tag:blogger.com,1999:blog-2198942534740642384.comments2017-12-10T15:37:53.379-08:00Econometrics Beat: Dave Giles' BlogDave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comBlogger3844125tag:blogger.com,1999:blog-2198942534740642384.post-11740913734259047312017-11-18T09:48:39.913-08:002017-11-18T09:48:39.913-08:00You would include the (negative) sign.
DG You would include the (negative) sign.<br />DG Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-61757215539280694012017-11-18T09:26:15.288-08:002017-11-18T09:26:15.288-08:00Hello Prof Giles,
First of all, thanks you for yo...Hello Prof Giles,<br /><br />First of all, thanks you for your very informative blog. I just want few precision. In your formula with C*, does C* come in the formula in absolute terms or we need to put its coefficient sign if C* is negative. <br /><br />Thanks You<br /><br />Sincerely,<br />Quebechttps://www.blogger.com/profile/03769268296619128578noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-25922237033745191052017-11-18T06:44:31.240-08:002017-11-18T06:44:31.240-08:00The dummy is just "shifting" the interce...The dummy is just "shifting" the intercept - so, you just list the dummy and intercept in the set of exogenous regressors.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-84999094010444347582017-11-18T06:33:42.656-08:002017-11-18T06:33:42.656-08:00Dear Prof. Giles,
How I use dummy(one structural b...Dear Prof. Giles,<br />How I use dummy(one structural break)in eq T-Y, in T-Y procedüre (I used dummy in Var) ?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-50858478412828293852017-11-17T09:25:40.341-08:002017-11-17T09:25:40.341-08:00Very low R-squared values often arise when cross-s...Very low R-squared values often arise when cross-section data are used. It's very common.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-43761746461600364852017-11-17T09:23:29.888-08:002017-11-17T09:23:29.888-08:00Dear Prof Dave,
I ran a regression on a cross-sec...Dear Prof Dave,<br /><br />I ran a regression on a cross-sectional firm level data across different countries, but my R2 is about 0.05.<br /><br />This worries me as it seems my model do not explain much of the variation in the dependent variable.<br /><br />What do you suggest I do about this please?<br /><br />okechukwu obiorahttps://www.blogger.com/profile/08115130223921442578noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-71893165586333482582017-11-16T07:16:04.528-08:002017-11-16T07:16:04.528-08:00If that's really the case, then you need 1 20 ...If that's really the case, then you need 1 20 for the lags and you add lags 21 and 22 of each variable in to the exogenous box.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-31804636565772780322017-11-14T13:00:56.658-08:002017-11-14T13:00:56.658-08:00Dear Prof Giles,
I have some questions regarding ...Dear Prof Giles,<br /><br />I have some questions regarding to the granger causality test in the last step, my X and Y variables are both I(2), and the lag information criteria suggest the lag length to be 3. The LM test shows the serial correlation cannot be removed until 20 lags. So, what I should input for my new VAR models in lag interval and exogeneous variables. It should be 1 to 5 in lag interval?<br />Looking forward to your guidance. Thank you in advance.<br /><br />Best regards.Elsie Wanghttps://www.blogger.com/profile/07561861718987628959noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-15209408510497236102017-11-14T03:59:05.884-08:002017-11-14T03:59:05.884-08:00very informative. Thanks.very informative. Thanks.prashant joshihttps://www.blogger.com/profile/04158240843900481451noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-28523848548743788162017-11-07T14:16:36.110-08:002017-11-07T14:16:36.110-08:00A new version of the R package 'plm' (ver ...A new version of the R package 'plm' (ver 1.6-6) is officially released; it contains said function.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-76755473517740998522017-11-05T16:00:14.649-08:002017-11-05T16:00:14.649-08:00need more articles Dave!!need more articles Dave!!not trampishttp://nottrampis.blogspot.com.au/noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-50175415425615846372017-10-26T07:59:46.577-07:002017-10-26T07:59:46.577-07:00The J-B is a test for Normality of the errors - it...The J-B is a test for Normality of the errors - it tests jointly for "no skewness and no excess kurtosis". Putting that to one side, non-normality of the errors is not an issue with the TY procedure, provided that the sample size is large enough. The chi square distribution of the (modified) Wald test for non-causality that you're using only holds asymptotically, and it doesn't require normality of the errors in the model.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-59939233314111473792017-10-26T01:33:11.826-07:002017-10-26T01:33:11.826-07:00Dear Professor Giles,
I do have one question with...Dear Professor Giles,<br /><br />I do have one question with regard to step 5:<br /><br />"Make sure that the VAR is well-specified. For example, ensure that there is no serial correlation in the residuals. If need be, increase p until any autocorrelation issues are resolved."<br /><br />As proposed, I apply the portmanteau test and check for dynamic stability (variables at level). However, what if the Jarque-Bera test indicates skewness of my data at almost all lags?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-61139653311541760852017-10-23T16:44:06.818-07:002017-10-23T16:44:06.818-07:00Dear Professor, on the premise that Cointegration ...Dear Professor, on the premise that Cointegration implies Granger Causality (GS)..If I have 10 variables that are found to be cointegration but after using the T-M GS procedure, I find that some but not all the variables at least a unidirectional relationship, do I conclude there is a problem? I understand that for two variables that are cointegrated, at least one must GS the other..However, things appear to be silent for 10 or 20 variables. Please provide some insights. Would be well appreciated..To put things in perspective, let's say X1, X2, X3 and X4 are found to be cointegrated, but X3 does not GS X4 neither does X4 GS X3..Howver, the rest variables at least GS each, is this a problem? Is there any inconsistency when put side by side with the fact that they are cointegrated? Please provide guidance.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-25837103222455595092017-10-18T06:36:15.990-07:002017-10-18T06:36:15.990-07:00Javaid - I'm afraid I don't understand wha...Javaid - I'm afraid I don't understand what could be meant by "uni-directional cointegration".Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-69255754261115203432017-10-18T00:25:21.255-07:002017-10-18T00:25:21.255-07:00Hello sir, thank you for such a heuristic illustra...Hello sir, thank you for such a heuristic illustration. <br />i wanted to know sir, the variables included in analysis should bear unidirectional co-integration, if there is a feedback then ARDL model is inappropriate? <br />Javaid Nazirhttps://www.blogger.com/profile/14105352234310654846noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-50475271775987139912017-10-16T01:35:40.807-07:002017-10-16T01:35:40.807-07:00Hi Prof. Thanks for the insights. However, suppose...Hi Prof. Thanks for the insights. However, suppose you need to do a multivariate test of heteroskedasticity for the whole SUR system rather than equation-by-equation. How do you do it in Eviews? Kindly refer to the following paper:<br />System Misspecification Testing and Structural Change in the Demand for Meats<br />Anya McGuirk, Paul Driscoll, Jeffrey Alwang, and Huilin Huang (1995).<br />There's need to take into account cross-equation correlation when conducting multivariate diagnostic tests.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-41558765937017069032017-10-16T00:08:10.813-07:002017-10-16T00:08:10.813-07:00It is much needed! I have always thought of this. ...It is much needed! I have always thought of this. It will be a significant stuffs. The journals are always biased for POSITIVE results which is unfair. Santosh Dashhttps://www.blogger.com/profile/02016226999263087762noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-90681617944990704952017-10-15T07:03:35.864-07:002017-10-15T07:03:35.864-07:00Thanks for that!!!Thanks for that!!!Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-58897830790330393532017-10-15T02:52:47.809-07:002017-10-15T02:52:47.809-07:00The development version of the R package "plm...The development version of the R package "plm" has the function "pgrangertest" to perform the panel Granger causality test by Dumistrecu/Hurlin (2012).Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-41576228402654901792017-10-11T22:59:37.312-07:002017-10-11T22:59:37.312-07:00Excellent and helpful post… I am so glad to left c...Excellent and helpful post… I am so glad to left comment on this. This has been a so interesting ..I appreciate your effort..Smithhttps://www.youniversitytv.com/noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-88890671314558698402017-10-11T04:57:44.694-07:002017-10-11T04:57:44.694-07:00Thanks a lot for the explanation!Thanks a lot for the explanation!Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-6536275454007533372017-10-09T05:01:57.907-07:002017-10-09T05:01:57.907-07:00Thank you so much Prof Giles! Thank you so much Prof Giles! DCRsilverhttps://www.blogger.com/profile/06356638588580939810noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-57454725817673912902017-10-08T07:11:03.981-07:002017-10-08T07:11:03.981-07:00"Yes" to the second part of your second ..."Yes" to the second part of your second question too. :-)Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-14104419000885461982017-10-08T07:07:51.179-07:002017-10-08T07:07:51.179-07:00Yes, absolutely - you can see this illustrated in ...Yes, absolutely - you can see this illustrated in equations (1) and (2), in fact. And it will always hold for any X and Y because the R-squared in such cases is just the square of the sample correlation between X and Y.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.com