tag:blogger.com,1999:blog-2198942534740642384.comments2014-10-01T07:55:28.535-07:00Econometrics Beat: Dave Giles' BlogDave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comBlogger2405125tag:blogger.com,1999:blog-2198942534740642384.post-52193737981010723732014-10-01T06:48:01.129-07:002014-10-01T06:48:01.129-07:00Thank you so much for prompt reply.Thank you so much for prompt reply.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-16106071833440762572014-09-30T22:57:51.469-07:002014-09-30T22:57:51.469-07:00Well, due to this specific procedure, I agree and ...Well, due to this specific procedure, I agree and I think it is nice to see that the procedure works. My point is just that a statistical package could estimate the model without the multicollinear variable, save the set of coefficients, then add the multicollinear variable, assign some arbitrary coefficient to it and adjust the the other coefficients accordingly given the results from the full-rank estimation. For meaningful interpretation, it must be done what you did in the last three bullet points anyway.Martin Sandersnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-65160328295650135422014-09-30T11:33:37.876-07:002014-09-30T11:33:37.876-07:00You need at least 2 I(2) variables for Haldrup'...You need at least 2 I(2) variables for Haldrup's results.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-44153519798768055922014-09-30T10:53:49.116-07:002014-09-30T10:53:49.116-07:00Once you've fitted the model, select "Vie...Once you've fitted the model, select "View", "ARMA Structure", "Roots", Graph".<br />This is all in "HELP".Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-63696096346969972812014-09-30T08:47:48.138-07:002014-09-30T08:47:48.138-07:00THey're unique if you are using the Moore-Penr...THey're unique if you are using the Moore-Penrose inverse.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-81919686789957931532014-09-30T05:43:26.230-07:002014-09-30T05:43:26.230-07:00Dear Prof. Giles,
Thanks for useful information. C...Dear Prof. Giles,<br />Thanks for useful information. Can you please give the procedure how inverse roots of AR/MA polynomials can be obtained and graphed in Eviews software ? I really appreciate your help.<br />Thanks.<br /><br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-3703707225441583392014-09-29T22:51:24.406-07:002014-09-29T22:51:24.406-07:00Very, very interesting, Prof. Giles, thank you. Th...Very, very interesting, Prof. Giles, thank you. Though I think the use for interpretation is somewhat limited as it seems to me that the estimate for $\beta_{3}$ could be any number as long as the other two betas are "adjusted" accordingly. I will take a look in Searle, however.Martin Sandersnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-47958511160130769652014-09-28T12:12:51.949-07:002014-09-28T12:12:51.949-07:00No - you need to use Johansen's method for tha...No - you need to use Johansen's method for that.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-28036830856098600722014-09-28T12:11:05.947-07:002014-09-28T12:11:05.947-07:00isARDL also a multivariate method for testing Co-i...isARDL also a multivariate method for testing Co-integrating relationship. if k = 4 can it give us 3 CE? Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-31252006047056850862014-09-28T08:43:11.934-07:002014-09-28T08:43:11.934-07:00Estimate a 6-equation VAR model in the levels of t...Estimate a 6-equation VAR model in the levels of the data and just use the usual Wald test. There is no need to use the modified test.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-71965421354590678482014-09-28T01:00:28.551-07:002014-09-28T01:00:28.551-07:00Dear Prof.
thank you for the excellent service t...Dear Prof. <br /><br />thank you for the excellent service that you provide to the community (industry and academic) by running this blog. It is very valuable. <br /><br />Follow up question on the above: what happens if you use, say, 6 times series of returns (stocks, forwards, options, etc) where all of them are stationary? How would you go about testing for granger causality among them?<br /><br />Thank youAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-21780704604375576732014-09-24T18:43:31.878-07:002014-09-24T18:43:31.878-07:00O.K. Thanks. O.K. Thanks. Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-85868290575160550392014-09-24T18:35:10.793-07:002014-09-24T18:35:10.793-07:00Can't be done - you have a singular covariance...Can't be done - you have a singular covariance matrix. You need more data or fewer equations.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-11288864054668045912014-09-24T18:28:45.872-07:002014-09-24T18:28:45.872-07:00Thanks Dave. I'm using 8 variables. And given ...Thanks Dave. I'm using 8 variables. And given sample size, I'm fixing the lag length at 1 when specifying the VAR model.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-39856093376433799212014-09-24T18:15:21.209-07:002014-09-24T18:15:21.209-07:00You don't say how many variables (& how ma...You don't say how many variables (& how many lags) you are using in the VAR. Obviously too many for this number of observations.<br />Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-14052113437464680902014-09-24T17:52:16.889-07:002014-09-24T17:52:16.889-07:00Hi Prof. Dave, my name is Isaac. I'm working ...Hi Prof. Dave, my name is Isaac. I'm working with time series data covering 17 years . I intend to use the data for time series analysis (including unit root test, Johansen/Bounds test for cointegration, ECM, impulse response functions etc). However, Eviews8 keeps giving me error message "near singular matrix" whenever I try Johansen test for cointegration. All the cointegrating estimators (FMOLS and DOLS) are also giving me similar messages. Please, what do you think is going wrong? Is the sample size not large enough for such an analysis?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-81762990483412141202014-09-21T14:45:39.054-07:002014-09-21T14:45:39.054-07:00No. If you have 2 variables, one I(1) and the othe...No. If you have 2 variables, one I(1) and the other I(2), they can't be cointegrated, BY DEFINITION. Also, ARDL bounds testing can't be used with I(2) data, but that's a different point.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-52503067866222326802014-09-21T14:19:34.971-07:002014-09-21T14:19:34.971-07:00For the sake of specification, my I(1) variable is...For the sake of specification, my I(1) variable is GDP per capita and my I(2) variable is Population for Mexico from 1960 to 2012. <br /><br />Logical sense dictates both variables could be cointegrated. I read that "hypothesis testing of the I(2) model" and/or running DOLS could be a solution for cointegration an I(1) variable with an I(2) variable. <br /><br />Is this correct? Thanks in advance, sir.Francisco RenterĂahttp://www.blogger.com/profile/03424205603263956700noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-61698868531444562142014-09-21T10:28:08.325-07:002014-09-21T10:28:08.325-07:00Only if the sample size is very large.Only if the sample size is very large.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-91283563075143749612014-09-21T10:20:38.297-07:002014-09-21T10:20:38.297-07:00So professor, just to be more clear over the "...So professor, just to be more clear over the "This is a result that applies only for very large T, and we still have non-standard distributions in the finite-sample case." : ONLY if we have very large T sample then we can use the DF test results in the case you mentioned above of including either an intercept (drift term) and / or a linear trend term in the equation? Or in any sample, no matter how large it is?Ralucanoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-67241952650149920062014-09-20T17:55:39.749-07:002014-09-20T17:55:39.749-07:00Hi Dave: I didn't know that either so thanks.Hi Dave: I didn't know that either so thanks.mark leedshttp://www.blogger.com/profile/13213841692738932471noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-84147844822820624312014-09-16T19:21:03.434-07:002014-09-16T19:21:03.434-07:00Dear Prof! You have helped a lot for the researche...Dear Prof! You have helped a lot for the researchers throughout the world. Hats off to you with my whole-heart. I am anxiously waiting for your next post for ARDL model in more endogenous varaibles. We all are watching your way for the post. May Allah pak bless you healthy life with prosperity! Thanks again Prof. Niaz Hussain Ghumrohttp://www.iba-suk.edu.pknoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-42841351764415038172014-09-16T15:40:09.092-07:002014-09-16T15:40:09.092-07:00Thanks for your quick response.
Actually, the issu...Thanks for your quick response.<br />Actually, the issue is more simple: my variables have different integration order:<br /><br />GDP per capita: I(1); Y<br />Population: I(2); X<br />Period: 1960-2012<br /><br />Would Haldrup (1998) still be valid?Francisco RenterĂahttp://www.blogger.com/profile/03424205603263956700noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-72801654437298912342014-09-15T07:59:29.484-07:002014-09-15T07:59:29.484-07:00You won't find them. The intercept should be r...You won't find them. The intercept should be retained, even if it is insignificant.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-87189093028041177002014-09-15T04:05:57.374-07:002014-09-15T04:05:57.374-07:00Professor Dave, thank you so much for posting such...Professor Dave, thank you so much for posting such a useful material for us to learn. I have a small question though. I am doing ARDL approach to cointegration and I reached to a situation of no intercept and unrestricted trend after removing highly insignificant variables. But I did not find F-table and t-table to compare my statistics with lower and upper bounds to use the bound test, I have very high F-stat (12.11) though. I know it is significant but still I want go in a formal way. Where can I get such tables? Dambar Upretynoreply@blogger.com