Empirical Researches Of Agricultural Futures Market In China Based On Fractal Analysis

As a cornerstone of modern financial analysis system, the Efficient Market Hypothesis has been playing a crucial role in the development of financial theory. The Efficient Market Hypothesis thinks that people are rational, and the information is valid at each time point. However, many financial phenomena in the real can not be explained by the Efficient Market Hypothesis. With the rapid development of nonlinear science and complexity science, the Fractal Market Hypothesis emerged from the fractal theory and chaos theory. In this dissertation, we analyze the price return series of cotton and strong wheat in Zhengzhou Commodity Exchange, and the price return series of corn and soybean in Dalian Commodity Exchange based on fractal theory and methods. Their asymmetric multifractal characteristics and multiscale multifractal characteristics are studied for the first time.Firstly, we reduce auto-correlation in price series by the use of logarithmic rate of return method, and analyze basic statistical characteristics of return series of futures prices, then perform normality test by P-P graphs. The results show that the averages of price return series of the four species are close to zero, which means that the sequences have rebalancing function; the four varieties price series have "peal and fat rail" features, meaning that they does not meet the normality and have clear non-linear characteristics.Secondly, we use asymmetric detrended fluctuation analysis (abbreviated A-DFA), to investigate agricultural futures prices state characteristics in the rise and decline. The results show that the rise and decline sequences of the four species have asymmetric, and corn, strong wheat, soybean, cotton, in turn weakened. Agricultural prices of the rise and decline trend series show obvious fractal characteristics.Thirdly, we use multifractal detrended fluctuation analysis (abbreviated MF-DFA) and multifractal spectrum analysis (abbreviated MF-SA) to investigate the structural characteristics of returns series of agricultural futures prices. The results show that the agricultural price series have multifractal features. The multifractal strength of strong wheat, cotton, corn, soybean meal in turn weakened, which means the investment risk of the corresponding futures is gradually reduced.Finally, the MF-DFA method will be extended to multiscale case (abbreviated MM-DFA). We investigate the multifractal characteristics of the return series in the multiscale case, and analyze the necessity of the research. Compared with the conventional MF-DFA method, this method can provide a more reliable and more interpretable description for the dynamic mechanism of price series.