(EViews10): How to Perform GARCH Diagnostics

[@CrunchEconometrix]

I want to appreciate all my subscribers from across the globe (Africa, Asia, Europe, the Middle East, The Americas, and The Pacific). Thank you all for your support. I am encouraged by your comments, questions, likes and critiques. They keep me focussed and poised to do better. I will continue to contribute my little quota such that every student and researcher will independently analyse his/her data. My teaching approach is very practical. I adopt a do-as-I-do style. Many thanks to those who have supported me by telling others. Once again, CrunchEconometrix loves to teach, support my Channel with your subscription, likes, feedbacks and sharing my videos with your cohorts. Please do not keep me to yourself (lol) inform your friends, students and academic networks about my Channel. Tell them CrunchEconometrix breaks down the econometric jargons and teaches with simplicity. Follow me on Facebook, Twitter and Reddit. Love you all, greatly!!! J

[@CrunchEconometrix]

I want to appreciate all my subscribers from across the globe (Africa, Asia, Europe, the Americas, and The Pacific). Thank you all for your support. I am encouraged by your comments, questions, likes and critiques. They keep me focussed and poised to do better. I will continue to contribute my little quota such that every student and researcher will independently analyse his/her data. My teaching approach is very practical. I adopt a do-as-I-do style. Many thanks to those who have supported me by telling others. Once again, CrunchEconometrix loves to teach, support my Channel with your subscription, likes, feedbacks and sharing my videos with your cohorts. Please do not keep me to yourself (lol) inform your friends, students and academic networks about my Channel. Tell them CrunchEconometrix breaks down the econometric jargons and teaches with simplicity. Follow me on Facebook, Twitter and Reddit. Love you all, greatly!!! 

[@AlagieBSowe]

The Gambia is following too

[@rachelcxuan1174]

Thank you so much for the detailed video presented

[@CrunchEconometrix]

Thanks for the encouraging feedback, deeply appreciated!

[@johnotalor8441]

Dr you are simply amazing. Thanks for the video

[@CrunchEconometrix]

Thanks Mr. Otalor, for the encouraging feedback. Deeply appreciated!

[@johnotalor8441]

@@CrunchEconometrix Actually Dr. Otalor

[@CrunchEconometrix]

Apologies Doc!

[@johnotalor8441]

@@CrunchEconometrix You are welcome Doc

[@dahyeeeby]

Excellent lecture 👍👍👍

[@CrunchEconometrix]

My pleasure, Dahyee! Thanks for the encouraging feedback, deeply appreciated!

[@dahyeeeby]

@@CrunchEconometrix thank you for the video madam,☺️☺️

[@rahulbhirani4767]

You are a star! Thank you for the amazing econometrics content.

[@CrunchEconometrix]

You are welcome, Rahul🙏🥰

[@mohamedkkamara3836]

Your videos are extremely helping me to do my dissertation work

[@CrunchEconometrix]

Thanks for your encouraging feedback, Mohd🥰🙏

[@AlagieBSowe]

Keep up the good job Ma, its been helpful. Thanks a million

[@CrunchEconometrix]

Thanks for the encouraging feedback, Alagie. I'll keep doing my best. Please may I know from where (location) you are reaching me? Gracias!

[@AlagieBSowe]

@@CrunchEconometrix I'm a Gambian but reaching you out from the UK.

[@drsunilbhardwaj]

Thanks madam for such meaningful videos. I am an Indian

[@CrunchEconometrix]

Compliments humbly taken, Sunil. Kindly refer your students and colleagues to my Channel. Grateful!

[@Amal-galos]

I am Egyptian, I like your videos very much

[@CrunchEconometrix]

Thanks so much for your encouraging feedback, deeply appreciated! 🥰🙏

[@Amal-galos]

@@CrunchEconometrix Really,I liked econometrics because of you You made me understand it it very easily

[@michaelbaah7162]

mommy thank you very much for your videos God bless you from Michael (ghana)

[@CrunchEconometrix]

U're welcome, Michael 😊. Hope you are doing great! I'll appreciate it if you can share the link to my KZbin Channel with your friends and academic community in Ghana 🇬🇭. They will find the content helpful too. Thanks! ❤️

[@tobibakare7023]

Thank you very much for your video's, I really appreciate your help! If we just want to model volatility using GARCH and we have no intention to forecast, does our GARCH model need to pass the no heteroscedasticity & auto-correlation condition, or are the 2 criteria only relevant for forecasting? I am asking as my mean models show ARCH effect, however I get serial correlation and heteroscedasticity for my GARCH equation. Thank you from the UK!

[@CrunchEconometrix]

Hi Tobi, thanks for the positive feedback. Deeply appreciated! Always remember: the presence of ARCH effects or heteroscedasticity is the MAIN reason for estimating ARCH/GARCH models...and forecast may or may not be done. That depends on your research objectives.

[@tobibakare7023]

​@@CrunchEconometrix THANK YOU VERY MUCH!

[@anwarqahtan2750]

Thank you so much, it's very helpful. Please, we want another video that explains and describes step by step how to get rid and correct Heteroscadisity and serial correlation in different ways not only convert data to (log) but other methods in the panel data. Thank you.

[@CrunchEconometrix]

Hi Anwar, thanks for the positive feedback. Deeply appreciated! I'll do my best with respect to your request. Please may I know from where (location) you are reaching me?

[@anwarqahtan2750]

@@CrunchEconometrix from china. Thank you.

[@CrunchEconometrix]

@@anwarqahtan2750 Awesome, Anwar! I'll appreciate it if you can share the link to my KZbin Channel with your friends and academic community in China 🇨🇳 for awareness, thanks 😊.

[@estevaomcs]

Great job, CrunchEconometrix. I ask you if in eviews there Is a test like Durbin Watson test for testing the residuals? And how to execute it?

[@CrunchEconometrix]

Hi Estevo, thanks for the positive feedback...on DW test for residuals: not to my knowledge. You may want to post this on an EViews platform for more constructive feedback.

[@windijordan2213]

HI! could please tell me how to fix if in the Ljung-BOX result for serial correlation within GARCH models is written that: "Probabilities may not be valid for this equation specification"?? I cant rely on probabilities of Q-stat to recognize existance of serial correlation for GARCH models. Thank you

[@CrunchEconometrix]

Hi Windi, I have not experienced such. You may want to check out other online resources for more information. Thanks

[@vitorbuarque1081]

Hi! Can you tell me exactly which paper/textbook i can find the explanation to choose between models? You wrote: "From the literature, the preferred model(s) should have:" but i couldn't find it anywhere. help me please! greetings from brazil

[@CrunchEconometrix]

Hi Vitor, you can always check published articles and several resources. Some are listed at the end of my ARCH and GARCH videos. Thanks

[@kamariannisgr]

How do we react on autocorrelation existence on garch models? I mean that on the Correlogram -Q statistics there are 4 observations from the 4th to 8th that their Pvalue is lower than 5%. How can we explain that? Thank you..

[@CrunchEconometrix]

Watch my ARCH video on DIAGNOSTICS on what to do.

[@SureshKumar-qq1yk]

Which model is suitable for the analysis of volatility spillover between two variables?

[@CrunchEconometrix]

Hi Suresh, an empirical review of related studies will guide you appropriately. Please may I know from where (location) you are reaching me?

[@SureshKumar-qq1yk]

@@CrunchEconometrix Thank you!!

[@SureshKumar-qq1yk]

@@CrunchEconometrix I can't get there right now.

[@SureshKumar-qq1yk]

@@CrunchEconometrix I am from Punjab (India). based on the review I thing Muptiple GARCH model ie used for the spillover volatility among time series variables. Please explain it and share a new video on the MGARCH model in the Eviwes. Kumar et al. (2019) [www.sciencedirect.com/science/article/abs/pii/S0301420718306251?via%3Dihub] for Correlations and volatility spillovers between oil, natural gas, and stock prices in India have been used VARMA-DCC-GARCH model.

[@CrunchEconometrix]

@@SureshKumar-qq1yk I have a video on GARCH-M and not on "multiple GARCH".

[@nidhidhankhar8220]

Hii if error message" near singular matrix" is obtained then what should we do for forecasting forecasting

[@CrunchEconometrix]

Nidhi, "near singular matrix" points to the multicollinearity problem. Kindly watch my video on MULTICOLLINEARITY. It guides on how to resolve the issue. Thanks

[@carolinazunigajuul2217]

What if only the GARCH parameter is significant? I should drop the model?

[@CrunchEconometrix]

Hi Carolina, no....but you may need to re-estimate if you feel less confident about your result.

[@radhikajain14]

I have a question regarding diagnostics. In my data series, all three tests are showing positive results in all three kinds of the model but I am getting R square, adj R square, and the log-likelihood as negative. What should I do?

[@CrunchEconometrix]

Perhaps you did not input the variables correctly. In particular, the Rsq should not be negative.

[@ahah9211]

The question regarding lags. In case of daily date is it fixed that i try first 1 then 36 ? Or i can choose any number ? I got tired trying to get this information

[@CrunchEconometrix]

Watch my video on Optimal Lag Selection.

[@kevinalejandro3121]

What if I have a GARCH model where it has certan non-significant coefficients and has the lowest Schwartz IC comparing with other models, even in the adj. R squared is the highest too, but i have too a GARCH model where all of it's coefficients are significant. ¿In that case which one to choose ? I have to choose the one that has lowest Schwartz IC even it has non- significant coefficients ??

[@CrunchEconometrix]

Hi Kevin, my explanation is quite clear on what to do. I will advise you adapt the knowledge to your results as applicable. Thanks.

[@shubhamgarg9540]

Good evening mam, what will happened if our q statistics value are showing less than .05 values for most of lags but our correlogram of squared residuals shows most of values greater than .05 after running the garch model

[@CrunchEconometrix]

Shubham, please watch the clips on GARCH again and adapt the information to your study. My explanations are quite clear.

[@danielstalmeisters2528]

Hi Ngozi, I hope that you are well. Just (hopefully) a quick question, why does all empirical research use the ARCH LM and the Q-statistic test for OLS testing but not for testing GARCH-type models?

[@CrunchEconometrix]

Hi Dan, it is always good to use the word "all" with caution as your assertion is not true. There are tons of diagnostic tests and every researcher use the ones they are familiar with considering data type and empirical techniques. So, regarding your query you will have to find out the underlying assumptions about those tests.

[@blessingakarue9010]

Good morning just joining how do i access the last seven video on Garch model you talked about

[@CrunchEconometrix]

Hi Blessing, my videos are nicely sorted into Playlists which you can check through. You can also do a search on my Channel for those videos. Thanks.

[@immaculatelum5102]

Cameroon ma

[@drsaghirghauri5361]

I could not find the video "estimation EGARCH model", kindly respond to my querries???

[@CrunchEconometrix]

Hi Dr. Ghauri, kindly browse my Channel to locate the video on Exponential GARCH. Always allow up to 5 days for responses to your queries. Thanks for your understanding.

[@danielstalmeisters2528]

Hi Ngozi, I love your videos and have found them extremely useful! Currently I am estimating a panel EGARCH model in Eviews, after applying the diagnostic tests above I have noticed that there is serial correlation in my estimation and was wondering whether you know how to correct this? If not then does having serial correlation make my model unreliable? Thank you!

[@CrunchEconometrix]

Hi Daniel, thanks for the positive feedback and kind remarks about my KZbin videos. Deeply appreciated! On a general note, serial correlation and heteroscedasticity are every researchers nightmare. Their occurrence violates OLS assumptions. You may need to refresh those and since I haven't undertaken panel EGARCH may not be in the best position to guide you correctly. My sincere apologies.

[@SureshKumar-qq1yk]

Dear ma'am, Thank a lot. Please tell me about volatility spillover between two variables. I studied a objective is to examine the volatility spillover between stock prices and gold prices. Which model would give appropriate answer to achive that objective.

[@CrunchEconometrix]

Hi Suresh, my suggestion will be that you engage empirical review of similar research for proper guidance. Thanks.

[@SureshKumar-qq1yk]

@@CrunchEconometrix I am from Punjab (India). based on the review I thing Muptiple GARCH model ie used for the spillover volatility among time series variables. Please explain it and share a new video on the MGARCH model in the Eviwes. Kumar et al. (2019) [www.sciencedirect.com/science/article/abs/pii/S0301420718306251?via%3Dihub] for Correlations and volatility spillovers between oil, natural gas, and stock prices in India have been used VARMA-DCC-GARCH model.

[@CrunchEconometrix]

@@SureshKumar-qq1yk I have a video on GARCH-M and not on "multiple GARCH". Have you watched it?

[@AsifAli-gj6us]

CrunchEconometrix hi Mam, would you please upload a video on multi variant GARCH. The nine video I watched is for single variable

[@CrunchEconometrix]

Thanks for the suggestion, Asif. I will work on it. Please share my videos with your colleagues and may God bless you, amen!

[@ahah9211]

Good morning mam can you explain how we can select the optimal number of lags? In case i have daily data should i just use 36 ? It can be different than garch lag ? I mean garch (1,1) does not mean the number of lag should be 1 ? Thanks for your effort

[@CrunchEconometrix]

Hi Ali, all you need to know about ARCH and GARCH are on my Channel. Watch again and adapt to your study. Also, watch my video on Optimal Lag Selection.

[@ahah9211]

Thank you so much 🙏🏻

[@mahadihasan153]

I am doing a volatility analysis with two variables Real effective exchange rate and use the volatility of reer variable to analysis the volatility of foreign direct investment. but i am getting negative garch coefficient while checking for volatility in reer and it has significant arch effect by arch Lm test. should i use egarch ? can you please suggest me?

[@mahadihasan153]

i have also tried egarch model and in this case i am getting negative garch coefficient and positive leverage coefficient. and almost same result in case of fdi whether i use volatility of reer in the mean equation or not. please help me. i could not find any paper that addressed this problem.

[@CrunchEconometrix]

Mahadi, if that is your result. Present it and discuss it within the scope of your study.

[@mahadihasan153]

thank you ☺

[@mahadihasan153]

I've 2 more questions if I may ask. 1. Can I use non stationary series in the mean equation of arch or it is necessary to convert the series to stationary before using it in the mean equation. 2. In the var max lag selection if I select 2 I get 2nd lag significant and if I select I get 3 and so on. I am using a yearly data and want to apply ardl. Should I select 2 in the lag selection and use it as the maximum lag in the ardl endogenous and exogenous variables? I would be very grateful if you answer these questions.

[@CrunchEconometrix]

1) Watch the background video on ARCH/GARCH. I explained what needs to be done. 2) Watch my videos on ARDL. Well explained. Thanks.

[@gregpandise4601]

Is it bad if my GARCH-M with arch-m(variance) model has a negative r-squared?

[@CrunchEconometrix]

Hi Greg, how did you come about a negative R-sq?

[@gregpandise4601]

For my full 10 year sample I get negative r squared. For shorter sample periods, the r squared is positive.

[@CrunchEconometrix]

Did you omit the constant?

[@usmankhurram9739]

After run the ARCH/GARCH fmaily model then in residual diagnostics, If the ARCH LM test passed but the "correlogram Q-statistics" P-Value0.005 but when i used the same distribution for another sub-period of the same stock return series KSE the Q stat P value

[@CrunchEconometrix]

Re-estimate the model at higher order lags.

[@usmankhurram9739]

@@CrunchEconometrix Thanks for quickly respond, you mean the lags of the ARCH/GARCH model like ARCH(2,1) or GARCH (2,2) etc, and please also tell about the sub-period sample analysis as question above. Thanks

[@CrunchEconometrix]

Yes, for the lags. But I don't understand your query about the "sub-sample" analysis. I have videos on those, though. You may want to check them out.

[@usmankhurram9739]

@@CrunchEconometrixNow i run TGARCH model again in Eviews (ARCH 4 and GARCH 1 and threshold order 1) and this time ARCH with 4 lags now result of residual diagnostics, "Q stats p-value" of 36 lags >0.005 except only one lag 22th lag. So is that output Ok now?? if ok then kindly guide me which alpha value of ARCH i need to report of this TARCH model in my paper result (alpha value for ARCH is lag 1st or the lag 4th outcomes of TARCH model) Only tell me about ARCH value, other GARCH and TARCH value i know. Thanks Further the sub-period mean i run ARCH family models on 15 markets returns one by one e.g. chinese index CSI-300 for full sample period 2002-2019 and then sub-sample period from 2002-2009 (Pre-crisis) and then 2nd sub sample 2010-2019 (post-crisis). so when i run GARCH/TARCH model on full sample e.g CSI300 then 1,1,1 TARCH model results significant but when i run 1,1,1 same order on pre-crisis/post crisis period then the residual diag. Q stat pvalue not sign. on 1,1,1 so can i use the 2,1,1 or 3,1,1, etc and also changed the error distribution like Student t or GED untill GARCH/TARCH residual diag Q stat p value on sub-sample period not significant. Thanks

[@CrunchEconometrix]

I will advise you to get published articles that used this technique for detailed guidance. There are lots of literature out there. Thanks.

[@mohamednimcale5507]

dr please can you help me how to estimate garchsk

[@CrunchEconometrix]

Hi Mohd, what is "garchsk"? No idea what that is, honestly.

[@mohamednimcale5507]

@@CrunchEconometrix GARCH with skewness and kurtosis and it was proposed first time by Leon etal (2005) and it is like four parts: 1. Mean equation 2. Variance equation 3. Skewness equation 4. Kurtosis equation

[@CrunchEconometrix]

@@mohamednimcale5507 Thanks. Never heard of this approach.

[@mohamednimcale5507]

@@CrunchEconometrix Never mind, how can I forecast n-step ahead using eviews e.g. 1st step, 2nd step, 3rd step. ........ n step

[@CrunchEconometrix]

@@mohamednimcale5507 No idea.