Predict the relationship between stock returns and information asymmetry using artificial neural networks

Document Type : Original Article

Authors

10.22034/iaar.2014.104325

Abstract

Given the importance of return on investment studies to estimate the relationship between information asymmetry and return is an important issue. Changes in efficiency, inadequacy studies, and existence of effective factors are cause of development new and intelligent methods to estimate the stock return of companies. The aim of this study is to predict stock returns using information asymmetry with an artificial neural network approach. The independent variable in this study is information asymmetry and stock return is the dependent variable. Therefore, the variables for the 100 companies in the stock exchange and has been collecting for 6 years. Estimated output of artificial neural networks and the results of the estimation using this approach, with evaluation criteria is (R2= 0.99, MSE= 0.064 and MAE= 0.21). Considering the random value (50%) compared with R2= 0.99, correlation between information asymmetry and stock return are observed. Also, the designed network has the least error (MSE = 0.064 and MAE= 0.21) than the other networks.

Keywords

Accounting and Auditing Research
Volume 6, Issue 24
January 2015
Pages 16-33

Files

Share

How to cite

Statistics