Hurst And Multifractal Analysis On The China Stock Market

EMH is necessary to prove that price follows a random walk model. But EMH is established on market, which is independent normal distribution and variance existing. But the market is a complicated interdependent system. A lot of cases are what EMH can't explain, such as scale effect, season effect etc. So a lot of theories that explain these special cases have appeared, such as catastrophe theory, behavior finance theory, coherent market theory and the fractal market theory, which let researchers taking more attention more about the use of non-linear theory in financial market.Fractal theory is one kind of nonlinear analysis and has only decades since it developed, because it can efficiently expenses many complicated problem, such as economics phenomena, so it has became the strong analysis tool of the nonlinear capital market theory.There are five parts in our paper, the first is introduced simply the theories of EMH and FMH; the second is introduced particularly the method of Hurst index and multifractal; and in the third we get the conclusion that China stock market is FMH with analysis using the date of Shanghai from 2 Jan. 1991 to 31 Aug. 2005 and Shenzhen from 1 Apr. 1991 to 8 Jan. 2006; in the forth we get the conclusion that China stock market is MFMH with analysis the date of Shanghai and Shenzhen. The fifth is conclusion and expectation.In our paper, researching on the index of shanghai and Shenzhen bond market, we find that our bond market has obvious fractal character, and working out the Hurst index and V index, which prove that our bond market can be test by non-linear theory. We can get the 300 days as a period, and the H is 0.626358 of shanghai bond market; the same as 1200 days and H is 0.590986 of Shenzhen bond market. In the last we use Box-counting method to prove that Shanghai and Shenzhen bond market is mutifractal.