tag:blogger.com,1999:blog-2198942534740642384.comments2015-10-05T12:24:24.752-07:00Econometrics Beat: Dave Giles' BlogDave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comBlogger3026125tag:blogger.com,1999:blog-2198942534740642384.post-64082587287988470942015-10-05T12:01:19.552-07:002015-10-05T12:01:19.552-07:00I had always thought the temptation with missing d...I had always thought the temptation with missing data was to impute missing values of independent variables, so that one does not have to drop a large number of rows. My understanding is that there are a number of open questions in econometrics as to how best to do this, e.g. for principal components analysis of large datasets.Evan Soltashttp://www.blogger.com/profile/06212305798151301158noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-84523223387119710882015-10-05T11:47:34.607-07:002015-10-05T11:47:34.607-07:00Uh, yes - that's what this post was about.Uh, yes - that's what this post was about.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-56488146653504229402015-10-05T11:13:39.785-07:002015-10-05T11:13:39.785-07:00hi prof.
can we use ardl model with two variable c...hi prof.<br />can we use ardl model with two variable caseNAVEED KHANhttp://www.blogger.com/profile/18085660321479828126noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-2953469971271355352015-10-05T08:48:50.499-07:002015-10-05T08:48:50.499-07:00Dear Dr. Giles,
I started to wonder about testing...Dear Dr. Giles,<br /><br />I started to wonder about testing OLS assumptions namely homoskedascicity and linear functional form. Two usual tests are Breush-Pagan and LINK test respectively. If we denote residuals from OLS as "e" and fitted values as "y_hat" the two tests use additional regressions:<br /><br />e^2 = alpha_0 + X*alpha + u (for BPT)<br /><br />and<br /><br />e = alpha_0 +y_hat*alpha1 + y_hat^2 *alpha2 + y_hat^3 *alpha3 + y_hat^4 *alpha4 + u (for LINK test)<br /><br />above test specifications are very similar (these are specifications from NLOGIT software I use) and therefore it seems to me that two effect maybe easily cofounded. How can I know which one is the real problem in given dataset? Is there any formal procedure to test i.e. both assumptions jointly?<br /><br />I figured that Quantile regression is a very easy way to detect heteroskedascicity, but i didn't found any formal tests which use it. Are there any? (and why not? - is it bad idea?)<br /><br />Best regards,<br /><br />WiktorUnknownhttp://www.blogger.com/profile/02837157963575983813noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-74301261748366844192015-10-04T08:12:34.217-07:002015-10-04T08:12:34.217-07:00Achim - no, the problem definitely doe NOT go away...Achim - no, the problem definitely doe NOT go away. The time variable addresses any (linear) "deterministic" trend in the data. Unit roots introduce "stochastic" trends - something entirely different. Indeed, if you see any regression where time is a regressor, be awfully suspicious of the results.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-52859523694503645162015-10-04T07:40:39.337-07:002015-10-04T07:40:39.337-07:00Hi Dave, thanks for your illustrations. It's l...Hi Dave, thanks for your illustrations. It's long ago since I had a time series course. Could you please tell me whether this problem is done away with if I regress not only on the other time series but also on time as a variable? <br /><br />Sorry if this comment appears multiple times; when I click "publish" it just disappears.Achimnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-79642047483464009392015-10-04T01:50:05.347-07:002015-10-04T01:50:05.347-07:00I am doing research on the topic of causal relatio...I am doing research on the topic of causal relationship between public debt and GSDP and public debt and total expenditure. While, i doing the calculation, co-integration is found in the data, but doing granger causality test, p value is found more than 5 %. It means there is no causal relationship between these variables. I have studied in the literature that, if there is co-integration in the time series data between different variables, then there must be granger causality either one way or both way. Now but is happening in my case. Is the literature i have studied is wrong or there is something that i don't know. Please suggest something, if there is no Granger causality, then what is the next method i can use. My PhD is on last stage, so what is the next best method for me the study the causal relationship.Rakesh Pushphttp://www.blogger.com/profile/02440729355082394541noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-14422625815680429782015-10-03T08:37:03.536-07:002015-10-03T08:37:03.536-07:00Not sure that I can. Maybe another reader can help...Not sure that I can. Maybe another reader can help?Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-28631376441658135762015-10-02T14:09:47.598-07:002015-10-02T14:09:47.598-07:00yesterday I started reading "Causality...&quo...yesterday I started reading "Causality..." and today I found this-nice!<br /><br />I have a question. Will reading the book equip me with the proper tools to illustrate casual relationships as in "Event E (eg the 2007-2008 crisis) caused variables V1, V2, ...., Vn to alter their course but not so for vars Vn+1 to Vn+m"?<br /><br />ps. I'll read the book irrespective of the answer<br /><br />-john-<br />Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-61550306222794574212015-10-01T23:14:46.614-07:002015-10-01T23:14:46.614-07:00Can you put this on the context of multiple imputa...Can you put this on the context of multiple imputation, which accounts for the uncertainty in y*_n? There's a lot of literature arguing for it, but the intuition has always escaped me. JChttp://www.blogger.com/profile/10619261817794851130noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-3679720253157145102015-10-01T12:05:57.719-07:002015-10-01T12:05:57.719-07:00You have to be careful, yes. The standard errors m...You have to be careful, yes. The standard errors may be greater or smaller, and the same is true of the R-squared. You can easily check this with an empirical example.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-67980344660107017682015-10-01T09:40:27.926-07:002015-10-01T09:40:27.926-07:00In addition, one has to be careful interpreting th...In addition, one has to be careful interpreting the variance of the estimated regression coefficient(s) and the R-squared if imputation was done before running the estimation. In your example I suspect the R-squared would be spuriously higher and the variance of beta would be lower than it should be.Daumantasnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-66043868685209818902015-09-28T03:47:44.666-07:002015-09-28T03:47:44.666-07:00Hello Prof. Giles,
Your blogs and codes are indee...Hello Prof. Giles,<br /><br />Your blogs and codes are indeed very useful. I got a question regarding the structural break point.<br /><br />Which model is used to identify the structural break point in this case? If you have a blog/codes related to it, kindly guide. <br /><br />Thanks in anticipationSatya Ranjan Sahoohttp://www.blogger.com/profile/02677971783175312223noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-9168593141149478862015-09-26T04:04:06.035-07:002015-09-26T04:04:06.035-07:00Hello! i am studying the impact of trade openness ...Hello! i am studying the impact of trade openness on manufacturing growth from 1967 to 2013 in Tunisia. my variables are Manufacturing value added(MVA), openness(OPEN), GFCF and manufacturing labor force (MLAB)<br />1/ MVA (lag1) and MLAB (lag4) are I (0) no constant , no trend<br />MVA, OPEN and GFCF are stationary at first difference and satisfy the 3 equations of ADF test but MLAB at first difference is only stationary when I include constant and trend. can I claim that MLAB is stationary at first difference (i mean is not I (2)) ??can I run ARDL model in this case?<br />2/I include trend and intercept in my ARDL model but their coefficients are not significant, what should I do? can delete them and estimate again my model? in this case what critical value I should consider for my ARDL model? unrestricted intercept and unrestricted trend //or no intercept and no trend because their coefficient are not significant..what best software I can use?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-78583850225142336342015-09-24T09:04:10.964-07:002015-09-24T09:04:10.964-07:00You just estimate along with all of the other coef...You just estimate along with all of the other coefficients.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-1703430042693819342015-09-23T21:32:47.730-07:002015-09-23T21:32:47.730-07:00How to find coefficient of Dummy variable manually...How to find coefficient of Dummy variable manually? I am confused. How do we add it in ALPHA to shift the curve?Muhammad Afzal Bughiohttp://www.blogger.com/profile/02669742197045169032noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-72255025598428889252015-09-23T12:22:25.745-07:002015-09-23T12:22:25.745-07:00Hello Dr. Giles,
I have recently had cause to con...Hello Dr. Giles,<br /><br />I have recently had cause to consider the implications of variance in variable effect estimates in probit models, which subsequently are to be used in prediction. I discuss the problem in more detail here (http://bit.ly/1JqhLLL), but in a nutshell, when estimating the probability that Y=1 on new data, is it reasonable to give equal weight to beta1 and beta2 if they have very different variances (even if the effect magnitude is identical)? Any insight would be greatly appreciated.Marvin Ward Jr.http://www.blogger.com/profile/14745665785616485338noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-31758431698735216352015-09-22T18:59:57.626-07:002015-09-22T18:59:57.626-07:00My writing's not THAT bad, is it? :-)My writing's not THAT bad, is it? :-)Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-10240741196800218032015-09-22T18:58:41.782-07:002015-09-22T18:58:41.782-07:00is there an interpreter for hieroglyphics?is there an interpreter for hieroglyphics?Not Trampishttp://nottrampis.blogspot.com.aunoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-49497894185196357582015-09-22T08:07:35.887-07:002015-09-22T08:07:35.887-07:00Thanks for that, Nick! :-)Thanks for that, Nick! :-)Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-67391537176212428292015-09-22T04:27:41.983-07:002015-09-22T04:27:41.983-07:00Back in grad skool, Randy Wigle invented the MVLS ...Back in grad skool, Randy Wigle invented the MVLS estimator, for cases like this. It stands for Minimum Variance Lucky Seven, and is calculated as B* = 7. Since Var(7) = 0, it works perfectly for h=1. Now, true, it *may* be biased, but we don't know whether or not it's biased without knowing the parameter B.Nick Rowehttp://www.blogger.com/profile/04982579343160429422noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-70925039580765395782015-09-21T14:28:02.744-07:002015-09-21T14:28:02.744-07:00Not quite - the answer is going to depend n an unk...Not quite - the answer is going to depend n an unknown parameter.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-46405369316496860642015-09-21T13:47:03.547-07:002015-09-21T13:47:03.547-07:00Hi Dave: Is the answer no because you don't kn...Hi Dave: Is the answer no because you don't know the bias of B* until after you solve for h ? So, it's a circular<br />problem ? Just a guess. Thanks.<br />mark leedshttp://www.blogger.com/profile/13213841692738932471noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-14517837335422211612015-09-19T09:37:29.094-07:002015-09-19T09:37:29.094-07:00I don't wish to sound rude, but I just don'...I don't wish to sound rude, but I just don't have time for this. Sorry!Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-11189532766261293422015-09-19T09:35:10.348-07:002015-09-19T09:35:10.348-07:00Reza - thanks for the concern - all is well - just...Reza - thanks for the concern - all is well - just exceptionally busy.Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.com