Asymmetric Information, Stock Price Volatility Mixed Transition Behavior Research

In most cases, the market is not completely balanced. With the establishment of the GEM market in China, so we once again cast our eyes on the GEM board. It makes us consider the practical significance of the GEM board's existence. Financial time series date has several "stylized facts", they are volatility clustering, fat tails of the distribution, leverage effects and so on.In 2005, Mahen & McCurdy discussed the statistical behaviors and applicability of the GARCH-JUMP model, in the use of the jump intensity and the jump size, and given the rules of asset price behavior and a series of economic significance conclusions. But the paper does not consider the bidirectional feedback effect of jumping and fluctuations.We will research this problem and provide empirical evidence.First of all, through analysis the characteristic of GEM market price behavior, we build GARCH-JUMP model, unlike typical models, we consider the feedback effect of information.Secondly, We here employ the assumption that the arrival of jump components follows a Possion distribution, The analogy is similar to Mahen & McCurdy's research thought, and gives the comprehensive consideration of GEM market price behavior, In particular of the statistical characteristics of the model, then we undertake some likehood-ratio tests for the parameters in the jump part, convert the maximum likelihood estimation into an optimization problem, using the simplex method to get the parameter estimations. Compared with the usual Newton-Raphson optimization method, it reduces the complexity of iterative computation.Finally, we apply our model to three indices. All indices display significant jump clustering. Our model provides superior out-of-sample conditional variance forecasts relatives to a popular benchmark model.