froni stepl to step by back ward method LIST OF FIGURES Figure 2.1: Structure ofprobability to get housing loan Figure 2.2: Structure of samples possibility to get a housing loan Figure 2.3: Growth oftotal assets and chartered capital in HDBank Figure 2.4: Growth ofloan outstanding debts in HDBank Figure 2.5: Outstanding ofhousing credit program and total loan outstanding in HDBank Figure 2.6: Regression standardized residual LIST OF ANNEXS: Annex 1: T-Test result Annex 2: The result of model 1- Enter method Annex3: The resultofmodell- Backward method Annex 4:Theresult ofniodel2- Enter method Annex 5: The result ofmodel2- Backward method viii ·ABBREVIATION HDBank: Housing Development Bank ADB: Asian Development Bank SBV: State Bank of VietNam SOB: State Owned Bank · DLH: Department of Land and House ix Omnibus Tests of Model Coefficients Chisquare Step Step 2(a) Step 3(a) Step 4(a) Step 5(a) Step 6(a) Sig df Step 346.401 10 000 Block 346.401 10 000 Model 346.401 10 000 -.052 820 Block 346.350 000 Model 346.350 000 -.131 717 Block 346.219 000 Model 346.219 000 -.984 321 Block 345.235 000 Model 345.235 000 -1.018 313 Block 344~217 000 Model 344.217 000 -2.600 107 Block 341.617 000 Model 341.617 000 Step Step Step Step Step a A negative Chi-squares value indicates that the Chi-squares value has decreased from the previous step Model Summary ~2Log Cox& SnellR Square Nagelkerke R Square Step likelihoo d 23.237(a) 697 968 ~3.289(a) 697 968 23.420(a) 697 967 24.404(a) 696 966 25.422(b) 695 964 28.022(b) 692 961 · a EstimatiOn termmated at Jteratwn number 20 because maximum iterations has been reached Final solution cannot be found 18 b Estimation terminated at iteration number I because parameter estimates changed by Jess than 00 I Classification Table(a) 1Observed Predicted B Step B Percentage Correct 94 96.9 1 192 99.5 Overall Percentage Step B 98.6 95 97.9 1 192 99.5 Overall Percentage Step B 99.0 95 97.9 191 99.0 Overall Percentage Step4 B 98.6 94 96.9 191 99.0 Overall Percentage Step B 98.3 94 96.9 191 99.0 Overall Percentage Step B 98.3 94 96.9 190 98.4 Overall Percentage 97.9 a The cut value IS 500 19 Variables in the Equation B Step 1(a) Sig df Exp(B) 009 12.523 000 967 029 015 3.626 057 1.030 SEX -1.412 1.423 985 321 244 AGE -.181 146 1.552 213 834 EDU -.613 2.752 050 824 542 903 309 8.553 003 2.467 -1.727 1.145 2.275 131 178 1.110 2.972 139 709 3.034 007 003 6.612 010 1.007 19.476 4396.54 000 996 2873999 26.014 -13.519 4396.55 000 998 000 -.033 009 12.518 000 967 028 015 3.577 059 1.029 SEX -1.366 1.424 921 337 255 AGE -.174 145 1.449 229 840 893 305 8.564 003 2.442 -1.737 1.152 2.273 132 176 1.043 2.971 ;123 726 2.836 007 003 6.582 010 1.007 19.006 4540.62 000 997 1795480 85.712 -13.650 4540.62 000 998 000 -.033 009 12.492 000 967 028 015 3.597 058 1.028 SEX -1.303 1.411 852 356 272 AGE -.171 143 1.442 230 843 305 9.072 003 2.508 -1.538 980 2.460 117 215 007 003 7.707 005 1.007 18.421 4617.60 000 997 1000706 MATUR SIZE HHNO HOUSEVAL COLL · Constant LA MATUR INCOME SIZE HHNO HOUSEVAL COLL Constant Step J(a) Wald -.033 LA INCOME Step 2(a) S.E LA MATUR INCOME SIZE HOUSEVAL COLL 919 20 B S.E Wald Sig df 75374 Constant Step 4(a) LA· MATUR AGE INCOME SIZE HOUSEVAL COLL Constant Step 5(a) 4617.61 000 998 000 -.031 008 13.848 000 969 024 012 3.638 056 1.024 -.151 134 1.269 260 860 873 286 9.296 002 2.394 -1.584 944 2.813 093 205 006 002 8.703 003 1.006 17.990 4795.83 000 997 6501364 7.068 000 998 000 4795.83 -.035 009 16.098 000 966 024 011 4.431 035 1.024 AGE -.245 093 6.971 008 782 INCOME 1.007 279 12.991 000 2.738 -1.239 837 2.192 139 290 007 002 9.818 002 1.007 Constant 6.660 3.261 4.171 041 780.672 LA -.033 008 17.913 000 968 025 011 4.844 028 1.025 -.298 109 7.474 006 742 INCOME 891 235 14.340 000 2.437 HOUSEVAL 007 002 9.155 002 1.007 5.387 3.171 2.887 089 218.605 LA MATUR HOUSEVAL MATUR AGE Constant -12.298 -12.559 SIZE Step 6(a) Exp(B) a Variable(s) entered on step 1: LA, MATUR, SEX, AGE, EDU, INCOME, SIZE, HHNO, HOUSEVAL, COLL~ Model ifTerm Removed Variable Model Log Likelihoo d Change in -2Log Likelihoo d 21 df Sig ofthe Change ·step LA -54.513 85.788 000 MATUR -14.261 5.284 022 SEX AGE -12.178 1.119 290 -12.446 1.655 198 EDU -11.644 052 820 INCOME -28.774 34.310 000 SIZE -13.130 3.023 082 HiiNo -11.693 148 700 -17.706 12.176 000 COLL -12.227 1.217 270 LA -55.937 88.586 000 -14.306 5.323 021 SEX -12.178 1.067 302 AGE -12.466 1.643 200 -31.260 39.232 000 SIZE -13.153 3.017 082 HHNO -11.710 131 717 HOUSEVAL -17.823 12.358 000 COLL -12.260 1.231 267 LA -58.543 93.665 000 MATUR -14.363 5.307 021 SEX -12.202 984 321 AQE -12.524 1.628 202 INCOME -31.336 39.252 000 SIZE -13.350 3.281 070 HOUSEVAL -18.816 14.213 000 ·caLL -12.260 1.101 294 LA -58.544 92.685 000 MATUR -14.571 4.739 029 AGE -12.893 1.381 240 INCOME -31.363 38.322 000 SIZE -13.922 3.441 064 iiOUSEVAL -18.930 13.456 000 COLL -12.711 1.018 313• LA -66.759 108.097 000 HOUSEVAL Step MATUR INCOME Step3 Step4 Step I 22 Step6 MATUR -15.402 5.383 020 AGE -20.020 14.618 000 INCOME -49.005 72.588 000 SIZE -14.011 2.600 107 HOUSEVAL -20.386 15.351 000 LA -69.172 110.322 000 -17.041 6.059 014 AGE -22.800 17.578 000 INCOME -49.013 70.005 000 HOUSEVAL -21.572 15.123 000 · MATuR Variables not in the Equation Score Step 2(a) Variables 049 824 049 824 EDU 033 856 HHNO 122 727 17J 918 SEX 907 341 EDU 000 993 HHNO 046 830 1.114 774 SEX 840 359 EDU 096 756 HHNO 018 895 COLL 722 395 1.620 805 SEX 1.046 306 EDU 003 955 SIZE 2.460 117 ·HHNO 618 432 COLL 101 751 4.461 485 EDU Overall Statistics Step 3(b) Variables Overall Statistics Step 4(c) Variables Overall Statistics Step 5(d) Variables Overall Statistics Step 6(e) Variables Overall Statistics a yanable(s) removed on step 2: EDU b Variable(s) removed on step Sig df 3: HHNO 23 c Variable(s)removed on step 4: SEX d Variable(s) removed on step 5: COLL e Variable(s) removed tm step 6: SIZE 24 ANNEX 4: THE RESULT OF MODEL 2- ENTER METHOD Regression Notes 09-DEC-2008 17:04:02 Output Created Comments Input ' Missing V~ilue · Handling Filter Weight Split File N ofRows in Working Data File 193 Definition of Missing User-defined missing values are treated as missing Cases Used Syntax Resources Statistics are based on cases with no missing values for any variable used REGRESSION /MISSING LISTWISE /STATISTICS COEFF OUTS R ANOVA /CRITERIA=PIN(.05) POUT(.l 0) /NOORIGIN !DEPENDENT LA /METHOD=ENTER MATUR SEX AGE EDU INCOME SIZE HHNO HOUSEVAL Elapsed Time 0:00:00.07 Memory Required 3980 bytes Additional ·Memory· Required for Residual Plots bytes yariables Entered/Removed(b) 25 Mode Variables Entered I I Variables Removed Method HOUSEV AL, SEX, MATUR,· EDU, HHNO, INCOME AGE,.· 'SIZE(a) Enter ·a All requested variables entered b Dependent Variable: LA Model Summary Mode I R Std Error of the Estimate Adjusted R Square R Square 858(a) 724 736 75.596 a Predictors: (Constant), HOUSEVAL, SEX, MATUR, EDU, HHNO, INCOME, AGE, SIZE ANOVA(b) Sum of Squares Model Regressi 2927304 on 453 Mean Square df 365913.057 Residual 1051502 108 184 3978806 561 192 Total Sig F 64.030 OOO(a) 5714.685 a Predictors: (Constant), HOUSEV AL, SEX, MATUR, EDU, HHNO, INCOME, AGE, SIZE b Dependent Variable: LA Coefficients(a) Model Unstandardized Coefficients Standardized Coefficients t 26 Sig B Std Error -220.666 60.377 353 140 SEX·· 1.799 AGE EDU (Constant) MATUR INCOME SIZE HHNO HOUSEVAL Beta -3.655 000 097 2.526 012 11.266 006 160 873 1.370 1.503 058 911 363 44.821 33.372 054 1.343 181 15.146 1.300 562 11.653 000 -10.074 10.446 -.067 -.964 336 35.802 18.183 101 1.969 050 131 014 394 9.356 000 a Dependent Variable: LA 27 ANNEX 5: THE RESULT OF MODEL 2- BACKWARD METHOD Regression Notes 09-DEC-2008 17:05:45 Output Created Comments Input Missing Value Handling Filter Weight Split File NofRowsin Working Data File 193 Definition of Missing User-defined missing values are treated as missing Cases Used REGRESSION /MISSING LISTWISE /STATISTICS COEFF OUTS R ANOVA /CRITERIA=PIN(.05) POUT(.1 0) /NOORIGIN /DEPENDENT LA IMETHOD=BACKWARD MATUR SEX AGE EDU INCOME SIZE HHNO HOUSEVAL Syntax ,, Resources Statistics are based on cases with no missing values for any variable used Elapsed Time 0:00:00.17 Memory Required· 4516 bytes Additional Memory Required for Residual Plots bytes 28 Variables Entered/Removed(b) Variables Removed Mode Variables I Entered Method HOUSE\TAL,· SEX~ MATUR · · ' EDU,HHNO, INCOME, AGE, SIZE(a) Enter SEX Backward (criterion: Probability ofFto-remove >= 00) AGE Backward (criterion: Probability ofFto-remove >= 00) SIZE Backward (criterion: Probability ofFto-remove >= 00) EDU Backward (criterion: Probability ofFto-remove>= 100) a All requested vanables entered b Dependent Variable: LA Model Summary Mode I a R Adjusted R Square R Square Std Error of the Estimate 858(a) 736 724 75.596 858(b) 736 726 75.396 857(c) 735 726 75,358 857(d) 734 727 75.204 855(e) 732 726 75.343 Predictors: (Constant), HOUSEVAL, SEX, MATUR, EDU, HHNO, INCOME, AGE, SIZE b Predictors: (Constant), HOUSEVAL, MATUR, EDU, HHNO, INCOME, AGE, SIZE c Predictors: (Constant), HOUSEV AL, MATUR, EDU, HHNO, INCOME, SIZE d Predictors: (Constant), HOUSEVAL, MATUR, EDU, HHNO, INCOME e Predictors: (Constant), HOUSEV AL, MATUR, HHNO, INCOME ANOVA(f) 29 Sum of Squares Model df Regression 2927304.453 Residual 1051502.108 184 3978806.561 192 Total Mean Square F Sig 64.030 OOO(a) 73.561 OOO(b) 85.7'73 OOO(c) 103.303 OOO(d) 128.232 OOO(e) 365913.057 5714.685 Regres.sion 2927158.717 Residual· 1051647.844 185 Total 3978806.561 192 Regression 2922545.608 Residual 1056260.953 186 Total 3978806.561 192 Regression 2921204.838 Residual 1057601.723 187 Total 3978806.561 192 Regression 2911625.704 727906.426 Residual 1067180.857 188 5676.494 Total 3978806.561 192 418165.531 5684.583 487090.935 5678.822 584240.968 5655.624 a Predtctors: (Constant), HOUSEVAL, SEX; MATUR, EDU, HHNO, INCOME, AGE, SIZE b Predictors: (Constant), HOUSEVAL, MATUR, EDU, HHNO, INCOME, AGE, SIZE c Predictors: (Constant), HOUSEVAL, MATUR, EDU, HHNO, INCOME, SIZE d Predictors: (Constant), HOUSEVAL, MATUR, EDU, HHNO, INCOME e Predictors: (Constant), HOUSEV AL, MATUR, HHNO, INCOME f Dependent Variable: LA Coefficients(a) Model · Unstandardized Coefficients 30 Standardized Coefficients t Sig B l 000 097 2.526 012 11.266 006 160 873 1.370 1.503 058 911 363 EDU 44.821 33.372 054 1.343 181 INCOME 15.146 1.300 562 11.653 000 -10.074 10.446 -.067 -.964 336 35.802 18.183 101 1.969 050 131 014 394 9.356 000 J.736 000 2.531 012 60.377 353 140 SEX 1.799 AGE SIZE HHNO HOUSEVAL (Constant) 351 139 AGE 1.341 1.489 .057 901 369 EDU 44.316 33.134 054 1.337 183 INCOME· 15.136 1.295 11.691 000 SIZE -9.852 10.326 -.066 -.954 341 iiHNo 35.787 18.135 101 1.973 050 131 014 395 9.394 000 -188.173 47.876 -3.930 000 361 138 099 2.611 010 39.720 32.722 048 1.214 226 INCOME 15.272 1.285 567 11.883 000 SIZE -3.815 7.852 -.025 -.486 628 HHNO 36.178 18.121 102 1.997 047 131 014 395 9.399 ;000 -189.006 47.747 -3.958 000 360 138 099 2.614 010 EDU 42.043 32.305 051 1.301 195 ·INCOME 15.11() L239 561 12.197 000 HHNO 31.498 15.317 089 2.056 041 131 014 394 9.407 000 -'140.962 30.339 -4.646 000 346 138 095 2.511 013 15.345 1.228 570 12.496 000 (Constant) MATUR EDU HOUSEVAL (Constant) MATUR HOUSEVAL 58.449 - · HOUSEVAL -218.346 096 MATUR Beta -3.655 -220.666 (Constant) MATUR Std Error (Constant) MATlJR INCOME 31 562 Unstandardized Coefficients B Model HHNO Standardized Coefficients Std Error Beta t Sig 26.925 14.936 076 1.803 073 129 014 389 9.308 000 HOUSEVAL a Dependent Vanable: LA Excluded Variables(e) Mode I Beta In T Sig Partial Correlation Collinearity Statistics Tolerance SEX · 006(a) 160 873 012 955 SEX ;002(b) 053 958 004 968 AGE 057(b) .901 369 066 362 SEX 001(c) 017 987 001 973 AGE 018(c) 370 712 027 626 SIZE -.o25(c) -.486 ·.628 -.036 522 SEX -.004(d) -.098 922 -.007 981 AGE 004(d) 088 930 006 655 SIZE -.035(d) -.670 504 -.049 533 EDU 051(d) 1.301 195 :095 932 a Predictors m the Model: (Constant), HOUSEV AL, MATUR, EDU, HHNO, INCOME, AGE, SIZE b Predictors in the Model: (Constant), HOUSEVAL, MATUR, EDU, HHNO, INCOME, SIZE c Predictors in the Model: (Constant), HOUSEVAL, MATUR; EDU, HHNO, INCOME d Predictors in the Model: (Constant), HOUSEVAL, MA TUR, HHNO, INCOME e DependentVariable: LA 32