The attached code estimates the MGARCH-M through the recursive (expanding) windows. We set one-lag VAR for the mean model. We fix the initial date and keep increasing the sample size in each recursive window until the end of the sample. We set the increment to 5 observations.
It saves several parameter estimates, test statistics, and information criteria in separate objects computed through the windows.
Also, it tests for causality (spillover) in the variance in each recursive window separately.
Interested readers may refer to the following papers, which test the volatility spillover through the recursive windows.
Tanin, T. I., Hasanov, A. S., Shaiban, M. S. M. & Brooks, R. (2022). Risk transmission from the oil market to Islamic and conventional banks in oil-exporting and oil-importing countries. Energy Economics. 115, 17 p., 106389.
https://doi.org/10.1016/j.eneco.2022.106389
Vellachami, S., Hasanov, A.S., and Brooks, R. (2023). Risk transmission from the energy markets to the carbon market: Evidence from recursive window approach. International Review of Financial Analysis, 89, 102715.
https://doi.org/10.1016/j.irfa.2023.102715
Recursive window-based MGARCH-M
Recursive window-based MGARCH-M
- Attachments
-
- recur_VAR1_MGARCH-M-BGD.RPF
- RATS code
- (4.38 KiB) Downloaded 894 times
-
- Data_BGD.txt
- Data file
- (143.16 KiB) Downloaded 890 times