基于指令驱动市场EKOP模型的异质期望研究  

Heterogeneous Expectations Study Based on EKOP Model under Order Driven Market

作　　者：王硕[1] 李鹏程[1] 杨宝臣[1] 

机构地区：[1]天津大学管理与经济学部,天津300072

出　　处：《管理科学》2016年第3期123-135,共13页Journal of Management Science

基　　金：国家自然科学基金(71171144;71471129);教育部长江学者和创新团队发展计划项目(IRT1028)~~

摘　　要：传统资产定价模型假设所有的投资者对于同一种资产未来收益的预期相同。但是,这一假设很难得到实证研究的证实。在现实市场中,投资者往往存在意见分歧,这样的意见分歧就是异质期望。理解异质期望如何影响资产价格是金融研究中一个极其重要也颇具争议的问题,这一课题的研究焦点在于,市场中投资者意见分歧的增大是否产生溢价。实际上,研究投资者意见分歧与资产定价的关系就是研究市场中的信息是如何进入价格的。投资者由市场信息产生或调整自己的预期,通过交易行为体现出来,最终反映在价格中。这一信息的融入和传导过程是市场微观结构研究的核心内容,而这一点在已有研究中往往被忽视。从市场微观结构的角度就投资者预期和异质期望对定价的影响进行研究,结合中国股市交易机制,建立指令驱动系统下的Easley-Kiefer-O'hara-Paperman模型,通过交易活跃度这一参数,以买卖订单交易量为基础,利用带机制转换的向量自回归模型估计信息状态概率,更准确地估计市场中的知情交易比例和不同信息状态下非知情交易者的预期。利用沪深股市29只股票2010年1月至2012年6月的5分钟高频分笔数据进行估计,并在此基础上构建异质期望衡量指标,利用多因素模型回归研究异质期望与股票定价的关系。研究结果表明,这一时期投资者意见分歧较大,市场逐渐低迷,交易活跃度从5.149%降至1.357%,其中知情交易比例为28.184%,比例较高;市场转冷时,知情交易量的减小远小于非知情交易量;非知情交易者对市场中的信息估计不足,但是可以较为准确地从市场交易中判断看涨的走势,却不能在价格下跌时有效止损,这也体现出其投机心理比较严重;垄断优势强的企业投资者意见分歧较小;异质期望水平较高会导致价格被高估,且二者成正比,这一高估会在两个月内得以完全�Traditional asset pricing modal assumes that all the investors should hold the same opinion for future returns of the same kind of assets. However, this hypothesis is difficult to get confirmed by empirical studies. In real markets, investors＇ opinion is different. This disagreement is heterogeneous expectations. Understanding how heterogeneous expectations affect asset prices in financial markets is one of the most fundamental issues in finance, and it is also a controversial one. The focus on the issue is whether the investors earn a premium when disagreement arises in the market. This paper studied how the heterogeneous expectations impact asset pricing from the perspective of market microstructure. The relation of investors＇ disagreement and asset prices is in fact a problem of how information is priced. Investors adjust their expectations based on information, and through the transaction behavior, price is the final reflection of investors＇ opinion. This process is a problem of market micrestructure research, and it has been largely ignored so far. In this paper, an EKOP model under order driven market is established, which is applicable for the Chinese security market. We propose a term called market activity percentage in the model, and estimate the model through trade volumes. The probability of different information is estimated by Markov-switching vector autoregressive model. The proportion of buying, selling and nontrading of non - information traders could be estimated more accurately in this way. 5 minute high-frequency data is selected of 29 individual stocks during the period of Jan 2010 to Jun 2012. We build a proxy for the divergence in opinion based on the esti- mate results, and analyze the proxy and stock price through multiple factor regression. Result shows that the divergence of opin- ion is rather large and investors become pessimistic during this period. Market activity falls from 5. 149% to 1. 357%, and the probability of informed trading is 28. 184%, which is relatively high. The

关 键 词：异质期望 EKOP模型 指令驱动市场 定价 市场信息 

分 类 号：F830.9[经济管理—金融学]