Design of time-variable stop losses and profit objectives using neural networks

摘要

Stop losses and profit objectives are important components in many trading strategies, and are used to limit losses and secure achieved profits. An investor, who buys an asset or a security at a particular price, decides that if the price drops beyond a certain level, called a stop loss, then the investor sells the security in an attempt to avoid further losses. On the other hand, if the price appreciates beyond a certain level (profit objective), the investor sells the security to secure the achieved profit, for fear that the security has become overpriced and might fall in price. In this study we propose time variable stop losses and profit objectives. Their values are not constant, but vary in the course of trade. We design a method to obtain optimal stop loss and profit objective curves, using multilayer neural networks. The method is tested on the stock market.