连续双向拍卖中的“增强零信息”策略研究  

Study of Zero-Intelligence Plus Trading Strategies in Continuous Double Auction Market

作　　者：詹文杰[1] 占俊英[2] 

机构地区：[1]华中科技大学管理学院,武汉430074 [2]中南财经政法大学,武汉430060

出　　处：《管理评论》2008年第5期44-50,共7页Management Review

基　　金：国家自然科学基金资助项目(70301013)

摘　　要：连续双向拍卖市场中的"增强零信息"(Zero-Intelligence Plus,简称ZIP)交易策略具有自学习能力,能根据交易成功与失败的价格,相应调整利润率来提高或降低报价。本文研究了ZIP策略的收益率自学习收敛过程,以及静态和动态条件下ZIP与理性"零信息"策略(Zero Intelligence with constraint,简称ZI-C)的优劣比较。研究发现:在静态条件市场条件下,由于ZIP的自学习能力,使得ZIP策略优于ZI-C策略优;而在动态市场条件下,由于ZIP的自学习能力失去作用,ZIP策略与ZI-C策略无差异。The Zero-Intelligence Plus (ZIP) trading strategy has the self-learning capability to adjust its profit margin according to every successful or failed transaction. This paper studies the learning process of ZIP's profit margin firstly, and then compares the ZIP strategy with Zero-Intelligence with constraint (ZI-C) strategy in statistic and dynamic demand-supply conditions. We find that the ZIP beats the ZI-C in the statistic conditions for its self-learning capability, but both of them seem not different in dynamic condition for the reason that the ZIP loses its learning capability

关 键 词：连续双向拍卖 增强零信息 仿真实验 

分 类 号：F713.359[经济管理—产业经济] F224