股票市场交易策略模拟研究-管理科学与工程专业论文.docxVIP

摘要 行为金融学和计算实验金融的出现为股票市场的研究提供了新的思路和手 段。本文利用行为金融学理论和计算实验金融方法，以 Netlogo 为平台，构建人 工股票市场。该人工股票市场以事件驱动，市场的出清机制和价格形成机制均由 供求关系决定。市场中有两种交易资产，有若干交易者，他们随机选择动量、反 向、噪音和理性四种投资策略。本文旨在分析动量交易者、反向交易者、噪音交 易者以及理性交易者的交易行为和策略转换，并通过收益对比判断这四类交易者 的市场存在情况，为真实的金融市场交易提供部分依据。 仿真分析结果指出，动量交易者的数量变动趋势和反向交易者的数量变动相 ，萨 反，噪音交易者和理性交易者的数量相对稳定，不会被驱逐出市场。羊群行为对 噪音交易者的数量变动有重大影响。当羊群效应成为噪音交易者考虑执行策略的 主要影响因素时，噪音交易者会以周围交易者的行为作为其选择投资策略的标 准。反向交易者和动最交易者之一可获得明显的超额收益，但是，无论是反向交 易者还是动量交易者获胜，均表现出相当大的收益波动率，也就是说投资者承受 的风险也相对大的多。噪音交易者的收益总是高于理性交易者，而且收益波动率 也不显著，这也证实噪音交易者不会被市场驱逐。 关键词:计算实验金融:动最策略;反向策略:噪音策略。 、队 、 ABSTRACT ηle appearance of Behavioral Finance and Agent-based Computational Finance provides new 由oughts and methodologies for the research on the stock market. This article makes full use of theories in Behavioral Finance and Agent-based Computational Finance，trying to establish an artificial stock market on the Netlogo platfonn. The artificial stock market is driven by 伽 special events and rules for market clearing and prices forming are determined by s叩ply and demand of 州附. 咀lere are 仰o assets 出 the markets and several traders. Each trader will randomly choose his or her initial trading strategy.ηlere are four kinds of strategies available to 护 choose，which are momen阳ms衍创egy，contrarian strategy，noise strategy as well as rational strategy. We aim to observe the trading behaviors of four kinds of traders and analyze the outcomes of strategy transfonnation in order ωveri命 the existence of these different kinds of traders. The research shows 由at the number of traders in momen阳m 耐ategies has a reverse changing tendency wi由that oftraders inωntrarian strategies. The number of traders in noi8e traders and rational trader百 tend to be stable and will not be 由iven out of the market. Heard behavior plays a key role in affecting the amount of noise traders. Traders in momentum 8tr副egy or trad时s in contrarian 8位ategy wi1l won the superior earnings; however，由e volatility of profit is extremely lar;肘， which