Comparison of the potential of chaotic models and artificial neural networks in explaining abnormal stock returns around the release date of annual financial statements

Document Type : Original Article

Authors

1 Ph.D. Student, Department of Accounting, Neyshabur Branch, Islamic Azad University, Neyshabour, Iran

2 Department of Accounting, Neyshabur Branch, Islamic Azad University, Neyshabur, Iran.

3 Assistant Professor, Department of Statistics, Mashhad Branch, Ferdowsi University, Mashhad, Iran

10.22034/iaar.2022.151103

Abstract

Nowadays, the most important criterion for evaluating the performance of business entities is the rate of stock returns. Since human beings have a great interest in predicting future events in order to control the effects of events and minimize the negative consequences of them, predicting and explaining the price and stock returns has always been noticeable issues In the academic field. Therefore, in the present study, abnormal returns' data of 177 listed companies in the Tehran Stock Exchange between 2008 and 2017 were investigated using technical analysis and the discovery of the past trend around the release date of annual financial statements.
SETAR and LSTAR models (chaotic models), the AR model (linear model) and artificial neural network model (using Fama-French's three factors) are also used to predict abnormal stock returns. Finally, the artificial neural network model is chosen as the optimal model.

Keywords

Accounting and Auditing Research
Volume 14, Issue 53
April 2022
Pages 135-154

Files

Share

How to cite

Statistics