基于网络学习策略的股市投资者行为演化研究

Abstract

随着全球经济一体化发展，股票市场波动性日趋频繁，导致投资者的行为决策呈现出一定程度的交易频繁、投资盲从等市场异象，而这又易反作用于市场，加剧股票市场的不稳定性，因此有必要对投资者行为演变特征进行深层次的分析。基于此，本文从股市投资者“学习”属性视角出发，首先分析了投资者行为特征，且在投资者行为特征的基础上分析了投资者学习策略；其次探讨了投资者行为结构，并由投资者间行为结构的差异性引出投资者网络结构的异同，最后总结了在网络学习策略下股市投资者行为演化的特征，并在此基础上构建了基于网络学习策略的股市投资者行为演化理论模型，为模拟或实证分析投资者行为演化特征提供了理论依据。 The stock market is an important part of the modern financial system, which plays an important role in improving the efficiency of capital allocation, promoting the healthy development of economy and guaranteeing the return of investors. With the development of global economic in-tegration, all kinds of uncertainty factors interweave increasingly frequent volatility in stock markets, causing some irrational behavior among investors, such as “gambling” investment, over-confidence, herd behavior. In turn, investors' irrational investment behavior has caused the vola-tility of the stock market, so it is necessary to deeply analyze the evolution characteristics of in-vestors’ behavior. Based on this, from the angle of view of the “learning” of stock investors, firstly, this paper analyzes the characteristics of investors’ behavior, and on this basis, analyzes the in-vestors’ learning strategies. Secondly, it discusses the structure of investor behavior and the structure of investor network. Finally, it summarizes the theoretical evolution characteristics of investors’ behavior in network, and then builds the evolution model of investors’ behavior in stock market based on learning strategy in network, which provides a theoretical basis for simulation or empirically analyzing evolution characteristics of investors’ behavior. The evolution model of investors’ behavior in stock market based on learning strategy in network is as follows. Due to the difference of the structure of investor behavior, different networks are formed among investors. Under the network, investors’ learning strategies are different because of the different characteristics of investors’ behavior. Based on the independence of investors, investors form en-dogenous learning strategies which mainly include reinforcement learning strategies and Bayesian learning strategies; and investors choose exogenous learning strategies because of social in-teraction among investors, which mainly include imitation learning strategies and game learning strategies. Investors with heterogeneous beliefs choose different learning strategies according to their own investment behavior characteristics, and based on the learning strategies which they choose, investors make investment decisions, so as to realize the evolution of investors’ behavior in stock market, and at the same time in the process of the evolution of investor behavior, the structure of investors behavior and the characteristics of investors behavior are also in constant change, and the environment of stock market will be affected.