Asset Bubble Model Based On Hazard Rate Function And Related Data Fitting

This thesis summarized most classical asset bubble pricing models based on the study in asset bubbles'historical, definition, existence, generating process and all kinds of model classification. We generalized the basic periodical bursting bubble model (Evans 1991) by connecting the probability of bubble bursting with the hazard rate function in the field of survival analysis. In the process of picking the hazard rate functions, we constructed a interaction model of agents in the markets in two different ways, 2D-Ising Model and hierarchical diamond lattice model. After picking the hazard rate function, we have made asset bubble numerical simulation with Monte Carlo technical, and got some satisfying results. Finally, we run a fitting of the model with the data from chinese stock markets using a two-step method. We chose the simplex downhill method based on simulating annealing as our optimization method.