Purpose - To use efficient volatility direction forecasts for option trading.
Design/methodology/approach - Presents an option trading methodology as a flow chart. Bases it on weekly closing values , and calculates historical volatility series using both a naïve forecast and a 13 week moving average forecast. Combines in a two-layer artificial neural network (ANN)with back-propagation. Adds an Imply Volatility Series, and forecasts one period ahead. Applies to the CAC 40, DAX and Greek FTSE/ASE 20, rolling 26 one-week forecasts.

Findings - Finds the combined method provided much more accurate forecasts and a profit over 26 weeks. However, notes some simpler ,methods yielded higher profits.

Research limitations/implications - Proposes research into more accurate directional predictions, limited to those over specific margins. Adds the need to use daily data.

Originality/value - Presents an apparently simple method of forecasting the direction of volatility.