tag:blogger.com,1999:blog-2198942534740642384.comments2015-03-02T10:01:18.419-08:00Econometrics Beat: Dave Giles' BlogDave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comBlogger2580125tag:blogger.com,1999:blog-2198942534740642384.post-71320258064662122472015-03-02T10:01:18.419-08:002015-03-02T10:01:18.419-08:00If all of the series are I(o) then they can't ...If all of the series are I(o) then they can't be cointegrated, by definition.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-76994207692854552472015-03-02T00:36:41.248-08:002015-03-02T00:36:41.248-08:00Thank you very much professor Giles for such a det...Thank you very much professor Giles for such a detailed explanation of the granger test. In step 6 you have “If two or more of the time-series have the same order of integration, at Step 1, then test to see if they are cointegrated, preferably using Johansen's methodology (based on your VAR) for a reliable result.”<br />But If this same order of integration is 0, then I believe the co-integration test is not needed right? Since both series are stationary. Thank you<br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-40650297652948448592015-03-01T12:05:59.495-08:002015-03-01T12:05:59.495-08:00Yaser - that's right. It's a classic &quo...Yaser - that's right. It's a classic "pre-test testing" problem. The reason is that the various test statistics are not independent.<br /><br />Asymptotically, this effect should not be substantial, but in finite samples it can be a serious issue, the extent of which has not hasn't been fully explored.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-6811286810583334222015-03-01T11:48:18.141-08:002015-03-01T11:48:18.141-08:00Dear Prof Giles,
Thanks a lot for your great note...Dear Prof Giles,<br /><br />Thanks a lot for your great note and sample. I have a question:<br /><br />To do our g-causality we have to do some requisite test like unit root test, cointegration test, stability test, and etc. each of these tests has their own error in estimation and in each step the errors will transform to the next steps until we reach to final step to use Wald test for causality.<br /><br />I would like to know how we can handle this errors in each steps and its accumulation ?<br /><br />RegardsAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-35190792027568884302015-02-27T14:57:35.625-08:002015-02-27T14:57:35.625-08:00You might be interested in this study on measuring...You might be interested in this study on measuring treatment for tax policy analysis by Caroline Weber: <br /><br />http://pages.uoregon.edu/cweber5/Weber_FBATE.pdf<br /><br />Caroline recognizes that the measurement error associated with the traditional ATE definition of a marginal-tax-rate change is correlated with the instruments, and opts for an ITT definition of treatment.Jakenoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-21093377815236250922015-02-26T09:55:11.417-08:002015-02-26T09:55:11.417-08:00Casey - sorry if I drive you to despair. As you co...Casey - sorry if I drive you to despair. As you correctly noted, in your last paragraph, the ARDL post is designed to show readers how to implement a particular technique using a particular package. That's all!<br />You make many points that would be valid if the post was a journal article and you were a referee. That's not the case, and I'm sure that the majority of readers know that all too well. However, thanks for your comments.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-5225013389913075522015-02-26T09:52:21.922-08:002015-02-26T09:52:21.922-08:00Casey: I must say that when this comment arrived, ...Casey: I must say that when this comment arrived, I was left puzzled. What on earth could it mean? I do get a LOT of really CRAZY comments. To answer your question (rthetorical or not) - yes, I AM serious.<br /><br />Then it occurred to me that you "moderator-snipping" remark may refer to another comment that you may have sent previously. So, I searched the "junk" box, and sure enough there was another comment from you (relating to this and another post) that had wrongly been given a junk assignation. I apologise for that. You'll find that your earlier comment, with my response is now there. Thanks for your interest.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-57205353509168298092015-02-26T09:32:47.422-08:002015-02-26T09:32:47.422-08:00Interesting study. Thanks for sharing!Interesting study. Thanks for sharing!Arturo Bujandahttp://www.blogger.com/profile/00186386138216582437noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-41943844281903396772015-02-25T12:56:58.107-08:002015-02-25T12:56:58.107-08:00Yes, I believe so.Yes, I believe so.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-1441657817910136622015-02-25T02:34:53.540-08:002015-02-25T02:34:53.540-08:00Dear David,
I came across your post looking for ...Dear David, <br /><br />I came across your post looking for an answer to the question if the robust standard errors (Wooldridge suggests in 13.8.2.) are correct without assuming strict exogeneity?<br /><br />To be more precise, is it sufficient to assume that:<br /><br />(1) D(y_it|x_it) is correctly specified and<br />(2) E(x_it|e_it)=0 (contemporaneous exogeneity)<br /><br />in the case of pooled Probit, for 13.53 (in Wooldridge p. 492) to be applicable?<br /><br />Thanks!Tobiasnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-13538030389920240782015-02-24T12:01:52.681-08:002015-02-24T12:01:52.681-08:00I am confused about how you get the impulse respon...I am confused about how you get the impulse response function from a single equation framework? Don't we need a square matrix of impact multipliers to calculate the impulse response functions?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-15967023245152179472015-02-24T08:24:08.143-08:002015-02-24T08:24:08.143-08:00I'm assuming that both series are I(1). In tha...I'm assuming that both series are I(1). In that case you should fit a VAR in the levels of the two original variables, determine the optimal lag length, and then go from there to test for cointegration.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-67309729055411895002015-02-24T02:29:54.564-08:002015-02-24T02:29:54.564-08:00Dear Professor,
I am testing two series which th...Dear Professor,<br /><br /> I am testing two series which them Patent app. and Export figures of some country. I am totally confused about Johansen cointegration test about using variable lag (for instance Pat (-1)). Can I use lag of patent variable while testing johansen cointegration.<br /><br />Moreover We are very thankfull sharing your knowledge and experiences with us.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-81645418786516435832015-02-23T15:29:27.418-08:002015-02-23T15:29:27.418-08:00You say: "In fact, it seems that I'm enco...You say: "In fact, it seems that I'm encountering more and more of this nonsense. This isn't "econometrics", and the purveyors of this rubbish aren't "econometricians. My real concern is that students who are exposed to these papers and seminars may not recognize it for what it is - just ad hoc empiricism."<br /><br />Are you serious? Maybe refer to my moderator-snipped comments on your ARDL blog.<br /><br />Keep morale high.<br />CaseyAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-51813818143080507442015-02-20T20:19:43.109-08:002015-02-20T20:19:43.109-08:00Thanks! Your posts are truly helpful!Thanks! Your posts are truly helpful!DJ Jeongnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-53017175167490637042015-02-20T14:37:16.167-08:002015-02-20T14:37:16.167-08:00Thanks - not quite the same. The DFBETAS are looki...Thanks - not quite the same. The DFBETAS are looking at the influence of each sample observation, rather than each regressor. More coming on diagnostics.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-85792167512345337322015-02-20T14:06:01.668-08:002015-02-20T14:06:01.668-08:00Great post, as always! Are partial residuals relat...Great post, as always! Are partial residuals related to DFBETAs? <br /><br />I look forward to your post about using the residuals to help understand the errors; this is something I struggle with quite a bit. I'm sure this question will reflect my ignorane, but how can we "test" for anything unobservable? For example (while not required for the Gauss-Markov theorem to hold), I've seen people plotting residuals to check if the assumption of normal errors holds. Why is this reasonable, because we assume a representative random sample? Thanks!Nickhttp://www.blogger.com/profile/16825362414374172386noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-7601849176419290592015-02-20T09:20:43.409-08:002015-02-20T09:20:43.409-08:00I have no problems with it. Why not contact the au...I have no problems with it. Why not contact the authors of the package?Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-48372794805569596672015-02-20T08:49:50.072-08:002015-02-20T08:49:50.072-08:00After installing the package I did the library(MVN...After installing the package I did the library(MVN) and got the following error: <br /><br />Error in loadNamespace(i, c(lib.loc, .libPaths()), versionCheck = vI[[i]]) : <br /> there is no package called ‘digest’<br />Error: package or namespace load failed for ‘MVN’<br /><br />Is this usual?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-10878977924615285532015-02-20T07:21:05.602-08:002015-02-20T07:21:05.602-08:00Thank you, Dave.Thank you, Dave.Daniel Hendersonhttp://mycba.ua.edu/~djhendernoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-2253416659309747542015-02-19T20:28:26.229-08:002015-02-19T20:28:26.229-08:00Daniel - thanks. I've re-phrased my last comme...Daniel - thanks. I've re-phrased my last comment to clarify this (I hope)!Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-31993143878859719162015-02-19T18:42:07.535-08:002015-02-19T18:42:07.535-08:00Thanks a lot for recognizing our book. We are ver...Thanks a lot for recognizing our book. We are very excited about our book being released.<br /><br />We do have one point of clarification. While our programming code has many of the same functions that are in the np package, there are many that are not in the np package (similarly, there are many functions in the np package that are not on our website). That being said, we wish we note that we wrote our code in such a way that it is easy to read (and hence learn) and modify, and thus may not be as efficient as the np package. Daniel Hendersonhttp://mycba.ua.edu/~djhendernoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-33152457450482542092015-02-19T08:36:00.958-08:002015-02-19T08:36:00.958-08:00In this case you should use TY. You can't use ...In this case you should use TY. You can't use ARDL if any of the data are I(2).Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-42183724959344532352015-02-19T02:53:13.015-08:002015-02-19T02:53:13.015-08:00Dear Dr.Giles! Thank you for your precious time yo...Dear Dr.Giles! Thank you for your precious time you spare for sharing your experiences and disseminating the knowledge about various issues in Econometrics. <br />Dr. David! if some of the data on variables are I(1), some is I(0) and some is I(2) , and the aim is to investigate the causality , then what do you suggest in such a case. Which technique is best suitable either VAR, VECM, ARDL, OR TY etc. <br />Please guide me in this regard. Appreciated if you could support the answer step by step with Eviews work-files. <br />I also request you that to write some blogs on forecasting techniques in E-views too.<br />Thanks in advance. Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-37730491801880484702015-02-19T01:07:26.886-08:002015-02-19T01:07:26.886-08:00please sir, can you give an example on how to go t...please sir, can you give an example on how to go through the nonlinear ARDL modelOyeleye Olalekanhttp://www.blogger.com/profile/13348286659686154092noreply@blogger.com