Course Name: Options Volatility Trading: Concepts and Strategies, Section No: 16, Unit No: 19, Unit type: Exercise
Hello,
I have an issue with the GARCH course, why would you model monthly volatility whereas the squared monthly returns are not autocorrelated comapred to daily sacred returns whihc are autocorrelated ?
In addition, you don’t include a constant in the GARCH model when including the Parkinson estimator which can lead to a variance of 0 when all other terms are 0. Also, your starting parameters values are all 0.1 including for the long persistence parameter beta which is typically close to 0.9 in the literature, which eventually lead to an estimated parameter value of 0 which do not make sense.
Also in the likelihood function you don’t include the past values of volaility for beta but instead you use the Parkinson estimator whihc again do not make sense.