Identification and Ranking of Effective Factors on Stock Return Synchronicity Using Neural Networks and Decision Tree Model

Autor: Mohsen Lotfi (Ph.D), Hamid Haghighat (Ph.D), Mohammadhossein Ghaemi (Ph.D)
Jazyk: perština
Rok vydání: 2019
Předmět:
Zdroj: مجله دانش حسابداری, Vol 9, Iss 4, Pp 1-36 (2019)
Druh dokumentu: article
ISSN: 2008-8914
2476-292X
DOI: 10.22103/jak.2018.11404.2571
Popis: Objective: This research aim at identification of the factors affecting the stock price synchronicity and their ranking in a sample of 1030 years-companies from the Tehran Stock Exchange in the years 2006 to 2015. Methods: This research takes a sample that was selected through systematic elimination method, and examines the effects of macroeconomic variables (economic growth, inflation and unemployment rate), corporate governance system (institutional shareholder, concentration of ownership, and independence of the board), audit quality, agency problems (free cash flow, complexity of operations and information asymmetry), and company information characteristics (profit sustainability, conservatism, earning smoothing and opaque) on stock price synchronicity based on data mining methods (including neural network model and C5 tree model). Results: The results showed in corporate governance, the focus of ownership (based on the percentage of free float) and the independence of the board; in agency problems, the information asymmetry; in company information characteristics, the conservatism and opaque; and from macroeconomic variables, the economic growth have the most impact on the stock returns synchronicity. Also, the information asymmetry has the most impact on stock return synchronization based on different criteria. Conclusion: Given that the stock price synchronicity is considered the indication of stock price informativeness (the extent of access to specific company information) and efficiency of capital allocation, therefore, it is recommended to the capital market activists to pay special attention to the factors affecting this variable to improve their decisions.
Databáze: Directory of Open Access Journals