基于情绪的投资者行为研究

【作者】 薛斐；

【导师】 姜波克；

【作者基本信息】 复旦大学 ， 金融学， 2005， 博士

【摘要】 现代金融理论认为在理性人假设前提下,市场是有效的,投资者个体行为对市场没有影响,即使存在犯错误的投资者,也会被套利者所淘汰,在这个理论框架下对投资者行为的研究就没有太大的意义。但是行为金融理论放松理性人的假设,在传统金融理论的基础上提出更贴近现实的解释,认为投资者的行为和决策会影响资产定价和金融市场,市场不是有效的。在这个框架下,研究投资者的决策模式和特征,受什么因素影响,投资者的行为会产生什么样的影响就显得非常重要。 本文从投资者情绪研究出发,尝试探索投资者情绪对其行为的影响机理,在投资者情绪影响下投资者行为对资产定价和资本市场以及实体经济的影响机制,通过实证研究方法寻找能够更好衡量我国投资者情绪变化的指标,最后在这个分析框架下以我国投资者的过度自信信念为例进行实证,分析这种信念的影响。 论文主体分四部分。首先,论述投资者情绪对其行为的影响。本文主要从认知结构、信息处理过程和决策动机三个方面来研究投资者的决策模式。人的决策主要由直觉系统和推理系统做出,推理系统对直觉系统的监控是松懈的,这就使得启发式判断有了可能。投资者在接受信息后并不总是进行贝叶斯推断,而是首先经过简化、心理账户、可得性和忽略等方式降低信息的复杂性,这个处理过程往往使得其判断产生系统性偏差。投资者的决策动机并不只是收入或者财富的最大化,其往往会为了心理的满足而付出收入的代价,这种心理的满足主要是两个:避免认知失调和满足控制欲。 其次,论述在以上决策模式影响下的投资者行为对资产定价和金融市场的影响。通过建模的方法说明在以上决策模式影响下的噪音交易者和套利者进行博弈的过程中未必会被淘汰,这时投资者选择情绪投资策略可能更加明智。在以上决策模式影响下的噪音交易者行为形成噪音交易者风险,这会增加资产价格的波动性,减少实体经济的融资量。从宏观调控的角度看,可以采取三种措施来降低噪音交易者的交易:公开市场操作;征收交易税;征收短期资本收益税。在噪音交易者和套利者的博弈过程中,套利者的作用并不总是稳定性的,当套利者可以理性预期到噪音交易者的行为模式的时候,会通过正反馈交易来获利,这会加剧市场的波动性。 第二部分认为情绪投资策略可能是投资者可选的最佳投资策略,这就要求准确衡量市场中的投资者情绪。第三部分通过比较自己编制的加权平均和算数平均的封闭式基金折价指数、国家统计局发布的消费者信心指数、深圳证券信息公司调查获得的个体投资者信心指数等指标,以实证的方法得出结论:消费者信心指数是衡量我国投资者情绪的最佳变量。以封闭式基金折价来衡量我国投资者情绪更多还原

【Abstract】 The agent is rational and the market is efficient in The Modern Financial Theory (MFT). So the investor behavior has no influence on financial market, even if there are irrational investors, they will be wash out by the arbitrageurs. So it is not significance to research on investor behavior. But the Behavioral Finance Theory (BFT) relaxes the rational agent hypothesis and gives the explanation more close to the reality than MFT. The investor behavior has more influence on asset pricing and financial market in BFT.The thesis focuses on investor sentiment and the influence of investor sentiment on investor behavior and tries to find the mechanism of the investor behavior based on investor sentiment influencing the asset pricing and economy. The thesis tries to measure the investor sentiment with the empirical studies through the compares of closed-end fund discount index, consumer confidence index, the investor confidence index. The thesis tests the framework through the overconfidence belief in Chinese online trader.The thesis contains four main parts. First, the mechanism of investor sentiment to behavior is focus. The thesis research investor decision-making model through cognition structure, information processing and decision motives. The agent decision-making is constrained by intuition system and reasoning system. The monitor of reasoning system to intuition system is lax. So the agent makes the heuristic judgments. When the agent gets the information, he will reduce the complexion through simple, mental accounting, availability and ignorance. The information process makes systematic biases. The agent decision motives is not only the income or wealth maximization, the psychology satisfactions is one of the most important motives. The psychology satisfactions are avoiding cognitive dissonance and satisfying the control feeling.Second, through the model construction, a conclusion is got that the noise trader cannot be washed out by the arbitrager in the game between the noise trader and the arbitrager. In this situation the sentiment investment strategic may be the best strategic. More noise trader can create more noise trader risks, more volatility of asset price and less financial quantity. From macro control aspects, three measures should be taken to reduce the noise trade: open market operation; trade tax and short-term capital income tax. In the game between noise trader and the arbitrageur, the role of arbitrageur is not更多还原