Hi everyone,
Will there be any impact of presence of structural breaks on conclusions derived from Volatility analysis using GARCH(1,1) Model in time series data.
If yes then how to incorporate this structural break effect on volatility analysis.
Thank you.
Sure, there is an impact.
You can use Markov switching models, or rolling windows. You can also run a test of structural break, you define the Date when the break occured. Then you divide your sample before and after the break, and estimate your volatility model.
Hope that this may help you.
Indeed, in the case of structural break there must be changes in GARCH coefficients.You might detect a break in series using one of the popular structural break tests, for instance the Chow test. Then you can use GARCH separately for parts of the time series indicated by the breakpoint. Another approach would be to use a dummy variable indicating a jump.
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series