tag:blogger.com,1999:blog-2198942534740642384.comments2016-08-24T02:03:09.075-07:00Econometrics Beat: Dave Giles' BlogDave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comBlogger3418125tag:blogger.com,1999:blog-2198942534740642384.post-18430983078118487612016-08-22T04:47:54.101-07:002016-08-22T04:47:54.101-07:00That's the whole point about the ARDL bounds t...That's the whole point about the ARDL bounds testing. You can use a mixture of I(0) and I(1) variables - the "bounds" are for "all I(0)" and "all I(1)" extreme situations.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-83509891949952779262016-08-22T01:52:41.991-07:002016-08-22T01:52:41.991-07:00Dear Professor,
Thank you for your help in unders...Dear Professor,<br /><br />Thank you for your help in understanding this model, there are not that much information about how to proceed and understand what the results means and what to select.<br />I just have one question: if you see that the variable LOG_CRUDE has a unit root, why you do the ARDL model without the 1st difference of that serie but with the non-stationary serie? for this model is not requeriment for the series to be stationary? if not, what about spurious correlations? <br /><br />Thank you so much for your help.<br /><br />Regards.Ana Ospinahttp://www.blogger.com/profile/12695750078561390790noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-49738819919711540492016-08-21T07:16:45.931-07:002016-08-21T07:16:45.931-07:00Mutawakil - you can do this by adding appropriate ...Mutawakil - you can do this by adding appropriate dummy variables, depending on whether the breaks are in levels or trends. If all of your series are I(1), I presume you've tested for cointegration. If it's present then you'd want to consider using a VECM model. Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-32514848871902230942016-08-21T07:12:08.838-07:002016-08-21T07:12:08.838-07:00Hi Dave,
I am estimating a VAR with four variables...Hi Dave,<br />I am estimating a VAR with four variables. I have tested all the variables for a unit root, and the variables are all I(1). The unit root tests also revealed that all the variables contain structural breaks but the breaks occurred at different dates. Is there a way that I can account for the different structural breaks in the VAR?<br />Thank you Mutawakil Zankawa Mumunihttp://www.blogger.com/profile/15658280362325534533noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-5150668749057037392016-08-17T21:55:14.893-07:002016-08-17T21:55:14.893-07:00Inspired Inspired INDIADIDIhttp://indiadidi.blogspot.in/2016/08/latest-at-india-didicom-bizz-buzz-india.htmlnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-74334430569847916082016-08-14T08:49:51.839-07:002016-08-14T08:49:51.839-07:00You can just use the ARDL model (estimated using O...You can just use the ARDL model (estimated using OLS), just as in this and my other posts.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-11812394535598972112016-08-12T13:22:23.033-07:002016-08-12T13:22:23.033-07:00Dear Prof, I have 6 variables. 4 of them are I(1) ...Dear Prof, I have 6 variables. 4 of them are I(1) including dependent variable and 2 are I(0). In this condition, can I use OLS estimation method? If yes, how can I use data?Pi Sinoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-28193812499828396842016-08-03T16:31:54.204-07:002016-08-03T16:31:54.204-07:00Wow,just learnt that today. I've been battling...Wow,just learnt that today. I've been battling with how on earth we got 52weeks in a year instead of 48. Thanks.Joylyn.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-73271769084206072092016-08-03T04:34:11.159-07:002016-08-03T04:34:11.159-07:00Found several courses concerning this topic on a c...Found several courses concerning this topic on a course. I think econometrics is an extremely useful instrument for social scientists. Especially in our age when so much data was collected. I admire a lot Piketty's work an global wealth and distribution. It was only possible because we have such effective instruments for huge data processing. Recently started working with <a href="http://www.databaseassignmenthelp.net/" rel="nofollow">databases</a> in SQL, hope would understand everything easily.Markus Millerhttp://www.blogger.com/profile/13112832724154446594noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-15496657135279793982016-08-02T12:42:59.945-07:002016-08-02T12:42:59.945-07:00Romian - you can just estimate each equation by OL...Romian - you can just estimate each equation by OLS. You can improve the estimator efficiency by using the Seemingly Unrelated Regression estimator.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-79250895293889608322016-08-01T13:42:56.597-07:002016-08-01T13:42:56.597-07:00Professor Giles,
Let's say I have a system of ...Professor Giles,<br />Let's say I have a system of equations with no contemporaneous endogenous variable (I only have exogenous variables and lagged endogenous variables as regressors). In that case, do I have to run the methods you dwell on (2SLS/3SLS...)?<br />Thanks<br />RomainRomain Faquethttp://www.blogger.com/profile/08353321882158361012noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-54472038711748357012016-07-28T06:32:12.176-07:002016-07-28T06:32:12.176-07:00"I plan to illustrate the application of seas..."I plan to illustrate the application of seasonal unit root and cointegration tests in a future blog post." ....looking forward to itAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-37270524448354202412016-07-26T17:01:04.315-07:002016-07-26T17:01:04.315-07:00Thanks for sharingThanks for sharingUnknownhttp://www.blogger.com/profile/04905815734623694896noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-70472511033350219792016-07-25T17:45:22.648-07:002016-07-25T17:45:22.648-07:00Daniel - yes it does. If you square the t-statisti...Daniel - yes it does. If you square the t-statistic you just get the Wald statistic for the case of one restriction. Follow your intuition and ass lags. The statistical "cost" of including more lags than you should is far less than the "cost" of including too few lags.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-57430398590176807522016-07-25T17:35:49.315-07:002016-07-25T17:35:49.315-07:00Dear Professor,
You mention that "Don't ...Dear Professor,<br /><br />You mention that "Don't use t-tests to select the maximum lag for the VAR model - these test statistics won't even be asymptotically std. normal if the data are non-stationary, and there are also pre-testing issues that affect the true significance levels."<br /><br />Is this also valid for Wald tests for joint significance of lags?<br /><br />For the sake of curiosity, if standard selection criteria recommend using 0 lags but this <br />conflicts with one's economic intuition, how should one go about it? Do not estimate the VAR at all, or still apply one's intuition? <br /><br />Thank you so much for the excellent blog! It is a source of learning and inspiration?<br /><br />DanielDaniel Pintonoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-90936761658352157702016-07-25T12:37:02.971-07:002016-07-25T12:37:02.971-07:00Often, we can predict the X variables using an ARI...Often, we can predict the X variables using an ARIMA model.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-75904930619330284142016-07-25T07:09:42.700-07:002016-07-25T07:09:42.700-07:00Dear Dave,
Thanks for the insightful explanation!...Dear Dave,<br /><br />Thanks for the insightful explanation! Can you elaborate some other ways of x variables in the forecasting process other than "guess"? <br /><br />LeeAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-36262725795459393162016-07-23T05:16:49.752-07:002016-07-23T05:16:49.752-07:00very useful to understand the concept of consisten...very useful to understand the concept of consistencyDhanasekaran Kuppuswamihttp://www.blogger.com/profile/15038889931233353679noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-63185272653096384282016-07-21T07:50:11.215-07:002016-07-21T07:50:11.215-07:00You should not use the model. Almost certainly the...You should not use the model. Almost certainly the maximum lag lengths have been mis-specified.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-60983008168107324012016-07-21T06:35:30.719-07:002016-07-21T06:35:30.719-07:00Dear Professor Giles,
I have benefited greatly fo...Dear Professor Giles,<br /><br />I have benefited greatly form your blogs. THANK YOU! My question is related to the question above - if my model is great in every aspect (significant and correct signs) but the coefficient of CointEq(-1) is -1.56 (which is over-correcting), what do you think may be happening? What do you mean by "over correcting" and what may be causing it? Can I use that model still or would it be inappropriate to keep that model? <br />I look forward to your response. Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-60173323548065345522016-07-13T00:15:54.399-07:002016-07-13T00:15:54.399-07:00..good work
..good work<br />Samuel Waiguruhttp://www.blogger.com/profile/13682091962586298809noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-87354841901181720582016-06-26T20:58:15.349-07:002016-06-26T20:58:15.349-07:00Hi, I am new to Panel VAR. I have balanced panel d...Hi, I am new to Panel VAR. I have balanced panel data for 10 years, 52 countries where I have one dependent variable and other 5 independent variables. I have already finished the panel regression estimation. As I want to do Panel Var estimation, shall I use all variables or use only those variables (2 or 3) on which I am more interested. for example, I am interested to see the impact of X2 and X3 on Y but I have other X1, X4 and X5 control variables. In doing Panel VAR shall I use all 6 variables or only Y X2 X3?<br /><br />Again, if one of my variable is non-stationary and others are stationary, what shall I do?<br /><br />Can anyone please help me? Subornohttp://www.blogger.com/profile/11265478885921619107noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-90169602942953989382016-06-22T11:32:41.928-07:002016-06-22T11:32:41.928-07:00You are using too many lags and/or too many regres...You are using too many lags and/or too many regressors. IN this respect an ARDL model is just like any other regression model.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-15655648329921093442016-06-22T11:21:21.833-07:002016-06-22T11:21:21.833-07:00HI Katherine - nothing changes in this situation. ...HI Katherine - nothing changes in this situation. The key thing is the log of the dependent variable.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-48380469755213723092016-06-22T11:18:37.987-07:002016-06-22T11:18:37.987-07:00That's right - if it's more negative than ...That's right - if it's more negative than -1 then the adjustement process is "over-correcting".Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.com