摘要
用一种新的信号处理工具—小波变换 ,对股价指数数据进行分析 ,发现股价指数数据类似于一类更广的噪声—分形噪声 ,从而推广了传统上处理股价指数时间序列时总假定其为白噪声或高斯噪声的假设 ,用分形噪声能更好地刻划股价指数数据的波动特性 .对上证指数和深证指数的实证分析显示 ,两市股价指数均存在正相关 ,我国股票市场不是弱式有效市场 .
Wavelet transform is a new signal process method. We have used it to analyze stock price index and found that it is well characterized by fractal noise that is more general than white noise or gauss noise. Therefore, fractal noise can describe the property of stock price index more precisely. The results show that the stock price index of Shanghai and Shenzhen exist positive correlation and the stock market is not low efficient. It also shows that wavelet transform is an effective method for study stock character.
出处
《经济数学》
2000年第1期25-30,共6页
Journal of Quantitative Economics