Hello, All papers I have read about time-varying copulas just produce tables, but none explains exact steps for empirical data. How do I calculate the time varying parameters for the copula? I have read Pattons work still I dont get it.
Suppose I have two returns R1 and R2 with 500 observations each. I fit some GARCH, get residuals, change it to copula data U1 and U2. I calculate kendals tau for U1 and U2 and get best fit static copula parameters.
What are the next steps for Time varying Copula? Do I split the U1 and U2 into different parts or does this has to do with Kendals Tau. Can someone please give a simplistic explanation for any one specific copula such as Gaussian or Gumbel.
Best Regards
Hi,
I worked on time varying copula, u can search one of my published papers on copula.
you can find reference papers from there too.
copula is quite complex model and its usage and explaination is dependent on the software used.
you have calculated residuals, that was first step, second step is to run selected copula type.
pl consult DR nasrullah, a comsats scholar for copula papers.
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series