Cluster analysis of share price: How firm characteristics relate to accounting metrics

Risk Governance & Control: Financial Markets & Institutions

article

Mfon Akpan¹  Guneet Dhillon²  Kim Trottier³ 

[1] Department of Accounting, Methodist University;Faculty of Computer Science Alumni, Dalhousie University;Department of Accounting

关键词: Share Price;    Earnings;    Financial Accounting;    EPS;    Cluster Analysis;    Accounting Numbers;    Industry;    Text Mining;   

DOI  :  10.22495/rgcv11i4p1

学科分类：社会科学、人文和艺术（综合）

来源: Virtus Interpress

【 摘 要 】

The purpose of this paper is to improve our understanding of the relationship between share price and accounting information. Much of the literature utilizes the earnings number to reflect firm value. However, the revenue number seems more relevant for high tech firms (Xu, Cai, & Leung, 2007), and cash flow figures are more informative for internet companies (Romanova, Helms, & Takeda, 2012). We build on this notion that share price may map out to different accounting numbers for different firms. We collect 629 accounting metrics for 3,365 firms in the U.S. and estimate their correlation with the firms’ share price. We analyze these correlations and find that many firms exhibit a low correlation between share price and earnings. Other accounting numbers are important for these firms, including book value of net assets, retained earnings, stock options, gain or loss items, special or non recurring items, and dividend rates. We are curious to learn what causes firms to anchor onto different metrics, therefore perform a cluster analysis to group similar firms together along three key accounting metrics. We examine the composition of each cluster and find that capital structure, dividend patterns, the persistence of operations, age, and industry can influence which accounting number is correlated with firm value. We encourage other researchers to continue this exploration as there are many interesting questions to answer.

【 授权许可】

CC BY-NC