(a) Distinguish between the terms ‘conditional variance’ and

‘unconditional variance’. Which of the two is more likely to be relevant for producing:

i. one-step-ahead volatility forecasts ii. twenty-step-ahead volatility forecasts.

(b) If ut follows a GARCH(1,1) process, what would be the likely result if a regression of the form in Question 1

(e) were estimated using OLS and assuming a constant conditional variance?

(c) Compare and contrast the following models for volatility, noting their strengths and weaknesses:

i. Historical volatility ii. EWMA iii. GARCH(1,1)
iv. Implied volatility.