Volatility is a crucial factor in option pricing models. The implied volatility is the estimated volatility from market trades and represents market sentiment in recent research. The implied volatility surface (IVS) based market analysis is also gathering more attention for its satisfying regression performance in risk management and related tasks. In this research, we develop an industrial application of the market trading system based on Support Vector Regression (SVR) and implement simulated trading through backtesting. According to the SPY option historical data experiment results, considering appropriate transaction cost, the system reaches satisfying performance in the fluctuating market.