This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: 1. USGDP The USGDP data has been downloaded with GDP, M1,M2 and tbill and these had been declared as time series. The natural logs for all variables except tbill has been taken, After that the lag values has been calculated. THe data has been detrended for GDP, M1 and T bill. Aftewr that the line plot of Detrended log values of GDP, M1 and IPI had been created in a single window as shown below. 60 120 180 240 300 360 420 480 Month 2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 F G H Y Figure: 1.c Detrended Log(IPI) has been regressed on the Tbill rate. The results are as shown below in the table: TABLE 1.d : OLS Results for Regression Model 1 (standard error) t pvalue IPI growth 0.368 (0.034) 15.78 0.00013 Tbill rate 0.0706 (0.076) 2.86 0.0003 R2 = .28 Fstat. pvalue = .0003 AIC = 3.223 SIC = .037543 DurbinWatson dstatistic( 5, 528) = .2127027 (the 5% critical range is 1.44…1.68) Since DW < 1.44, we strongly reject the no serial correlation null hypothesis at the 5% level The residules has been calculated and regressed on the lagged Tbill value: Table 1.e Source SS df MS Model 1373.632 5.332 1536.44269 Residual 4345.40457 99 71.16911937 Number of obs = 527 F( 2, 525) = 3452.54 Prob > F = 4.0000 Adj Rsquared = 0.9625 Rsquared = 0.9633 The results of the LM statistics are as follows: (Table 1.e) The formula for the LM statistics for tye residuals u at 1 to 6 is as follow: Random Effects Model: v(i,t) = e(i,t) + u(i) Estimates: Var[e] = .361260D02 Var[u] = .119159D01 Corr[v(i,t),v(i,s)] = .767356 Lagrange Multiplier Test vs. Model (3) = Lagrange Multiplier Test vs....
View
Full Document
 Fall '08
 Staff
 Econometrics, Null hypothesis, Statistical hypothesis testing, log likelihood

Click to edit the document details