基于无标度网络的自组织金融模型研究

发布时间：2018-03-21 18:04

本文选题：无标度网络　切入点：逾渗理论　出处：《中国科学技术大学学报》2014年01期 　论文类型：期刊论文

【摘要】：针对CB模型及其改进模型中由于规则网络描述的均质群体结构与真实金融市场中投资者相互作用的异质性相悖,提出基于表征异质投资群体结构的无标度网络的自组织金融模型.通过投资者在交易规则约束下的自组织聚簇行为,模拟金融市场的动态演化过程.模型生成的价格波动序列与实际股指具有相似的演化动力学:价格收益的概率分布具有尖峰胖尾的特征,且它的中心峰值与时间尺度存在幂律关系,这表明价格波动序列的演化是一个自相似过程;易变性具有明显的聚簇行为,说明价格波动序列具有连续的巨幅涨落和长程关联性.这些统计特性与金融市场实证相符,验证了模型的有效性.
[Abstract]:In view of the heterogeneity of investor interaction in real financial market, the homogeneous group structure described by regular network in CB model and its improved model is contrary to the heterogeneity of investor interaction in real financial market. This paper proposes a scale-free financial model based on scale-free network, which represents the structure of heterogeneous investment groups, through the self-organizing clustering behavior of investors under the constraint of trading rules. Simulation of the dynamic evolution of financial markets. The price volatility series generated by the model has similar evolutionary dynamics to the actual stock index: the probability distribution of price returns has the characteristics of spike and fat tail. And its central peak value has a power law relation with time scale, which indicates that the evolution of price fluctuation series is a self-similar process, and the variability has obvious clustering behavior. It is shown that the price volatility series has continuous large fluctuations and long term correlations. These statistical characteristics are consistent with the empirical results of the financial markets and verify the validity of the model.
【作者单位】：中国科学技术大学电子科学与技术系;
【基金】：国家自然科学基金(61004102)资助
【分类号】：F830.9;O211.3

【参考文献】