TSP Version 4.5 (04/27/00) Windows32 128MB Copyright (C) 2000 TSP International ALL RIGHTS RESERVED 02/04/04 4:21 PM In case of questions or problems, see your local TSP consultant or send a description of the problem and the associated TSP output to: TSP International P.O. Box 61015, Station A Palo Alto, CA 94306 USA PROGRAM COMMAND *************************************************************** | 1 OPTIONS MEMORY = 128; | 2 ?************************************************************* | 2 ? * | 2 ? THIS COMPUTER PROGRAM IS FOR THE SPECIFICATION * | 2 ? * | 2 ? Y=ALPHA + LAMDA*WEIGHTINGMATRIX*Y + X*BETA + U, * | 2 ? U=RHO*WEIGHTINGMATRIX*U + E * | 2 ? * | 2 ?************************************************************* | 2 ?************************************************************* | 2 ? * | 2 ?EXPLANATION OF THE VARIABLES USED IN THE PROGRAM * | 2 ?Y.........VECTOR OF DEPENDENT VARIABLE * | 2 ?X.........VECTOR OF EXOGENOUS EXPLANATORY VARIABLE * | 2 ?U.........VECTOR OF DISTURBANCES * | 2 ?MMAT......SPATIAL WEIGHTING MATRIX * | 2 ?E.........VECTOR OF INNOVATIONS * | 2 ?LAMDA.....SPATIAL AUTOREGRESSIVE PARAMETER IN DEPENDENT VAR * | 2 ?RHO.......SPATIAL AUTOREGRESSIVE PARAMETER IN DISTURBANCE * | 2 ?A2SLS.....2SLS ESTIMATE OF ALPHA (STEP 1) * | 2 ?AH2SLS....2SLS ESTIMATE OF ALPHA (STEP 2) * | 2 ?B2SLS.....2SLS ESTIMATE OF BETA (STEP 1) * | 2 ?BH2SLS....2SLS ESTIMATE OF BETA (STEP 2) * | 2 ?LAM2SLS...2SLS ESTIMATE OF LAMDA (STEP 1) * | 2 ?LAMH2SLS..2SLS ESTIMATE OF LAMDA (STEP 2) * | 2 ?RHOGM.....GENERAL MOMENTS ESTIMATOR OF RHO * | 2 ?NS........SAMPLE SIZE * | 2 ?************************************************************* | 2 ?************************************************************* | 2 ?READ THE DATA WHICH IS USED IN THE SAMPLE TSP PROGRAM * * | 2 ?************************************************************* | 2 | 2 SET NS = 100; | 3 SMPL 1 NS; | 4 READ (FILE='C:\TEMP\VAR2.DAT', FORMAT = FREE) Y X1 X2; | 5 READ (FILE='C:\TEMP\MMAT.DAT') M1-M100; | 6 MMAKE MMAT M1-M100; | 7 MMAKE YMAT Y; | 8 MAT WYMAT = MMAT*YMAT; | 9 MMAKE X1MAT X1; | 10 MMAKE X2MAT X2; | 11 MAT WX2MAT = MMAT*X2MAT; | 12 MAT WWX2MAT = MMAT*WX2MAT; | 13 UNMAKE WYMAT WY; | 14 UNMAKE WX2MAT WX2; | 15 UNMAKE WWX2MAT WWX2; | 16 | 16 ?************************************************************* | 16 ?IN THE FIRST STEP ESTIMATE THE MODEL USING TWO-STAGE LEAST * | 16 ?SQUARES AND GET THE ESTIMATED DISTURBANCES * | 16 ?************************************************************* | 16 | 16 2SLS (SILENT) (INST=(X1,X2,WX2,WWX2)) Y X1 WY X2; | 17 SET A2SLS = @COEF(1); | 18 SET LAM2SLS = @COEF(2); | 19 SET B2SLS = @COEF(3); | 20 SET S2SLS = (@SSR)/(NS-1); | 21 TITLE 'STEP 1: 2SLS ESTIMATORS'; | 22 PRINT A2SLS LAM2SLS B2SLS S2SLS; | 23 GENR UH = Y - A2SLS*X1 - LAM2SLS*WY - B2SLS*X2 ; | 24 MMAKE UHVEC UH; | 25 | 25 ?************************************************************* | 25 ?IN THE SECOND STEP USE THE ESTIMATES OF DISTURBANCES * | 25 ?FOR THE GENERAL MOMENTS(GM) ESTIMATOR * | 25 ?************************************************************* | 25 | 25 INPUT C:\TEMP\GMPROC1; | 25 | 25 ?GMPROC PROCEDURE HAS BEEN DEFINED BELOW; | 25 ?**************************************************************** | 25 ?GENERAL MOMENTS ESTIMATION PROCEDURE FOR ESTIMATION OF RHOH * | 25 ?MOMENT EQUATIONS INVOLVING * | 25 ?E’E, (ME)’(ME), AND (ME)’E, WHERE ME=WEIGHTINGMATRIX*E * | 25 ?**************************************************************** | 25 | 25 SMPL 1 NS; | 26 MMAKE UHVEC UH; | 27 MAT VHVEC = MMAT*UHVEC; | 28 MAT WHVEC = MMAT*VHVEC; | 29 UNMAKE VHVEC VH; | 30 UNMAKE WHVEC WH; | 31 | 31 ?**************************************************************** | 31 ?CALCULATE VARIOUS SAMPLE MOMENTS APPEARING IN THE SYSTEM * | 31 ?OF EQUATIONS FOR RHO FROM ESTIMATED DISTURBANCES * | 31 ?**************************************************************** | 31 | 31 GENR U2H = UH*UH; | 32 GENR V2H = VH*VH; | 33 GENR UVH = UH*VH; | 34 GENR VWH = VH*WH; | 35 GENR W2H = WH*WH; | 36 GENR UWH = UH*WH; | 37 | 37 MSD (SILENT) U2H V2H UVH VWH W2H UWH ; | 38 | 38 SET U2HM = @MEAN(1); | 39 SET V2HM = @MEAN(2); | 40 SET UVHM = @MEAN(3); | 41 SET VWHM = @MEAN(4); | 42 SET W2HM = @MEAN(5); | 43 SET UWHM = @MEAN(6); | 44 | 44 ?CALCULATE TR(MMAT'MMAT)/(SAMPLE SIZE) | 44 ?*************************************************************** | 44 MAT MMM = TR(MMAT'MMAT)/NS; | 45 | 45 ?**************************************************************** | 45 ?PREPARE ESTIMATION OF RHO AND S FROM THREE EQUATIONS BY NLS * | 45 ?**************************************************************** | 45 | 45 FRML EQU HMY = HM1*RHOGM + HM2*RHOGM**2 + HMS2*SGM; | 46 | 46 SMPL 1 3; | 47 GENR HM1 = 0; | 48 GENR HM2 = 0; | 49 GENR HMS2 = 0; | 50 GENR HMY = 0; | 51 | 51 SMPL 1 1; | 52 GENR HM1 = -2*UVHM; | 53 GENR HM2 = V2HM; | 54 GENR HMS2 = -1; | 55 GENR HMY = -U2HM; | 56 | 56 SMPL 2 2; | 57 GENR HM1 = -2*VWHM; | 58 GENR HM2 = W2HM; | 59 GENR HMS2 = -MMM; | 60 GENR HMY = -V2HM; | 61 | 61 SMPL 3 3; | 62 GENR HM1 = -(V2HM+UWHM); | 63 GENR HM2 = VWHM; | 64 GENR HMS2 = 0; | 65 GENR HMY = -UVHM; | 66 | 66 ?**************************************************************** | 66 SMPL 1 3; | 67 PARAM RHOGM SGM ; | 68 LSQ (SILENT,MAXIT=500,MAXSQZ=20,TOL=0.001) EQU; | 69 TITLE 'STEP 2: NLS ESTIMATORS OF RHO AND SIGMA^2'; | 70 PRINT RHOGM; | 71 PRINT SGM; | 72 | 72 | 72 | 72 | 72 ?************************************************************* | 72 ?IN THE FOURTH STEP USE THE GM ESTIMATES FOR SPATIAL * | 72 ?AUTOREGRESSIVE PARAMETER, CORRECT FOR SPATIAL CORRELATION * | 72 ?IN DISTURBANCES AND ESTIMATE THE CORRECTED MODEL BY 2SLS * | 72 ?************************************************************* | 72 | 72 INPUT C:\TEMP\TSLSPROC2; | 72 | 72 ?GLSPROC PROCEDURE HAS BEEN DEFINED BELOW | 72 | 72 ?GENERALIZED LEAST SQUARES PROCEDURE | 72 ?************************************************************* | 72 SMPL 1 NS; | 73 MMAKE YVEC Y; | 74 MMAKE WYVEC WY; | 75 MMAKE X1VEC X1; | 76 MMAKE X2VEC X2; | 77 MAT YSVEC = YVEC - RHOGM*MMAT*YVEC; | 78 MAT WYSVEC = WYVEC - RHOGM*MMAT*WYVEC; | 79 MAT X1SVEC = X1VEC - RHOGM*MMAT*X1VEC; | 80 MAT X2SVEC = X2VEC - RHOGM*MMAT*X2VEC; | 81 MAT WX2SVEC = MMAT*X2SVEC; | 82 MAT WWX2SVEC = MMAT*WX2SVEC; | 83 UNMAKE YSVEC YS; | 84 UNMAKE WYSVEC WYS; | 85 UNMAKE X1SVEC X1S; | 86 UNMAKE X2SVEC X2S; | 87 UNMAKE WX2SVEC WX2S; | 88 UNMAKE WWX2SVEC WWX2S; | 89 | 89 2SLS (SILENT) (INST=(X1S,X2S,WX2S,WWX2S)) YS X1S WYS X2S; | 90 SET AH2SLS = @COEF(1); | 91 SET LAMH2SLS = @COEF(2); | 92 SET BH2SLS = @COEF(3); | 93 SET SH2SLS = (@SSR)/(NS-1); | 94 TITLE 'STEP 3: 2SLS ESTIMATOR'; | 95 PRINT AH2SLS LAMH2SLS BH2SLS SH2SLS; | 96 | 96 | 96 | 96 | 96 | 96 ?************************************************************** | 96 DATE;STOP;END; EXECUTION ******************************************************************************* | 0 Current sample: 1 to 100 STEP 1: 2SLS ESTIMATORS ======================== A2SLS LAM2SLS B2SLS S2SLS Value 0.38747 0.50005 1.00002 2.10280 Current sample: 1 to 100 Current sample: 1 to 3 Current sample: 1 to 1 Current sample: 2 to 2 Current sample: 3 to 3 Current sample: 1 to 3 STEP 2: NLS ESTIMATORS OF RHO AND SIGMA^2 ========================================== RHOGM = 0.65760 SGM = 0.97590 Current sample: 1 to 100 STEP 3: 2SLS ESTIMATOR ======================= AH2SLS LAMH2SLS BH2SLS SH2SLS Value -0.0074573 0.50009 1.00002 1.00898 02/04/04 4:21 PM ******************************************************************************* END OF OUTPUT. MEMORY USAGE: ITEM: DATA ARRAY TOTAL MEMORY UNITS: (4-BYTE WORDS) (MEGABYTES) MEMORY ALLOCATED : 31500000 128.0 MEMORY ACTUALLY REQUIRED : 59245 2.3 CURRENT VARIABLE STORAGE : 39780