You can find the spreadsheets for this video and some additional materials here: drive.google.com/drive/folders/1sP40IW0p0w5IETCgo464uhDFfdyR6rh7 Please consider supporting NEDL on Patreon: www.patreon.com/NEDLeducation
I was reading my time series analysis book and had some questions on MA models. Then I wake up today and NEDL's got this video up. Thank you for your great timing 😎
I was searching for the excel version of MA implementation but couldn't find it anywhere on the internet. Thanks man, this is special, and I feel at ease knowing now how MA is solved mathematically.
Hello, first of all I'd like to thank you for sharing all these great tutorials and your knowledge. It helps me a lot, especially these time series analysis videos. It would be absolutely awesome if you created some subsequent tutorials, starting with easy AR-Models, then introducing MA-Models and a comparison between both, their assumptions and necessary conditions for applying them to make forecasts, than combining them to ARMA-Models. By the fact that you already have created vids on ARCH and GARCH, you could perfectly combine them to a "Time Series Analysis Playlist"
Thanks for the amazing content. a question concerning the error term : it has to be with mean of zero and of constant variance. In your illustration how could we know error term would fulfil above requirements?
Hi, Sava.Great content as always! I'm not sure how to formulate this question... is there a way, or could this model be applied on pairs trading using cointegration? What would be used as returns? Returns of both pairs? Cointegration itself maybe? Thanks!
Hi, and glad you liked the video! It is true that sometimes you want to "filter out" short-term dependence from your stationarity test, but you generally use augmented Dickey-Fuller test for it, regressing the change in the dynamic equilibrium (Price B - a - b*Price A) onto lagged values of the dynamic equilibrium as well as lagged changes of set order. So augmented Dickey-Fuller is an AR specification by design. I do not see fundamentally why augmented Dickey-Fuller cannot be reformulated as a MA model instead, it is just unconventional and I have not seen it applied in the literature or practice. Hope this makes sense!
@@NEDLeducation Thank you so much. Thats so great. I really hope, that you will go on with your channel and give us some more vids to analyze Time series data e.g. with extensions like ARMA(X) and ARIMA(X). For so long I try to understand time series analysis and now i really think, thanks to your channel I will manage hit. It would be awesome If you would create a vid on basics of time series analysis - Topics like how to choose between different models (how to choose between AR, MA, ARMA, ARIMA) and how to check wheater a time series fullfills stationarity assumptions ort not Please Please go on with this channel!
Thanks for the very helpful video bro. But I have a question why did you shift the lag of the error terms downward. In my opinion, it should be shifted upward. for example, e(-1) sholud be (-0.04%,0.24%,-0.31%....), and e(-2) should be (0.24%,-0.31%,0.17% ....). Thanks
You can find the spreadsheets for this video and some additional materials here: drive.google.com/drive/folders/1sP40IW0p0w5IETCgo464uhDFfdyR6rh7
Please consider supporting NEDL on Patreon: www.patreon.com/NEDLeducation
I was reading my time series analysis book and had some questions on MA models. Then I wake up today and NEDL's got this video up. Thank you for your great timing 😎
I was searching for the excel version of MA implementation but couldn't find it anywhere on the internet. Thanks man, this is special, and I feel at ease knowing now how MA is solved mathematically.
Best finance channel ever 👍
Great content! Thank you, Savva!
Hey man great!!!!!! This is one of the building blocks of time series econometrics...very nice
Hello, first of all I'd like to thank you for sharing all these great tutorials and your knowledge. It helps me a lot, especially these time series analysis videos. It would be absolutely awesome if you created some subsequent tutorials, starting with easy AR-Models, then introducing MA-Models and a comparison between both, their assumptions and necessary conditions for applying them to make forecasts, than combining them to ARMA-Models. By the fact that you already have created vids on ARCH and GARCH, you could perfectly combine them to a "Time Series Analysis Playlist"
Amazing!
Thanks for the amazing content. a question concerning the error term : it has to be with mean of zero and of constant variance. In your illustration how could we know error term would fulfil above requirements?
Hi, Sava.Great content as always! I'm not sure how to formulate this question... is there a way, or could this model be applied on pairs trading using cointegration? What would be used as returns? Returns of both pairs? Cointegration itself maybe? Thanks!
Hi, and glad you liked the video! It is true that sometimes you want to "filter out" short-term dependence from your stationarity test, but you generally use augmented Dickey-Fuller test for it, regressing the change in the dynamic equilibrium (Price B - a - b*Price A) onto lagged values of the dynamic equilibrium as well as lagged changes of set order. So augmented Dickey-Fuller is an AR specification by design. I do not see fundamentally why augmented Dickey-Fuller cannot be reformulated as a MA model instead, it is just unconventional and I have not seen it applied in the literature or practice. Hope this makes sense!
@@NEDLeducation Thanks! 👍👍
@@NEDLeducation is this spreadsheet going to be available on google drive?
Please give us all a tutorial on AR Modells and then a combination with this MA-Approach to ARMA-Modells
Hi, and thanks for the suggestion! Here is a new video on the AR model, if you are interested: th-cam.com/video/ANFOzqNCK-0/w-d-xo.html
@@NEDLeducation Thank you so much. Thats so great. I really hope, that you will go on with your channel and give us some more vids to analyze Time series data e.g. with extensions like ARMA(X) and ARIMA(X). For so long I try to understand time series analysis and now i really think, thanks to your channel I will manage hit.
It would be awesome If you would create a vid on basics of time series analysis - Topics like how to choose between different models (how to choose between AR, MA, ARMA, ARIMA) and how to check wheater a time series fullfills stationarity assumptions ort not
Please Please go on with this channel!
Thanks for the very helpful video bro. But I have a question why did you shift the lag of the error terms downward. In my opinion, it should be shifted upward. for example, e(-1) sholud be (-0.04%,0.24%,-0.31%....), and e(-2) should be (0.24%,-0.31%,0.17% ....). Thanks
Sorry I have made a mistake, your data is ordered from the farthest day to the recent day, not the way around. Sorry
Will you also be doing AR and ARMA models?
Hi Joe, and thanks for the question! Yes, they are in my plans for the future :)