GARCH Model. Model One. STATA

[@7648-d5b]

hi, many thanks for your detailed explaination, I also get stagger and I know the reason-- after spending such a long time in reading and writing, getting along with journal articles and data bases, everyone got stagger, so dont be sorry, that is a proof of your knowledge and effort.

[@NicolasRodriguez-hc7cj]

Dude, you have just saved my final project for Macroeconomics and Finance class. Keep up the good work! Greetings from Colombia

[@caspersky6473]

Thank you so much for this interesting video... God bless you

[@mijwokopecakokjakamaykway944]

Thank you so much! This video is very much helpful for my thesis. God bless you Sir.

[@henryburdon4012]

Fantastic explanation. But what to do if all 3 models are undesirable like in this case?

[@sayedhossain23]

Henry Burdon Then you have to redesign all your variables and re run model again to get optimum result

[@jounmichael6415]

very good step by step explanation. thank you for the effort. If there is a chance we need more videos on ARIMA and ARMAX and NARX if you could. thank you in advance.

[@sayedhossain23]

You are welcome. In future I shall do those you are asking.

[@sayedhossain23]

You are welcome to Hossain Academy

[@hongthao4117]

Thank you so much. This video is very useful for my thesis.

[@sayedhossain23]

+Hồng Thảo Thank you. I would like to invite you to join Hossain Academy Facebook for greater interaction about economics, finance and econometrics with me. Thank you Sayed Hossain from Hossain Academy. Please join below and post your question.facebook.com/groups/hossainacademy/

[@hongthao4117]

Thank you sir, I did join to your group facebook some days ago.

[@kyithatun2277]

You saved my life.

[@siljeund2429]

Thank you for a great video on a difficult topic Sayed Hossain:) If possible I would like to know more about the mean and variance eqaution. Around minure 32 you say that IBR, PBR and ARCH is the variance eqution. Why do you include only IBR and PBR, but not BSR. Is it because PBR and IBR are exogenous shocks? I also see that you do not include them in what you describe to be mean equation, is there some reason for that? I am asking because I am trying to figure out how to apply the ARCH to my own hypothesis..

[@sayedhossain23]

You are talking about which model? Model 1 or Model 2 of GARCH model as there are two models in Hossain Academy dealing with GARCH model? However, volatility of BBR can be influenced by two shocks. They are external and internal shock. External shock or exogenous factors such as IBR and PBR while ARCH and GARCH are internal shock that influence the volatility of BBR. I guess I have answered you.

[@siljeund2429]

Sayed Hossain Dear Sayed - thank you so much for your notice:) I am talking about model one of Garch. I understand that IBR and PBR are external, because the Indian and Pakistan Bond return will be external to the Bangladesh bond return. However, I do not understand why IBR and PBR are included in the variance equation, but not BSR. Why is that? My main problem is understanding which factors should be included/excluded in the variance equation and why.. Arch and Garch term should be included of course, but I am wondering about the other variables. Thank you so much for your guidance, I will recommend your videos to all my friends:)

[@sayedhossain23]

BSR is Bangladesh stock return....should not be external factor...It is internal factor that influence the volatility of BBR....IBR and PBR are external factor....outside of bangladesh....

[@siljeund2429]

Sayed Hossain Okay, so external factors are variance equation, while mean equation are internal factors?

[@sayedhossain23]

Mean equation means variables are related..that is there should have relationship between independent and dependent variables in reality.

[@12artyom]

Can someone help me with the data ? I downloaded different monthly datas from the yahoo.finance and test for arch effect but never receiveed a significant effect, I supposed that my data was not stationary so took the first difference and then re-run but again I got high value of p, What should I do ?

[@isabellac2927]

Dear Sayed, Im trying to do a GARCH-M and have trouble figuring out how many lags to use. What is the difference between 'specifying maximum lags' and 'specifying a list of lags'. When we used 'specify maximum lags' state said "flat log likelihood encountered, cannot find uphill direction". Could you please try to explain how I can solve this problem? Best regards, Isabella

[@areebakhan9726]

I am also encountering the same problem

[@dancunoguta6893]

Thank you very much, this video was very useful. Be blessed.

[@sayedhossain23]

Thank you Oguta, I would like to invite you to join Hossain Academy Facebook at below link and post your question there if you have any. Thank you once again, Sayed Hossain from Hossain Academy. facebook.com/groups/hossainacademy/

[@LentScope]

Hello, useful video! I have a question: why 10 freedom degrees when stimating by t-student distribution and why 1,5 when you do it by generalized error distribution? Thanks again!

[@hfhjjhc9464]

we saw that, none of the model was desirable . So is there any way to get rid of serial correlation and non normal distrubution?

[@tsegatefera8348]

Great video, thank you!

[@sayedhossain23]

+Tsega Hailu Thank you. I would like to invite you to join Hossain Academy Facebook for greater interaction about economics, finance and econometrics with me. Thank you Sayed Hossain from Hossain Academy. Please join below and post your question.facebook.com/groups/hossainacademy/

[@saurabhsingh9623]

Thank you very much Sir for this video, it cleared my doubts. Only few things : what if all the independent variables are exogenous variables. if the residuals are not normal and have serial correlation then how to select the best model? if my mean equation is - Dollar = c + C1*Dollar(-1) + e , then do i need any other variables for my variance equation? or ARCH and GARCH term are sufficient

[@ishansood7499]

my first value in autocorrelation is 0.35 and the rest are almost zero. how do i interpret that?

[@sayedhossain23]

+ishan sood Thank you. I would like to invite you to join Hossain Academy Facebook for greater interaction about economics, finance and econometrics with me. Thank you Sayed Hossain from Hossain Academy. Please join below and post your question.facebook.com/groups/hossainacademy/

[@matheusramalho9523]

Dear Sayed, First of all, thank you very much for your class. It is helpfull for all students, like me. In the beginnig of the vídeo, you show if a model does not have ARCH effect, it is not possible to use ARCH-GARCH method. And the way to check is testing by statistic Q, in a correlogram. My question is: when you take the residuals, you did not apply square in the residuals. Why? As I see in Enders, Applied Econometrics Time Series, page 130, to test Ljung-Box Q-statistics you must squared residuals and, after that, test if the model has or not seriall correlation, as you can see in this phrase: "Rejecting the null hypothesis that the {𝜀 ̂ 2t } sequence is serially uncorrelated is equivalent to rejecting the null hypothesis of no ARCH or GARCH errors." Once again, thank you very much for your patience and dedication on this video. Your faithfully, Matheus Ramalho

[@sayedhossain23]

Thank you. I would like to invite you to join Hossain Academy Facebook Group at below link and join our group discussion. Thank you. Sayed Hossain from Hossain Academy. facebook.com/groups/hossainacademy/

[@moctarndiaye5830]

Many Thanks! What's about Panel ARCH/Garch?

[@sayedhossain23]

Not done yet. Thank you Sayed Hossain from Hossain Academy at www.sayedhossain.com

[@UKCSnooker]

Great explanation - I've tried doing something similar for my dissertation, but for different countries... however, my results are demonstrating no serial correlation (this is good) but the normality tests all are = 0, regardless of changing the variables multiple times. I know this is undesirable & we can therefore not accept the model. Please can you explain in greater detail how you can change the variables in order to get the desired result in terms of normal distribution? For example: Would smaller sample size work? Or, Fewer or more exogenous variables? Thank you.

[@sayedhossain23]

***** Increase the sample size could be one options to get normal distribution

[@UKCSnooker]

Great - thanks, Why are the exogenous variables (i.e. IBR and PBR) not squared?

[@UKCSnooker]

Also, please can you inform what kind of data is best to use? Is it best to use Daily/ weekly/monthly/annual returns? Thanks

[@sayedhossain23]

monthly

[@sayedhossain23]

***** Dear UKC, Thank you. Please join Hossain Academy Facebook below for greater interaction with me regarding data analysis. Thank you Sayed Hossain from Hossain Academy facebook.com/groups/hossainacademy/

[@chhorntheara7587]

Why u selected garch(1,1)? If so, how?

[@sayedhossain23]

Dear Chhorn, Thank you. I would like to invite you to join Hossain Academy Facebook at below link and post your question there. If I know the answer I shall certainly respond. Thank you once again, Sayed Hossain from Hossain Academy. facebook.com/groups/hossainacademy/

[@amaranwar1249]

Good work.

[@sayedhossain23]

Thank you. Sayed Hossain from Hossain Academy at www.sayedhossain.com

[@sayedhossain23]

Printable copies are not possible indeed. Sorry for that

[@mswoonc]

Stop repeating jesus christ.