Ox version 2.20 (Sun) (C) J.A. Doornik, 1994-2000 Arfima package version 1.00, object created on 20-09-2000 ---- Database information ---- 6 variables, 115 observations name sample period min mean max stddev Dsde 1979 (5) 1988 (11) -1.68 -0.0047826 1.18 0.39588 Dsbe 1979 (5) 1988 (11) -1.09 -0.013652 1.32 0.34664 Ddfr 1979 (5) 1988 (11) -1.63 0.0093913 3.02 0.61302 Ddir 1979 (5) 1988 (11) -2.93 -0.031043 3.2 0.90494 Ddit 1979 (5) 1988 (11) -1.51 -0.0043478 2.7 0.55317 Ddnl 1979 (5) 1988 (11) -3.47 -0.012522 2.17 0.66384 var: Dsde 1979 (5) - 1988 (11) ---- Maximum likelihood estimation of ARIMA(0,d,0) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter 0.234639 0.07846 2.99 0.003 log-likelihood -51.4406991 no. of observations 115 no. of parameters 2 AIC 106.881398 AIC/T 0.929403463 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.142726 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.34825 ---- Maximum likelihood estimation of ARIMA(0,d,1) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter 0.165763 0.1203 1.38 0.171 MA-1 0.100146 0.1376 0.728 0.468 log-likelihood -51.1957824 no. of observations 115 no. of parameters 3 AIC 108.391565 AIC/T 0.942535347 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.142337 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.39458 ---- Maximum likelihood estimation of ARIMA(0,d,2) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter 0.0866720 0.1438 0.603 0.548 MA-1 0.178159 0.1630 1.09 0.277 MA-2 0.0965652 0.1147 0.842 0.402 log-likelihood -50.8570679 no. of observations 115 no. of parameters 4 AIC 109.714136 AIC/T 0.954035964 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.141621 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.28378 0.23358 ---- Maximum likelihood estimation of ARIMA(0,d,3) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter 0.0278430 0.1762 0.158 0.875 MA-1 0.236030 0.1925 1.23 0.223 MA-2 0.143116 0.1514 0.945 0.347 MA-3 0.0583704 0.1116 0.523 0.602 log-likelihood -50.7182041 no. of observations 115 no. of parameters 5 AIC 111.436408 AIC/T 0.969012245 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.141312 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.26211 0.20609 0.13378 ---- Maximum likelihood estimation of ARIMA(1,d,0) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter 0.104354 0.1975 0.528 0.598 AR-1 0.177646 0.2393 0.742 0.459 log-likelihood -51.0713638 no. of observations 115 no. of parameters 3 AIC 108.142728 AIC/T 0.940371545 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.142139 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.30147 ---- Maximum likelihood estimation of ARIMA(1,d,1) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter -0.626681 0.3568 -1.76 0.082 AR-1 0.916154 0.1154 7.94 0.000 MA-1 -0.0391186 0.2671 -0.146 0.884 log-likelihood -50.0430749 no. of observations 115 no. of parameters 4 AIC 108.08615 AIC/T 0.939879564 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.138852 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.61205 -0.41020 ---- Maximum likelihood estimation of ARIMA(1,d,2) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter -0.702263 0.4212 -1.67 0.098 AR-1 0.932464 0.1043 8.94 0.000 MA-1 0.0189538 0.3366 0.0563 0.955 MA-2 0.0307168 0.1553 0.198 0.844 log-likelihood -50.0249729 no. of observations 115 no. of parameters 5 AIC 110.049946 AIC/T 0.956956051 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.138709 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.57653 -0.31035 0.014391 ---- Maximum likelihood estimation of ARIMA(1,d,3) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter -0.708845 0.5435 -1.30 0.195 AR-1 0.933599 0.1184 7.89 0.000 MA-1 0.0242909 0.4384 0.0554 0.956 MA-2 0.0329781 0.1981 0.166 0.868 MA-3 0.00219821 0.1192 0.0184 0.985 log-likelihood -50.0248028 no. of observations 115 no. of parameters 6 AIC 112.049606 AIC/T 0.974344397 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.138699 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.77790 -0.51306 -0.043053 -0.043638 ---- Maximum likelihood estimation of ARIMA(2,d,0) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter -0.624472 0.3276 -1.91 0.059 AR-1 0.874118 0.3363 2.60 0.011 AR-2 0.0377259 0.2190 0.172 0.864 log-likelihood -50.0413584 no. of observations 115 no. of parameters 4 AIC 108.082717 AIC/T 0.939849711 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.138849 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.27042 0.10299 ---- Maximum likelihood estimation of ARIMA(2,d,1) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter -0.624382 0.1529 -4.08 0.000 AR-1 0.0168311 0.1402 0.120 0.905 AR-2 0.818195 0.1311 6.24 0.000 MA-1 0.854163 0.1366 6.25 0.000 log-likelihood -49.64074 no. of observations 115 no. of parameters 5 AIC 109.28148 AIC/T 0.950273739 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.13783 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.54205 0.021101 -0.30873 ---- Maximum likelihood estimation of ARIMA(2,d,2) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter -0.729958 0.2487 -2.94 0.004 AR-1 0.0457664 0.1349 0.339 0.735 AR-2 0.839566 0.1228 6.84 0.000 MA-1 0.944780 0.2428 3.89 0.000 MA-2 0.0896162 0.2066 0.434 0.665 log-likelihood -49.5667231 no. of observations 115 no. of parameters 6 AIC 111.133446 AIC/T 0.966377793 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.137537 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.27042 0.10299 0.0045609 -0.031078 ---- Maximum likelihood estimation of ARIMA(2,d,3) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter -0.593723 0.5136 -1.16 0.250 AR-1 0.0213214 0.1728 0.123 0.902 AR-2 0.815301 0.1546 5.27 0.000 MA-1 0.840035 0.4038 2.08 0.040 MA-2 -0.0670270 0.4565 -0.147 0.884 MA-3 -0.0718503 0.1553 -0.463 0.645 log-likelihood -49.4638444 no. of observations 115 no. of parameters 7 AIC 112.927689 AIC/T 0.981979902 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.137454 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.27042 0.10299 0.0052041 -0.031190 0.022433 ---- Maximum likelihood estimation of ARIMA(3,d,0) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter -0.730489 0.5756 -1.27 0.207 AR-1 0.979561 0.5825 1.68 0.095 AR-2 0.00124696 0.2999 0.00416 0.997 AR-3 -0.0381238 0.1893 -0.201 0.841 log-likelihood -50.0208159 no. of observations 115 no. of parameters 5 AIC 110.041632 AIC/T 0.956883755 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.138663 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.26751 0.095344 0.028275 ---- Maximum likelihood estimation of ARIMA(3,d,1) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter -0.719764 0.3075 -2.34 0.021 AR-1 0.131496 0.3657 0.360 0.720 AR-2 0.834076 0.1315 6.34 0.000 AR-3 -0.0748724 0.2375 -0.315 0.753 MA-1 0.844254 0.1377 6.13 0.000 log-likelihood -49.5872595 no. of observations 115 no. of parameters 6 AIC 111.174519 AIC/T 0.966734947 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.137595 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.26751 0.095344 0.028275 0.012461 ---- Maximum likelihood estimation of ARIMA(3,d,2) model ---- The estimation sample is 1979 (5) - 1988 (11) The dependent variable is: Dsde (in deviation from sample mean) Coefficient Std.Error t-value t-prob d parameter -0.806157 1.723 -0.468 0.641 AR-1 0.508299 3.761 0.135 0.893 AR-2 0.824874 0.1943 4.24 0.000 AR-3 -0.387588 3.143 -0.123 0.902 MA-1 0.545708 2.302 0.237 0.813 MA-2 -0.262564 2.096 -0.125 0.901 log-likelihood -49.614272 no. of observations 115 no. of parameters 7 AIC 113.228544 AIC/T 0.984596036 mean(Dsde) -0.0047826 var(Dsde) 0.156718 sigma^2 0.137535 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): Strong convergence Used starting values: 0.0000 0.26751 0.095344 0.028275 0.011187 -0.021693 ---- Maximum likelihood estimation of ARIMA(3,d,3) model ---- The estimation sample is 1979 (5) - 1988 (11) Model estimation failed: No convergence (no improvement in line search) Free coefficients: // [0][0]...[0][6] -1.0000 0.82138 0.80751 -0.64734 0.43788 -0.43450 -0.088768 BFGS estimation using numerical derivatives (eps1=0.0001; eps2=0.005): No convergence (no improvement in line search) Used starting values: // [0][0]...[0][6] 0.0000 0.26751 0.095344 0.028275 0.011040 -0.021617 -0.0094748 time: 11.83