tag:blogger.com,1999:blog-2198942534740642384.comments2016-06-24T11:46:45.081-07:00Econometrics Beat: Dave Giles' BlogDave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comBlogger3396125tag: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.comtag:blogger.com,1999:blog-2198942534740642384.post-26822595267739699792016-06-22T11:16:10.825-07:002016-06-22T11:16:10.825-07:00Only the ones that have a break, but when it comes...Only the ones that have a break, but when it comes to cointegration testing in this situation yo'll need to allow for the break(s).Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-78352473755774877212016-06-22T10:37:21.227-07:002016-06-22T10:37:21.227-07:00Sir, that's a really useful post on Break Poin...Sir, that's a really useful post on Break Point test. <br />I have one query, if some of the variables in a regression equation are showing a structural break and some are not, should I use simple unit root test for the latter category or a break point test for all of them.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-56692382074770213742016-06-22T08:07:19.527-07:002016-06-22T08:07:19.527-07:00Dear Professor Giles,
Thank you for your fantasti...Dear Professor Giles,<br /><br />Thank you for your fantastic posts !<br /><br />I have a question regarding the “cointegrating form” of the model. If the coefficient of CointEq(-1) is negative but lower than -1, would the ECM between the two variables of interest still be validated ? Would it mean that there is no “stable” long-run relationships between the two variables?<br />According to the Engle & Granger and the Johansen methodologies, this coefficient must be negative but higher than -1<br /><br />Thank you so much for your reply<br /><br />Kind regards<br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-91303824294140808712016-06-21T10:12:17.640-07:002016-06-21T10:12:17.640-07:00Sorry for a stupid question, but what if we have a...Sorry for a stupid question, but what if we have a mixed log level model<br />let's say, log(y)=b0+b1*log(x1)+b2*x2+b3*log(x3), how do you forecast then?Katherine Nicanorovahttp://www.blogger.com/profile/06183649660347981639noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-41474193401323840552016-06-21T07:38:58.917-07:002016-06-21T07:38:58.917-07:00Hello proffesor, in trying to run cointegration te...Hello proffesor, in trying to run cointegration test in eviews 7 (my data are on gdp, cpi, personal consumption expenditure, exchange rate and intrest rate) I get a message that says 'Near Singular Matrix'. My time series data spans 31 years. What can I do about this? Thank you.<br />Patrick Onodjehttp://www.blogger.com/profile/10754363914893547654noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-32748788113175395872016-06-18T18:36:35.438-07:002016-06-18T18:36:35.438-07:00It's just like any other regression model. You...It's just like any other regression model. You're trying to use too many lags relative to the sample size.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-8201758274299450782016-06-18T08:10:40.464-07:002016-06-18T08:10:40.464-07:00Dear Sir,
I run the ARDL model and face the "...Dear Sir,<br /><br />I run the ARDL model and face the "singular matrix" error. How can i solve this?Bích Ly Lêhttp://www.blogger.com/profile/03198163453693374031noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-45725741819741964572016-06-18T06:01:24.490-07:002016-06-18T06:01:24.490-07:00There's no "canned" routine for this...There's no "canned" routine for this in EViews so you'll have to write some code. Also, check the EViews forum at eviews.comDave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-59964029512420018812016-06-18T04:22:46.315-07:002016-06-18T04:22:46.315-07:00Nothing comes to mind - sorry.Nothing comes to mind - sorry.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-75827455622901756472016-06-17T09:57:13.086-07:002016-06-17T09:57:13.086-07:00Hi Dave,
Are you aware of any software procedures ...Hi Dave,<br />Are you aware of any software procedures (codes) that perform seasonal cointegration test for monthly data?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-19574886168267788332016-06-15T23:50:55.340-07:002016-06-15T23:50:55.340-07:00dear professor,
thank you very muc...dear professor,<br /> thank you very much for your post i am working with panel data in eviews 9 as i have done with the ardl but please help me whether it is possible to test the ardl bound test for panel data in eviews vijay modihttp://www.blogger.com/profile/12996924290698932043noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-83145947824169552562016-06-11T08:25:51.584-07:002016-06-11T08:25:51.584-07:00Just do what you would do with any other regressio...Just do what you would do with any other regression model.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-38334987330934009242016-06-11T08:21:33.749-07:002016-06-11T08:21:33.749-07:00After running the ARDL model, I found that it is h...After running the ARDL model, I found that it is heteroskedastic, what should be done now?Parulnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-81947156805661443392016-06-06T17:58:46.086-07:002016-06-06T17:58:46.086-07:00But the cointegrating vector must be (1,-1)But the cointegrating vector must be (1,-1)Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-77073751255987249082016-06-06T07:09:17.529-07:002016-06-06T07:09:17.529-07:00With the negative values you simply can;t use logs...With the negative values you simply can;t use logs. You SHOULD NOT add a value to make the number positive - NEVER!Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-38208429098714785412016-06-06T03:55:12.058-07:002016-06-06T03:55:12.058-07:00Respected Sir,
Thank you so much for your pro...Respected Sir, <br /> Thank you so much for your prompt reply. It really helped me. I have one more query. You have suggested differencing of all the series irrespective of stationarity test but after differencing I am getting negative values to a large number. Now for log transformation, I need to add minimum positive value to almost all series. On the other hand if log transformation is followed by stationarity test, then differencing log transformed data will actually mean rate/ ratio. What do you suggest? I am in dilemma. In anticipation of your advice.<br /><br />Warm Regards<br /><br />ArjunArjunhttp://www.blogger.com/profile/14032931920180996766noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-72665676242091865822016-06-05T07:15:55.099-07:002016-06-05T07:15:55.099-07:00Yes, you can legitimately estimate an ARDL model i...Yes, you can legitimately estimate an ARDL model in this case, but it's not really the right basis for Granger causality testing.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-50061517365928779112016-06-05T06:42:08.459-07:002016-06-05T06:42:08.459-07:00Hello Sir
Its great to read your post.
Although y...Hello Sir<br /><br />Its great to read your post.<br />Although you have made clear regarding the integration order of the variables, but I am still confused that whether I can use ARDL approach if both of my variables are stationary at level?<br />I tried using VAR & Granger but both of them are giving different lags significant, so hot struck & thought of using ARDL.<br />Kindly suggest.<br /><br />Thanks<br />ParulParulnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-53402437619015955232016-06-03T17:43:38.892-07:002016-06-03T17:43:38.892-07:00I'd discriminate between the two using one the...I'd discriminate between the two using one the of the available information criteria. A useful paper on this is: G. Chen & H. Tsurumi, "Probit and Logit MOdel Selection", Communications in Statisics - Theory & Methods, 2010, 40, 159-175. Here's the abstract:<br /><br />Abstract:<br />"Monte Carlo experiments are conducted to compare the Bayesian and sample theory model selection criteria in choosing the univariate probit and logit models. We use five criteria: the deviance information criterion (DIC), predictive deviance information criterion (PDIC), Akaike information criterion (AIC), weighted, and unweighted sums of squared errors. The first two criteria are Bayesian while the others are sample theory criteria. The results show that if data are balanced none of the model selection criteria considered in this article can distinguish the probit and logit models. If data are unbalanced and the sample size is large the DIC and AIC choose the correct models better than the other criteria. We show that if unbalanced binary data are generated by a leptokurtic distribution the logit model is preferred over the probit model. The probit model is preferred if unbalanced data are generated by a platykurtic distribution. We apply the model selection criteria to the probit and logit models that link the ups and downs of the returns on S&P500 to the crude oil price."Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-45456993936379379452016-06-02T13:09:31.102-07:002016-06-02T13:09:31.102-07:00Thank you! I have not seen this in any text.Thank you! I have not seen this in any text.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-24968454165927318922016-06-02T11:20:51.841-07:002016-06-02T11:20:51.841-07:00The Johansen results will be the superior ones, an...The Johansen results will be the superior ones, and I'd rely on those - as long as you have specified the underlying VAR model appropriately.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-70409893402150448782016-06-02T10:56:38.882-07:002016-06-02T10:56:38.882-07:00If X and Y are both I(1), then there is no reason ...If X and Y are both I(1), then there is no reason at all for (X/Y) to be I(0).<br />However, if log(X) and log(Y) are I(1) and cointegrated, then log(X)- log(Y) will be I(0). That is, log(X/Y) will be I(0).Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.com