Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • VAR-MGARCH Spillover effects help. # DCC-MGARCH #CCC-MGARCH # BEKK-MGARCH

    Dear Statalists:

    I am fairly a new person to Stata software and it's my first time here to post some questions here.

    I am trying to run a VAR-MGARCH model with BEKK and DCC and CCC specifications. My dataset contains 4 financial indices(each of them follows I(1) process ) and I want to investigate the potential return and volatility spillovers between the underlying variables. However, it seems that both DCC and CCC models do only provide me with the main diagonal matrices and do not provide further specifications of the full ARCH and GARCH parameters. It also seems that Stata 16 does not offer BEKK regressions.

    My question here: can anyone here able to share me some command or ado files to guide me on how to operate those VAR-DCC/ VAR-CCC models with full specifications of ARCH/ GARCH specifications to test for spillover effects between the underlying variables.

    For, the codes I used are listed below:

    Code:
    tsset t
    
    varsoc r_cnne r_eco r_wti r_cqqq, maxlags(10) # use information criterion to find the opitmal lag choice for VAR operation.
    
    mgarch dcc (r_cnne r_eco r_wti r_cqqq = L.r_cnne L.r_eco L.r_wti L_r.cqqq), arch (1) garch(1) distribution(normal)  nolog  # VAR-DCC with lag order 1 for mean euqation and GARCH(1,1) process for DCC residual.
    
    mgarch ccc (r_cnne r_eco r_wti r_cqqq = L.r_cnne L.r_eco L.r_wti L_r.cqqq), arch (1) garch(1) distribution(normal)  nolog  # VAR-CCC with lag order 1 for mean euqation and GARCH(1,1) process for CCC residual.
    Thank you in advance and You kindly advise and help is highly appreciated.

    Best regards
    Ben
    Last edited by Binyi Zhang; 16 Aug 2019, 02:20.

  • #2
    Cross-posted at https://stackoverflow.com/questions/...rch-ccc-mgarch

    Please note our policy on cross-posting, explicit in the FAQ Advice: you are asked to tell us about it.

    Comment

    Working...
    X