TH-cam recently changed the way my content will be monetised. My channel now needs 1,000 subscribers. So it would be amazing if you show your support by both watching my videos and subscribing to my channel if you haven’t done so already. Monetising my videos allows me to invest back into the channel with some new equipment so this small gesture from you will be extremely huge for me. Many thanks for your support….CrunchEconometrix loves to teach, support my Channel with your subscription and sharing my videos with your cohorts.
You are an amazing teacher, I am a student. I have learned more from your videos on ARIMA model than I learned in my entire semester. Once again thank you
Very useful. Your tutorial on VAR models made me get a 76% in my empirical finance assignment. Keep up with the good work. Your interpretations are top notch
They are really informative videos, they helped me a lot when I was studying the empirical methods in Aberdeen University, God bless you and keep you safe.
Thank you so much for the detailed explanation and analysis throughout the ARIMA model forecasting. I very much appreciate your efforts to make this videos available for us and your determination shown in doing analysis. God bless you and stay safe.
Hi Imti, I'm encouraged by your positive feedback and kind words. Thanks for the appreciation. I will continue to create quality videos. May God bless you too and stay safe 🙏
@@CrunchEconometrix one doubt here. How did u managed to change the forecasting RANGE to upto 1991Q4 when the SAMPLE is only upto 1990Q4? I'll be grateful to know how you did that. Thanks alot.
Thank you Dr. I am highly delighted with your videos, I can now run ARIMA model in my stata with ease. We thank you for impacting a lot on us. remain blessed.
Hi Rizal, thanks for the positive feedback and kind remarks about my TH-cam video. Deeply appreciated! Compliments humbly taken too😊. Please may I know from where (location) you are reaching me?
I have a question. Please forgive my ignorance. With ARIMA we are forecasting values that have already passed. Let's say, you forecasted four quarters, but this data were already realised. What is the sense of forecasting something that has already happened? Or is this forecast for the next quarters (1992q1 - 1992q4)? Thank you for the great job. Although I'm a Business PhD student, your videos have been very useful to me.
Hi Kujtim, forecasting is engaged using existing data to project what is to come. I restricted the sample to show just that and how the forecast "fit" the realised data. You may need to check other online resources for more forecasting resources. Thanks.
thank you Dr. your videos are so helpful and simplified and would you please add another video on ARIMA models (forecasting) but done in STATA 16, i would really appreciate it. Thanks once more for such amazing content
Awesome informative video! Helped me a lot! I have a couple of questions to ask. When does the forecast loses its prediction power? I guess ARMA forecasts are more short-term forecasts. Is a 10-year forecast considered valid or should it be less than 5 years? I would also like to ask if GARCH models produce better and more long-run forecasts.
Hi Theo, thanks for the positive feedback. Glad to be of help. A forecast too far into the future is "meaningless" and loses predictive power. Investors won't be able to keep the forecast "in mind". So a 5-year forecast has greater predictability than a 10-year. I'm yet to compare the outcome with a GARCH, hence, cannot say much about it. May I know from where (location) you are reaching me?
@@CrunchEconometrix Thank you for your response! I'm reaching you from Greece. I've already compared the outcome between ARIMA and GARCH models. In my case, both models seem to produce the same forecast. I hope I've done everything right though. Thanks once again!
A very clear explanation to be understood. I wonder why the forecast is only for 4 quarters (1991q1 0 1991 q4), could we do forecast for the longer period? Thank You.
Okki, my opinion is that too long future forecast will lose its meaning. No one will remember (lol). The shorter the forecast, the better. So, 3-5 years forecast, 3 quarters forecast, 3-6 months forecast are more intuitive, if you see my point.
Hi Tuki, I appreciate your positive feedback. Subscribe to my Channel (it's FREE) and watch all ARIMA videos. Please may I know from where (location) you are reaching me?
This is very informative maam. Can you please consider seasonal arima? And also how to forecast future values using arima?i still dont understand the part of getting the next quarters when u all used up all your sample in your data. For example range is from 1998q1 to 2019 q4 and my sample is also 1998q1 to q4.
Thank you very much for such informative videos on arima. I am a Stata user and following the Stata videos, but I can't see the Stata arima f forecasting video
No code yet. The reason why I'm yet to have the video. I'll appreciate if you share my TH-cam link with your colleagues in China. God bless you as you do, amen!
Thanks! i learn alot from your videos but how can i avoid forecasting negative values? my data is not stationery at level but after first order difference it became statinery but when i forecast i got negative values. from your video why did you forecast with gdp not with dgdp or loggdp? because your data is not statinery from the beginning but u made d forecast with gdp.
Hi Musa, thanks for complimenting my videos. Deeply appreciated! Regarding your query: I used the actual GDP and not the differenced or log because I want to simulate if my forecast will be as close to what was realised. Hence, using the GDP values as downloaded from source. Try it out and give me a feedback....may I know where you are reaching me from?
@@CrunchEconometrix Thanks for your reply! am reaching you from Malaysia but am doing my phd research on forecasting real estate residential price in Abuja. if i had watch this your videos while at home i will have try to come and see you personally on my analysis. i will appreciate if i can get your email adress.
@@musaalkali5406 My email address cruncheconometrix@gmail.com. Please know that due to time constraints, and as a matter of principle, I don't check results sent to me. Hence, I only offer guidance in accordance with my video tutorials. Thanks for the understanding.
Hello ma'm, I have gone through the whole process and understood it except the last thing - plotting the graph. As we have data set of 1970-1990, how could we match the actual gdp of 1991 with forcasted gdp of 1991. Cause we don’t have the actual gdp 1991.
Thank u for the video. But I want to ask. Your bias proportion is 0.5 is it too high? Which I read it must under 0.2. And how you explained it ? Thank you in advance. 😊
@@CrunchEconometrix okay thank you. Anyway I am done using AR(1) AR(2) MA (12) how to write the model? Since arima model (p,d,q), and the it is significant at first difference level. Thank you maam.
Thank you so much for making this extremely useful series. However i have encountered a problem when using Auro Arima function for my time series. It returned a ARIMA(0,1,0) model and that make me so confused. How can I interpret it. My time series is a series of daily stock prices with 500 observes. Thanks again!
Hi Phurong, please it is VERY IMPORTANT that you watch the introductory video to the ARMA/ARIMA series. You cannot estimate any model with zero (0) AR and MA features. What you have is just a series that is stationary at 1st difference. At best, estimate an ARCH model.
Hi Zhale, thanks for the positive feedback and kind remarks about my TH-cam videos. Deeply appreciated! The code you requested is not available. The reason I don't have the video, yet. Please may I know from where (location) you are reaching me?
I have tried to run ARIMA modeling in more than 10-time series data(GDP, Stock Price, Index Price) for practice but in every model, my correlogram of the residual square is significant (less than 5 percent). what should I do? I am trying to find the solution for the same but I failed to find an answer. Even after doing the adjusted ARIMA model the results are the same as mentioned above
Thank u so much I took the data for 60 years n for the very first time correlogram squared residuals were insignificant. Really happy to see the results 😊😊 Thanks for guiding me .
Firstly Jo, you are using the wrong comment section. You should have posted this comment on the video section on ARDL and ECM models that way anybody reading can follow and contribute to the discussion. Secondly, re-post this on any ARDL video and I will guide you appropriately. Thanks!
Hello,Thank you for the vidéo. I wonder if you could prepare à video about how To brun Impulse response fonction with Jorda's local projection method on Stata . Thank you very much
douda dada Hi never heard of that method b4 but may have to read up on it and if I understand the procedure, I'll do a video on it. Thanks for watching my videos...share the link with your social media community and academic networks too😉
Dear mam, when I do forecasting in Eviews, my system is getting hanged and is not responding. Is there any way to do forecasting other than the model which was mentioned in this video and automatic ARIMA forecasting? Please do help me mam,.
Hello! Could you please advice me on this - I am trying to examine the determinants of economic diversification by using herfindahl index as the dependent variable and the independent variables are gdp, inflation, exchange rate, trade openness, domestic credit to private sector and gross fixed capital formation.. the problem is that while doing the unit root test I found that some of my variables are stationary at first difference, some at level and gdp at second difference.. could you please suggest me which econometric model should I apply in this case for my study?? Will be really thankful for your help!!!
@@saimashadab4799 Increase time span to 30 years. Either drop the I(2) series and use ARDL model or include it and use the Toda-Yamamoto model. Find out more about this procedure online because I've not used it before. Thanks 😊.
@@CrunchEconometrix thank you so much for your reply! I can't expand the time series data to 30 years because it is not available anywhere! That is why I have to take 1995-2017. I am researching on economic diversification in the UNITED ARAB EMIRATES and data is not available for several variables prior to 1995..
@@CrunchEconometrix in this case should I simply apply FMOLS regression to test for the determinants of economic diversification in UAE instead of dropping I(2) variable and applying ARDL?? Which one do you think shall be better considering the single country case and time period 1995-2017? Thank you very much in advance and sorry if I am asking dumb questions, actually my econometrics is not good at all!
@@saimashadab4799 Ok, if that is the case, I will suggest that put up this query in other econometric platforms for more expert advice. May I know from where (location) you are reaching me? Thanks.
@@CrunchEconometrix sorry crunshEconometrix not for you that for auther vids ho explain ARIMA spinking french i said for him that fench becam old language sorry anthor time I am a follower of all your TH-cams
TH-cam recently changed the way my content will be monetised. My channel now needs 1,000 subscribers. So it would be amazing if you show your support by both watching my videos and subscribing to my channel if you haven’t done so already. Monetising my videos allows me to invest back into the channel with some new equipment so this small gesture from you will be extremely huge for me. Many thanks for your support….CrunchEconometrix loves to teach, support my Channel with your subscription and sharing my videos with your cohorts.
You should add a couple ads in the beginning of the video, so you can earn more money. Thank you very much for the good content
Thanks for the encouraging feedback and good thoughts, Lexxy! :)
You are an amazing teacher, I am a student. I have learned more from your videos on ARIMA model than I learned in my entire semester. Once again thank you
Happy to hear that, Hisam...thanks!
100000000000000000000000000 Thanks!!!! You´re the best!
Thanks, Ronaldo for your encouraging words. Deeply appreciated 🤗.
Very useful. Your tutorial on VAR models made me get a 76% in my empirical finance assignment. Keep up with the good work. Your interpretations are top notch
Great to hear, Milton...thanks!
Excellent. Well explained and easy to understand. Keep up the good work Doc.
Thanks for the encouraging words, deeply appreciated!
They are really informative videos, they helped me a lot when I was studying the empirical methods in Aberdeen University, God bless you and keep you safe.
Thanks, Ahmed for the encouraging feedback. Deeply appreciated!
Thank you so much for the detailed explanation and analysis throughout the ARIMA model forecasting. I very much appreciate your efforts to make this videos available for us and your determination shown in doing analysis. God bless you and stay safe.
Hi Imti, I'm encouraged by your positive feedback and kind words. Thanks for the appreciation. I will continue to create quality videos. May God bless you too and stay safe 🙏
@@CrunchEconometrix one doubt here. How did u managed to change the forecasting RANGE to upto 1991Q4 when the SAMPLE is only upto 1990Q4? I'll be grateful to know how you did that. Thanks alot.
Actual sample size ends at 1991Q4. I restricted the size to 1990Q4 to enable me forecast the next 4 quarters.
@@CrunchEconometrix thanks a lot for your prompt response. Really appreciate it.
U're welcome, Imti😊
Thank you for guiding me through your video on forecasting. Much appreciated work. Kudos to you.
U're welcome, Yogesh 😊. Please may I know from where (location) you are reaching me?
I am from India and found your videos very useful to work on softwares like stata and eviews.
Thank you Dr. I am highly delighted with your videos, I can now run ARIMA model in my stata with ease. We thank you for impacting a lot on us. remain blessed.
Hahahaha Abdulazeez, good to hear this encouraging feedback!!! May I know from where (location) you are reaching me?
Thank so much .I learn a lot from your videos.g ood luck
U're welcome, Manori...kindly share my videos with your students and colleagues too! 💕 😊
This tutorials are excellent! However, I would like to request for tutorial on Nowcasting GDP and inflation using ARIMA. Thank you!
Thanks Eric, for the encouraging feedback. Deeply appreciated! Please may I know from where (location) you are reaching me?
great tutorial.
Thanks Syamraj, for the positive feedback. Deeply appreciated! Please may I know from where (location) you are reaching me?
you're the best econometrician ever...
Thanks Alpha...compliment is humbly taken😀. May I know from where (location) you are reaching me?
Thank you so much!! I love your video. it is really help me for finish my thesis. thank you thank you.
U're welcome, Weka. Thanks for the encouraging feedback. Deeply appreciated! Please may I know from where (location) you are reaching me?
@@CrunchEconometrix I am from Indonesia
Great work! What is your advice for ARIMA on predicting the effect of terrorism attacks on tourist arrivals?
Hi Charles, if you have high frequency data to model the scenario why not? Try it to see the outcome.
thanks so much, youre the best
Hi Rizal, thanks for the positive feedback and kind remarks about my TH-cam video. Deeply appreciated! Compliments humbly taken too😊. Please may I know from where (location) you are reaching me?
CrunchEconometrix I’m from indonesia 🇮🇩 your video help me to do my exam, big thanks for u
Bravo!!! Gr8!!! Proud of you!!!
I have a question. Please forgive my ignorance. With ARIMA we are forecasting values that have already passed. Let's say, you forecasted four quarters, but this data were already realised. What is the sense of forecasting something that has already happened? Or is this forecast for the next quarters (1992q1 - 1992q4)? Thank you for the great job. Although I'm a Business PhD student, your videos have been very useful to me.
Hi Kujtim, forecasting is engaged using existing data to project what is to come. I restricted the sample to show just that and how the forecast "fit" the realised data. You may need to check other online resources for more forecasting resources. Thanks.
thank you Dr. your videos are so helpful and simplified and would you please add another video on ARIMA models (forecasting) but done in STATA 16, i would really appreciate it. Thanks once more for such amazing content
Noted, Jones. I will do my best!
Awesome informative video! Helped me a lot! I have a couple of questions to ask. When does the forecast loses its prediction power? I guess ARMA forecasts are more short-term forecasts. Is a 10-year forecast considered valid or should it be less than 5 years? I would also like to ask if GARCH models produce better and more long-run forecasts.
Hi Theo, thanks for the positive feedback. Glad to be of help. A forecast too far into the future is "meaningless" and loses predictive power. Investors won't be able to keep the forecast "in mind". So a 5-year forecast has greater predictability than a 10-year. I'm yet to compare the outcome with a GARCH, hence, cannot say much about it. May I know from where (location) you are reaching me?
@@CrunchEconometrix Thank you for your response! I'm reaching you from Greece. I've already compared the outcome between ARIMA and GARCH models. In my case, both models seem to produce the same forecast. I hope I've done everything right though. Thanks once again!
@@theo6699 I'm applauding your efforts. Well done! Right or wrong? I'll say we learn both ways.
A very clear explanation to be understood.
I wonder why the forecast is only for 4 quarters (1991q1 0 1991 q4), could we do forecast for the longer period?
Thank You.
Okki, my opinion is that too long future forecast will lose its meaning. No one will remember (lol). The shorter the forecast, the better. So, 3-5 years forecast, 3 quarters forecast, 3-6 months forecast are more intuitive, if you see my point.
@@CrunchEconometrix Thank you for the explanation
please also make video tutorial for arima forecasting in STATA
I'll do my best, Ralph. At the moment, no resource yet on the forecasting part. Thanks for the suggestion. Gracias!
thank you this video, we would like to study arima models
Hi Tuki, I appreciate your positive feedback. Subscribe to my Channel (it's FREE) and watch all ARIMA videos. Please may I know from where (location) you are reaching me?
This is very informative maam. Can you please consider seasonal arima? And also how to forecast future values using arima?i still dont understand the part of getting the next quarters when u all used up all your sample in your data. For example range is from 1998q1 to 2019 q4 and my sample is also 1998q1 to q4.
Hi Rachel, thanks for the encouraging feedback. Deeply appreciated! 🙏 I'll consider seasonal ARIMA. It's written down on my to-do-list😊
awesome video!!!
will you make a video with SARIMA approach??
Thanks for the video
Thanks Gus, I will definitely do videos on SARIMA once I clear my backlog of videos. I value your support...keeps me going :)
CrunchEconometrix Much love.Thanks.Your videos are just amazing
@@michaelmwanza355 Thanks for the compliments...very humbling Michael💕💃
Please make a video on two variable by ARIMA Modell in EViews(from Pakistan) thank you
Alright Muhd, I have noted your suggestion. Thanks!
Thank you very much for such informative videos on arima. I am a Stata user and following the Stata videos, but I can't see the Stata arima f forecasting video
Thanks, Veks for the kind remarks. The 4th step is still in the works. Not ready yet. Please may I know from where (location) you are reaching me?
@@CrunchEconometrix I'm recently in China
@@CrunchEconometrix can you please tell me the code for the forecasting? Cos I also need to generate a time trend from my arima model
No code yet. The reason why I'm yet to have the video. I'll appreciate if you share my TH-cam link with your colleagues in China. God bless you as you do, amen!
Thanks! i learn alot from your videos but how can i avoid forecasting negative values? my data is not stationery at level but after first order difference it became statinery but when i forecast i got negative values. from your video why did you forecast with gdp not with dgdp or loggdp? because your data is not statinery from the beginning but u made d forecast with gdp.
Hi Musa, thanks for complimenting my videos. Deeply appreciated! Regarding your query: I used the actual GDP and not the differenced or log because I want to simulate if my forecast will be as close to what was realised. Hence, using the GDP values as downloaded from source. Try it out and give me a feedback....may I know where you are reaching me from?
@@CrunchEconometrix Thanks for your reply! am reaching you from Malaysia but am doing my phd research on forecasting real estate residential price in Abuja. if i had watch this your videos while at home i will have try to come and see you personally on my analysis. i will appreciate if i can get your email adress.
@@musaalkali5406 My email address cruncheconometrix@gmail.com. Please know that due to time constraints, and as a matter of principle, I don't check results sent to me. Hence, I only offer guidance in accordance with my video tutorials. Thanks for the understanding.
Hello ma'm, I have gone through the whole process and understood it except the last thing - plotting the graph. As we have data set of 1970-1990, how could we match the actual gdp of 1991 with forcasted gdp of 1991. Cause we don’t have the actual gdp 1991.
Exactly, Munem. You can't.
So how could you do that in graph?
Please explain it. I am stucked at this last particular part where you resize the range and plot the graph.
Please I would like to know how to make future forecast of an ARIMA MODEL.
I covered forecast in this video.
Thank u for the video. But I want to ask. Your bias proportion is 0.5 is it too high? Which I read it must under 0.2. And how you explained it ? Thank you in advance. 😊
Hi Kotik, you can always make the necessary adjustments. Thanks.
@@CrunchEconometrix okay thank you. Anyway I am done using AR(1) AR(2) MA (12) how to write the model? Since arima model (p,d,q), and the it is significant at first difference level. Thank you maam.
-in forecasting why should we click on GDP rather than D(GDP). since ARIMA(1,1,8,12) model is based on d(GDP). please comment.
Hi Kanchan, I used the differenced GDP. Kindly watch the clip again, thanks.
Thank you so much for making this extremely useful series. However i have encountered a problem when using Auro Arima function for my time series. It returned a ARIMA(0,1,0) model and that make me so confused. How can I interpret it. My time series is a series of daily stock prices with 500 observes. Thanks again!
Hi Phurong, please it is VERY IMPORTANT that you watch the introductory video to the ARMA/ARIMA series. You cannot estimate any model with zero (0) AR and MA features. What you have is just a series that is stationary at 1st difference. At best, estimate an ARCH model.
@@CrunchEconometrix Thanks again! I'm from Vietnam.
Please, do this class on Matlab.
Hi Edilson, thanks for the suggestion but I have no idea about Matlab.
Thanks for your useful videos🙏🙏 May I have the forecast codes in Stata?
Hi Zhale, thanks for the positive feedback and kind remarks about my TH-cam videos. Deeply appreciated! The code you requested is not available. The reason I don't have the video, yet. Please may I know from where (location) you are reaching me?
CrunchEconometrix yes I’m from Iran
I have tried to run ARIMA modeling in more than 10-time series data(GDP, Stock Price, Index Price) for practice but in every model, my correlogram of the residual square is significant (less than 5 percent). what should I do? I am trying to find the solution for the same but I failed to find an answer.
Even after doing the adjusted ARIMA model the results are the same as mentioned above
That may be due to the very short time dimension of 10 years. Have a longer time series of at least 30 years and re-estimate the model.
@@CrunchEconometrix thank u... I will try the same
Thank u so much
I took the data for 60 years n for the very first time correlogram squared residuals were insignificant.
Really happy to see the results 😊😊
Thanks for guiding me
.
@@ramansingh7560 U're welcome...kindly share my my videos with your friends and colleagues too 😊
Sure ..
Hi, is there a possible way of combining both ARDL and ECM into a single equation and output? thank in advance.
Yes Johnny, it's very possible. Please watch my videos on ARDL and ECM estimations and you'll know how to go about them.
Thank you very for a reply, i have watching it but fail to capture where to combine them and which output should I interpret.
Firstly Jo, you are using the wrong comment section. You should have posted this comment on the video section on ARDL and ECM models that way anybody reading can follow and contribute to the discussion. Secondly, re-post this on any ARDL video and I will guide you appropriately. Thanks!
Hello,Thank you for the vidéo.
I wonder if you could prepare à video about how To brun Impulse response fonction with Jorda's local projection method on Stata .
Thank you very much
douda dada Hi never heard of that method b4 but may have to read up on it and if I understand the procedure, I'll do a video on it. Thanks for watching my videos...share the link with your social media community and academic networks too😉
CrunchEconometrix Thank you very much.
Dear mam,
when I do forecasting in Eviews, my system is getting hanged and is not responding. Is there any way to do forecasting other than the model which was mentioned in this video and automatic ARIMA forecasting?
Please do help me mam,.
I know that automatic ARIMA forecasting is possible but I don't know the procedure yet. Unless I find the time to learn and then do videos on it.
Hello! Could you please advice me on this - I am trying to examine the determinants of economic diversification by using herfindahl index as the dependent variable and the independent variables are gdp, inflation, exchange rate, trade openness, domestic credit to private sector and gross fixed capital formation.. the problem is that while doing the unit root test I found that some of my variables are stationary at first difference, some at level and gdp at second difference.. could you please suggest me which econometric model should I apply in this case for my study?? Will be really thankful for your help!!!
BTW, just for more details - all data is time series data taken from wdi for the period 1995-2017
@@saimashadab4799 Increase time span to 30 years. Either drop the I(2) series and use ARDL model or include it and use the Toda-Yamamoto model. Find out more about this procedure online because I've not used it before. Thanks 😊.
@@CrunchEconometrix thank you so much for your reply! I can't expand the time series data to 30 years because it is not available anywhere! That is why I have to take 1995-2017. I am researching on economic diversification in the UNITED ARAB EMIRATES and data is not available for several variables prior to 1995..
@@CrunchEconometrix in this case should I simply apply FMOLS regression to test for the determinants of economic diversification in UAE instead of dropping I(2) variable and applying ARDL?? Which one do you think shall be better considering the single country case and time period 1995-2017? Thank you very much in advance and sorry if I am asking dumb questions, actually my econometrics is not good at all!
@@saimashadab4799 Ok, if that is the case, I will suggest that put up this query in other econometric platforms for more expert advice. May I know from where (location) you are reaching me? Thanks.
I can't find your video on arima forecasting in stata.. please help
Hi Wilfred, it's still a W.I.P. May I know from where (location) you are reaching me?
How do we forecast for year 1992 and 1993?
If you have the data, yes. Simply follow the guide shown in the video. Please may I know from where (location) you are reaching me?
Mam , do you have any videos on ARCH and GARCH models?
Thank-you in advance
Not yet, Jay...working on it.
@@CrunchEconometrix How soon will they be ready? I have final exams coming soon? thank-you in advance for your time
can i forecast out of sample for more variables in arima??
How do you mean?
want to forecast out of sample in a single equation...
@@md.mahfujurrahman864I suggest that you seek other online resources...I have no idea what you intend doing. Thanks.
no french please
Hi Mustapha, but I did not speak French.
@@CrunchEconometrix sorry crunshEconometrix not for you that for auther vids ho explain ARIMA spinking french i said for him that fench becam old language sorry anthor time I am a follower of all your TH-cams
Alright, Mustapha. I appreciate your support. May God bless you, amen 🙏