Stata Tutorial: Out of Sample Forecasts

Generating Dynamic out-of-sample forecasts of financial data based on ARIMA time-series models in Stata.
Textbook Link
Introductory Econometrics for Finance
amzn.to/2YJc2Tu
My Twitter is:
/ michaelrjonas

[@alexholmes4128]

This was a fantasticaly made video sir, super concise and clear. You have me a lot of headache over my coursework!

[@mikejonaseconometrics1886]

Glad it helped!

[@manalpandey4118]

Dear Mike, I O U Best Wishes, Manal (Gorny says thanks Mike)

[@FVisionK]

Hi! Thank you for these amazing and helpful videos, I was wondering if you could do a video on how to choose the ARIMA specifications as it is something I'm still confused by.

[@mikejonaseconometrics1886]

This will get you part of the way there: kzbin.info/www/bejne/fXvYnH14m66gnZI But I haven't done a public version of the full lecture yet.

[@maggiemullane5390]

Very helpful video! Comment: have you tried using “tsappend” to expand your dataset? It expands the table and fills in the new date observations (for whichever var you set as the date when tsset). Question/request: I’d be curious to see the differences between the predicted values for the expost estimations, and the exante estimations (I noticed in your data browser that the exante values were different than the expost values; would this ever be a problem?). When I did a similar analysis following process, the estimates were the same until reaching the point for starting the dynamic forecast. • Thanks again!!

[@sayedalimsamim551]

Dear professor, than you so much for your nice explanation. after prediction how to get the actual value. such as I predicted cereal production till 2030. in 2022 the value of (rspf_exant= -0.0101488) I need to change the (rspf_exant=-0.0101488) to the number of products such as(? ton). kindly request please respond.

[@louis-davidturcotte3439]

Hey! If your model was using more than one variables, could it be possible to make ex ante forecast, where some of these future variables are already known, so that the forecast can take that into account. Thanks !

[@moneyfornothing3666]

Hi, Thank you for the insightful video. However, I'm facing an error when I use the set obs '=_N+ ' command, stating that observation number out of range. I'm doing ex-ante post forecasting based on your video.

[@robertnicholson7532]

I had the same problem and I think there was an update, So for anyone that has this problem again. You can substitute this with insobs 5

[@gojol00]

can you help my project