This paper investigates whether the combination of technical analysis techniques: box theory and moving average can predict the stock price movement and well define buying and selling signals. This paper also proposes a stock trading model based on these techniques. The experiment on historical data of S&P 500 components and some indexes (S&P 500, Hang Sheng, KOSPI, TWSE, NASDAQ composite, NASDAQ100, DJIA, NIKKEI225, CAC40 and FTSE100) from 2005/1 to 2015/1 show a promising performance with outstanding potential profit that outperforms buy-and-hold strategy, even in the global financial crisis 2008. The result which tested by statistic methods shows that the proposed model is a significant tool of investing.