This paper presents a general optimization framework to forecast put and call option prices by exploiting the volatility of the options prices. The approach is flexible in that different objective functions for predicting the underlying volatility can be modified and adapted in the proposed framework. The framework is implemented empirically for four major currencies, including Euro. The forecast performance of this framework is compared with the forecast performance of the Multiplicative Error Model (MEM) of implied volatility and the GARCH(1,1). The results indicate that the proposed framework is capable of producing reasonably accurate forecasts for put and call prices.