基于小波低频分量的量化择时策略及仿真模拟

Quantitative Timing Strategy and Its Simulation Based on Wavelet Low Frequency Component

文章运用Best Basis Selection（BBS）算法选取最优小波包基,对上证综指收盘价进行小波包非线性阈值消噪,在消除随机性干扰的基础上,针对传统均线策略买卖信号滞后性的不足,根据不同分解水平的小波低频分量能够反映信号基本和次级趋势且不具滞后性的特点,提出了一种基于小波低频分量的量化择时策略,并对该策略和传统的均线策略分别用R语言进行仿真模拟交易和回测,实验表明在类似风险的情况下,该策略在提示基本和次级趋势买卖信号的同时可以缩短交易信号的滞后性,具有更好的投资表现。

Based on Best Basis Selection（BBS） algorithm, this paper selects the optimal wavelet package basis to denoise the wavelet packet nonlinear threshold of the Shanghai Composite Index closing price. Eliminating the random interference, aiming at the shortcoming of the buy-sell signal hysteresis of the traditional average strategy, and according to the characteristics that wavelet low frequency component of different decomposition level is able to reflect the basic signal and the secondary trend and even has no lag, this paper proposes a quantitative timing strategy based on wavelet low frequency component, and uses R language to make a trading simulation and back-test of the strategy and the traditional average strategy respectively. Experiment results show that in the case of a similar risk, the proposed quantitative strategy presents a buy-sell signal of the basic and a secondary trend,and at the same time is able to shorten the lag of the trading signal, thus having a better investment performance.