有限关注视角下投资者本地偏好对股价同步性的影响研究

Research on the Impact of Investor Local Bias on Stock Price Synchronicity from the Perspective of Limited Attention

摘要: 中国资本市场“散户主导”的结构特征与股价“同涨同跌”现象并存，引发了对微观主体行为与市场效率关系的深层思考。本文以行为金融学中的“有限关注”理论为视角，利用2022年“IP属地实名制”全面实施后的数据契机，整合东方财富股吧用户互动行为与百度搜索指数，构建了双因子投资者本地偏好指数，并以2024年沪深A股上市公司为样本，实证检验了本地偏好对股价同步性的影响及其作用机制。研究发现：第一，中国个人投资者对本地上市公司表现出显著的关注聚集特征，且相较于单一维度的关注度指标，整合了“深度互动”与“主动搜寻”的双因子指标在度量本地偏好时具有更强的稳健性与解释力。第二，投资者本地偏好显著推高了股价同步性，支持了“行为噪音假说”。这表明在有限关注的认知约束下，本地偏好并非基于信息优势的理性挖掘，而是更多体现为基于熟悉感的地缘情绪共振与羊群行为，这种同质化交易强化了市场噪音，削弱了定价效率。第三，异质性分析显示，在信息透明度较低(低分析师覆盖)和散户定价权较大(小市值)的公司中，本地偏好对股价同步性的推高效应更为显著。本研究厘清了有限关注下本地偏好发挥“行为噪音”作用的边界条件，为监管部门优化信息披露制度及引导投资者理性行为提供了经验证据。

Abstract: The coexistence of the “retail-dominated” structure and the “co-movement” phenomenon of stock prices in China’s capital market has triggered extensive investigation regarding the relationship between micro-level investor behavior and market efficiency. From the perspective of “limited attention” theory in behavioral finance, leveraging the unique data opportunity provided by the full implementation of the “IP address location mandate” in 2022, this paper constructs a dual-factor Investor Local Bias Index (LBI) by integrating user interaction data from Eastmoney stock message boards and the Baidu Search Index. Using a sample of Shanghai and Shenzhen A-share listed companies in 2024, this study empirically examines the impact of local bias on stock price Synchronicity and its underlying mechanisms. The findings indicate that: First, Chinese individual investors exhibit significant geographic concentration of attention toward local listed companies. Compared to single-dimensional attention metrics, the dual-factor index, which integrates “deep interaction” and “active search”, demonstrates superior robustness and explanatory power in measuring local bias. Second, investor local bias significantly increases stock price Synchronicity, supporting the “behavioral noise hypothesis”. This suggests that under the cognitive constraints of limited attention, local bias does not stem from rational information exploitation based on information advantages, but rather reflects geographic emotional resonance and herding behavior driven by familiarity. Such homogenous trading reinforces market noise and impairs pricing efficiency. Third, heterogeneity analysis reveals that the positive effect of local bias on stock price Synchronicity is more pronounced in firms with lower information transparency (low analyst coverage) and stronger retail pricing power (small market capitalization). This study clarifies the boundary conditions under which local bias acts as “behavioral noise” amid limited attention, providing empirical evidence for regulators to optimize information disclosure mechanisms and guide rational investor behavior.