MD&A纵向文本相似度与分析师盈余预测准确性

MD&A Longitudinal Text Similarity and Analysts' Earnings Forecast Accuracy

中文关键词:  管理层讨论与分析  纵向文本相似度  分析师盈余预测  增量信息  资本市场信息效率

英文关键词:management discussion and analysis  longitudinal text similarity  analysts' earnings forecast accuracy  incremental information  information efficiency in the capital market

中文摘要:

      注册制下,信息的真实、准确、完整是投资者合理判断证券价值的关键,分析师在信息作用于股价这一过程中扮演着重要的信息中介角色,而年报所披露的信息对分析师盈余预测至关重要。以2012—2021年沪深A股上市公司为样本,探究了管理层讨论与分析(MD&A)纵向文本相似度对分析师盈余预测准确性的影响。研究发现,MD&A纵向文本相似度越高,分析师盈余预测准确性越低,证实了“增量信息”假说。异质性分析表明,这种影响显著存在于规模小、高科技行业、媒体关注度低的公司中,并且公司位于市场化程度低省份时受此影响更大。进一步研究发现,MD&A纵向文本相似度越高,分析师盈余预测的正向偏差越大。此外,分析师盈余预测准确性的下降会进一步降低资本市场的信息效率。因此,上市公司应重视文本信息披露,避免披露内容样板化；分析师和监管部门应加强对上市公司文本信息披露的监督,优化资本市场信息环境。

英文摘要:

      In the registration system, the truthfulness, accuracy and completeness of information are crucial for investors to make a reasonable assessment of the value of securities. Analysts play a vital role as information intermediaries in the process of information impacting stock prices, and the disclosure of information in annual reports is critical for analysts' earnings forecasts. This paper takes A-share listed companies in Shanghai and Shenzhen from 2012 to 2021 as samples and explores the impact of management discussion and analysis (MD&A) longitudinal text similarity on analysts' earnings forecast accuracy. The findings indicate a negative correlation between MD&A longitudinal text similarity and analysts' earnings forecast accuracy, confirming the “incremental information” hypothesis. Heterogeneity analysis shows that this correlation becomes more significant in companies characterized by small scale or low media attention, and those operating in the high-tech industry as well as located in provincial-level regions with low levels of marketization. Further research finds that an increased MD&A longitudinal text similarity leads to a higher positive bias in analysts' earnings forecasts, and a decrease in analysts' earnings forecast accuracy will further reduce the information efficiency in the capital market. Therefore, listed companies should pay attention to textual information disclosure to avoid excessive similarity in content. At the same time, analysts and regulators need to enhance their supervision over textual information disclosure of listed companies to improve the information environment of the capital market. 