tag:blogger.com,1999:blog-2198942534740642384.comments2018-06-20T18:31:09.930-07:00Econometrics Beat: Dave Giles' BlogDave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comBlogger3915125tag:blogger.com,1999:blog-2198942534740642384.post-33900286383420027232018-05-06T12:57:21.488-07:002018-05-06T12:57:21.488-07:00Yes.Yes.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-31494398283025507932018-05-06T10:47:02.420-07:002018-05-06T10:47:02.420-07:00Hello my Professor, thank you...
So it's j...Hello my Professor, thank you...<br /> So it's just confirmed:<br /> If K = 1, and the maximum order of integration dmax = 2 (because I have a variable I(2)), then I can use the VAR(1 + 2). i.e. VAR(3)?<br /><br />Best wishesecointelligencyhttps://www.blogger.com/profile/09764228437124629520noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-51460843661506108322018-05-06T10:38:24.042-07:002018-05-06T10:38:24.042-07:00Hello my Professor, thank you for your message...
...Hello my Professor, thank you for your message...<br /><br /> So it's just confirmed:<br /> If K = 1, and the maximum order of integration dmax = 2 (because I have a variable I(2)), then I can use the VAR (1 + 2)?<br /><br />Cordially<br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-85144828088742727472018-05-05T15:01:39.465-07:002018-05-05T15:01:39.465-07:00Diego - that's fantastic!!!! I'm so glad i...Diego - that's fantastic!!!! I'm so glad it worked out for you. DGDave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-33105033656277756092018-05-05T13:15:21.887-07:002018-05-05T13:15:21.887-07:00Thank you for your prompt response. I followed you...Thank you for your prompt response. I followed your suggestion to open the program file with a text editor and it worked. It was possible to manually execute the commands and I finally got my result (my model is fine). It took some time to wrap my head around the process but I've hopefully learned one thing or two!Diegonoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-86864140865536759032018-05-04T11:44:07.426-07:002018-05-04T11:44:07.426-07:00Diego - you won't be able to perform these tes...Diego - you won't be able to perform these tests in EViews without being able to run a program file. Usually, I wouldn't make the following offer, but if you want to send me your EViews workfile with the data in it, I'll run it for you. Send it dgiles@uvic.ca BTW you can open EViews program files with any text editor (e.g., Notepad). So, if you go to my website here - http://web.uvic.ca/~dgiles/downloads/binary_choice/index.html<br />you can at least look at the program files to see the code that I have written. These files have various comments in them to explain what is going on.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-76067244478801534712018-05-04T10:13:17.461-07:002018-05-04T10:13:17.461-07:00Dear Professor Giles, thank you for these posts. I...Dear Professor Giles, thank you for these posts. I went to your website but I'm still an undergraduate and I only have a student version of EViews. This does not allow me to open program files. How could I perform tests for heteroskedasticity and functional form in EViews? I want to investigate a logit with a binary response, 0 or 1. I have a constant and three independent variables with about 500 values in each of them. Thank you in advance.Diegonoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-65409367706168120232018-05-03T05:59:53.649-07:002018-05-03T05:59:53.649-07:00The answer is right here in the abstract for the T...The answer is right here in the abstract for the T-Y paper:<br />"This paper shows how we can estimate VAR’s formulated in levels and test general restrictions on the parameter matrices even if the processes may be integrated or cointegrated of an arbitrary order. We can apply a usual lag selection procedure to a possibly integrated or cointegrated VAR since the standard asymptotic theory is valid (as far as the order of integration of the process does not exceed the true lag length of the model). Having determined a lag length k, we then estimate a (k + dmax)th-order VAR where dmax is the maximal order of integration that we suspect might occur in the process.<br />The coefficient matrices of the last dmax lagged vectors in the model are ignored (since these are regarded as zeros), and we can test linear or nonlinear restrictions on the first k coefficient matrices using the standard asymptotic theory." <br /><br />Note that "the true lag" that is referred to in the passage in parentheses is NOT NECESSARILY "k". The latter is the lag that has been SELECTED empirically. In fact, we won't ever know the value of the "the true lag". However, this is not a big problem in practice. With economic data "dmax" will be a small number, either 0, 1, or 2. Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-78000342046773480072018-05-03T03:31:22.362-07:002018-05-03T03:31:22.362-07:00Good Morning, Hello,
please Professor, It's...Good Morning, Hello,<br /> please Professor, It's true this remark:<br />Toda-Yammamoto procedure requires that the maximum order of integration among the variables should not exceed the lags of the initial VAR?<br />Best wishesAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-18469551224621343202018-04-24T06:01:12.341-07:002018-04-24T06:01:12.341-07:00The Normality assumption is in fact needed and use...The Normality assumption is in fact needed and used for the bias and variance expressions. See the reference to Cramer's paper in this earlier post: http://davegiles.blogspot.ca/2013/05/good-old-r-squared.htmlDave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-56681866188642871172018-04-24T02:47:36.239-07:002018-04-24T02:47:36.239-07:00Great post! I was looking for some tips on which o...Great post! I was looking for some tips on which of the two, $R^2$ or $R^2_{adj}$ is a better estimator (say, in MSE sense) of the population $R^2$, and here it is. Of course, the really interesting case is when at least some (if not all) of the regressors truly belong in the model, but I understood that there is no general result (that would not depend on the true slope coefficients) for that case. On a side note, it does not seem you have made use of the normality assumption anywhere in the derivation. If so, why include it at all? And if you include it, a note on its irrelevance could be handy. It is my perception that too many people do not realize how little the normality assumption matters in deriving the standard results for OLS estimators. In any case, thank you for the great post!Daumantasnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-7776713601460276672018-04-21T15:12:25.621-07:002018-04-21T15:12:25.621-07:00If all of your series have been transformed approp...If all of your series have been transformed appropriately to make them stationary, then the 2SLS estimator will be consistent.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-43270320029831411882018-04-21T12:42:06.503-07:002018-04-21T12:42:06.503-07:00Thanks for your prompt reply. After transformation...Thanks for your prompt reply. After transformation, all of my variables are stationary at first difference with constant. is 2SLS correct to use now?Najia Mughalhttps://www.blogger.com/profile/17566253642073390422noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-56404896285251111282018-04-18T13:36:35.779-07:002018-04-18T13:36:35.779-07:00For 2SLS to be consistent, all of the data need to...For 2SLS to be consistent, all of the data need to be stationary - so, you will have to transform your variables before applying 2SLS.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-62239826225417496972018-04-18T13:10:45.106-07:002018-04-18T13:10:45.106-07:00Hello Sir,
Thanks alot for writing such a useful p...Hello Sir,<br />Thanks alot for writing such a useful post. Howevr, i have a question about the stationarity of the time series. I am trying to estimate an equation in which dependent variable is stationary at 2nd difference while all other variables are stationary at level or 1st difference. Will it be appropriate to use 2SLS on such data? if not, which technique should i move towards.<br />Najia Mughalhttps://www.blogger.com/profile/17566253642073390422noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-14903090633612920132018-04-14T12:00:41.640-07:002018-04-14T12:00:41.640-07:00Very thorough and clear, can you please talk more ...Very thorough and clear, can you please talk more about dynamic stability, I think that all stable processes are stationary, but vice-versa is not true.Amrendra Singhhttps://www.blogger.com/profile/08642319242703680042noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-22204432571862505402018-04-07T07:13:03.062-07:002018-04-07T07:13:03.062-07:00Sean - I had a lot of posts in 2013 - which one, s...Sean - I had a lot of posts in 2013 - which one, specifically, are you referring to?Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-70063347087553244022018-04-06T16:30:52.641-07:002018-04-06T16:30:52.641-07:00Hi Dave,
I have read your previous post in 2013 r...Hi Dave,<br /><br />I have read your previous post in 2013 regarding how you "tricked" eviews into generating multiple lags in order to choose the maximum lag. However, I have tried to replicate your method on eviews 10 and it did not work out. Can you give me some advice on this issue. Thanks!<br /><br />By the way, I am new to eviews.Seannoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-82984622044466609332018-04-04T13:32:02.003-07:002018-04-04T13:32:02.003-07:00Ivan - there could be all sorts of reasons. For in...Ivan - there could be all sorts of reasons. For instance: 1. Inappropriate choices of lag lengths in the ARDL model; 2. Same problem in the ECM; 3. Autocorrelation in errors of one or other of the 2 models; 4. Existence of unit roots at the seasonal frequencies (if, for instance, you are using quarterly or monthly data).Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-87316213662193873852018-04-04T12:10:06.654-07:002018-04-04T12:10:06.654-07:00Informative and useful!! It is pleasure to read su...Informative and useful!! It is pleasure to read such scientific masterpieces.Benyhttps://www.blogger.com/profile/13109569218146019654noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-26746969999785354412018-04-04T09:22:43.793-07:002018-04-04T09:22:43.793-07:00Great post, thanks a lot! More useful than much of...Great post, thanks a lot! More useful than much of my classesRafael Cattanhttps://www.blogger.com/profile/14094572834400556505noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-23225559008542655462018-04-04T01:18:01.051-07:002018-04-04T01:18:01.051-07:00Dear professor, I am using the Pesaran approach an...Dear professor, I am using the Pesaran approach and the general one to ECM I would like to know and I don't understand why I get a non significant Bound test in the former but a very significant and negative ECT in the latter. Thank you very much.Iván Sotohttps://www.blogger.com/profile/01422968472273266007noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-41214770103888853732018-04-02T13:10:58.111-07:002018-04-02T13:10:58.111-07:00Thank you Dr Giles! I have learned a lot from your...Thank you Dr Giles! I have learned a lot from your blog. It's very generous of you to share these materials. Nguyen Hue Lienhttps://www.blogger.com/profile/12841037751956118328noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-54286874195551399622018-03-19T13:17:44.645-07:002018-03-19T13:17:44.645-07:00Thanks - sorry the posts haven't been as regul...Thanks - sorry the posts haven't been as regular lately.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-83047944555018495912018-03-19T13:14:09.504-07:002018-03-19T13:14:09.504-07:00Hi, I read your blogs on a regular basis. Your hum...Hi, I read your blogs on a regular basis. Your humoristic style is awesome, keep up the good work!Anonymousnoreply@blogger.com