基于序次Probit模型的离散股价研究  

Research on discrete stock price based on ordered Probit model

作　　者：陈磊[1] 李平[1] 曾勇[1] 

机构地区：[1]电子科技大学经济与管理学院

出　　处：《系统工程学报》2009年第5期561-567,共7页Journal of Systems Engineering

基　　金：国家自然科学基金资助项目(70703002);教育部新世纪优秀人才支持计划资助项目(教技司[2005]2号)

摘　　要：分析股票日内价格行为需关注高频分笔交易的价格离散特征.本文使用沪市个股高频分笔交易数据,采用序次Probit模型对离散股票价格建模分析.结合我国股市限制卖空与订单驱动的制度背景,考察了交易时间间隔、交易量和订单信息对价格变动的影响.研究结果表明,序次Probit模型可捕捉价格离散特征.较长交易时间间隔可能预示坏消息,使股价下跌.相对于卖方交易量,买方交易量更具有信息含量.订单信息影响价格变动,买方较多时股价上涨,卖方较多时股价下跌.Stock price discreteness should be taken into account in research on intraday stock price behavior. Using high frequency trade-by-trade data of the stocks in Shanghai stock market, this paper analyzes discrete stock price by ordered Probit model. The paper analyzes the effect of interval be- tween trades, volume, and limit order information on price movements under short-selling forbidden and order-driven mechanism. The results show that the stock price discreteness can be described by ordered Probit model. The longer trading interval can be taken as a signal of bad news, and makes price go down. The buyer volume has more information content than the seller volume. Limit order information can affect price movements. More buyers make price go up, and more sellers make price go down.

关 键 词：高频数据 分笔交易 离散股价 序次Probit模型 限制卖空 

分 类 号：F830.91[经济管理—金融学]