I am working on a model for stochastic volatility. In short, the model try to capture that the volatility goes up suddenly after a shock (war, policy, financial events, etc) and then goes down slowly, so-called a shot-noise form or a impulse response function form. Now, I want to find some support in empirical data. I have two questions:
- According to your experience, which period data correspond this shot-noise phenomena?
- What kind of test (auto-correlation, etc) can mathematically show the existence of this feature?