智能财务分析与诊断机器人的开发及实证检验——来自我国A股上市公司的经验证据  

Development and Empirical Test of Intelligent Financial Analysis and Diagnosis Robot:Evidence from China’s A-Share Listed Companies

作　　者：吴世农[1] 林晓辉[2] 李柏宏 王举明 Wu Shinong;Lin Xiaohui;Li Baihong;Wang Juming(不详)

机构地区：[1]厦门大学管理学院,福建厦门361005 [2]厦门大学公司金融研究中心,福建厦门361005

出　　处：《证券市场导报》2021年第2期62-71,78,共11页Securities Market Herald

基　　金：国家自然科学基金重大项目“中国制度和文化背景下公司财务理论与实践研究”(批准号:NSF717906001)

摘　　要：近年来,随着人工智能技术的发展,金融、财务、投资、审计与会计的智能化问题引起广泛关注。本文应用人工智能技术,依据财务分析的基本理论和方法,结合中国上市公司财务特征,率先尝试开发了"智能财务分析与诊断机器人",用于评价上市公司的综合财务绩效。该款智能机器人具有"速度快""智能化程度较高""专业能力较强""相对公正客观"和"实用性较强"五大特点。本文应用该款机器人设计了两个投资决策实验,个股和组合实验结果均表明该款智能机器人能有效地评价和区分上市公司的财务绩效类型。根据其输出的综合绩效,买入高绩效型股票或组合,或买入绩效成长型的股票或组合,具有显著的超额收益;买入高绩效型股票或组合,比买入绩效成长型的股票或组合,具有更高和更稳定的超额收益。可见,基于人工智能的财务分析与诊断机器人具有稳健和有效的择股能力。In recent years,with the development of artificial intelligence technology,the intellectualization of finance,investment,auditing and accounting has drawn wide attention.Based on the basic theories and methods of financial analysis,combined with the financial characteristics of listed companies in China,this article uses artificial intelligence technology to firstly try to develop“intelligent financial analysis and diagnosis robot”to evaluate the comprehensive financial performance of listed companies.The results show that this intelligent robot has five characteristics:“fast speed”,“higher intelligence”,“strong professional ability”,“relatively fair and objective”,and“strong practicality”.On this basis,the study applies this AI robot to design two investment decision-making experiments,and the results of both individual stocks experiments and portfolios experiments show that this intelligent robot can effectively evaluate and distinguish the types of financial performance of listed companies.According to the comprehensive financial performance it outputs,we find that buying high-performance stocks or portfolios,or buying increasingperformance stocks or portfolios,can obtain significant cumulative abnormal returns.Moreover,buying high-performance stocks or portfolios has higher and more stable cumulative abnormal returns than buying increasing-performance stocks or portfolios.This shows that the AI robot has robust and effective stock selection capabilities.

关 键 词：智能财务 AI机器人 绩效评价 择股能力 

分 类 号：F830.59[经济管理—金融学]