TLDR: How many variables can I have in a VAR or VECM model?
I am writing my thesis and I am using a VECM (VAR model with error correction for cointegration) model for analyzing the relationship between the prices of an energy exchange and some other factors. So far I have 4 variables in my model and I am thinking of adding more.
My question is that after how many variables does the model become unusable and unstable or can I add as much as I like?
Thank you for your answers in advance!
One approach here is to add only those variables that improve modelling performance. In order to measure modelling performance, you can use metrics based on the out-of-sample rolling-origin approach. Please let me recommend this publication:
Article Assessing Point Forecast Bias Across Multiple Time Series: M...
You can use the OPc and AvgRelMAE metrics (see the FEW-L1 modelling performance evaluation workflow) to evaluate performance and choose the combination of variables that gives the best performance.
See if the two attached searches are helpful. Best wishes, David Booth
See these links:
https://www.researchgate.net/topic/VECM
http://web.vu.lt/mif/a.buteikis/wp-content/uploads/2018/04/Lecture_07.pdf
https://towardsdatascience.com/vector-autoregressions-vector-error-correction-multivariate-model-a69daf6ab618
https://www.reed.edu/economics/parker/s14/312/tschapters/S13_Ch_5.pdf
If you are not constrained by degrees of freedom, then you should be guided by the theory. Include all the relevant variables suggested by theory, that should constitute the minimum set of variables that you should include. Should you add more variables after that? That depends. What is the justification for those additional variables? You should be able to provide a justification for those added variables. You should also avoid a "kitchen sink" approach.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression