Volatility modelling and trading using option strategies

Thesis covers theoretical topics of realized and implied volatility, exchange traded products associated with volatility trading and financial instruments that can be used to trade volatility such as options as well as option strategies. Models ARIMA, ARFIMA, HAR-RV, semi-variance HAR-RV, GARCH, EGARCH and GJR-GARCH were selected for the volatility modelling study. These models are firstly explained to the reader and then parameters are estimated using ordinary least squares and maximum likelihood method on an underlying asset UVXY. The viability of the model is evaluated using the lowest information criterion and most viable models are then used for volatility forecast. Models with lowest AIC and BIC were ARIMA(1,0,1), ARFIMA(2,1,1), ARFIMA(1,d,1), semi-variance HAR and EGARCH(1,1). Strategies straddle, strangle, iron butterfly and iron condor are then tested based on volatility forecast of the models. Entry into the trade is based on estimated volatility and we test two holding periods weekly and monthly according to the option expiration calendar. The performance of each trading system was evaluated by standard performance measures such as net profit, maximum drawdown, profit factor, Sharpe and Sortino ratio. For the monthly holding period, the best strategy was strangle and all the models achieved the same results. In the weekly holding period models ARFIMA(1,d,1), semi-variance HAR and EGARCH(1,1) achieved best results when using strategy strangle.