B101 Wavelet解析による改良グランビルの法則に基づく株価予測

抄録

In order to predict the stock market price, modified Grandville law based on the wavelet analysis were developed. The present method improved profit of the stock trading. Since the conventional Grandville-law is not suitable to estimate the turnover period in the collapse region, we cannot estimate the suitable time for the stock trading. The wavelet analysis estimates the higest and lowest price in the long period. However the wavelet analyses are sometimes misreading the turnover. The present method modified theses weak points of Grandville and wavelet method. This new method can be controlled stock trading in the collapse period. Moreover, the stock was purchased in the lowest price with this method. The simulated data showed the market price was increasing after the stock buying. The simulation result showed that the average profit of 7 descriptions was over 35% of the capital. It was 10% larger than that of the conventional Grandville law.