The Equity Premium Puzzle: An Artificial Neural Network Approach
Shee Q. Wong
Labovitz School of Business
Nik R. Hassan
affiliation not provided to SSRN
Ehsan H. Feroz
University of Washington - Tacoma-Milgard School of Business; Vernon Zimmerman Center, University of Illinois; US Government Accountability Office
August 8, 2008
Review of Accounting and Finance, Vol. 6, No. 2, pp. 150-161, 2007
This paper presents evidence suggesting that artificial neural networks approach (ANNs) outperform traditional statistical methods and can forecast equity premiums reasonably well. The study replicates out-of-sample estimates of regression using ANN with economic fundamentals as inputs. The dividend yield variable was found to produce the best out-of-sample forecasts for equity premium. This result is useful for capital asset pricing model and in asset allocation decisions.
Number of Pages in PDF File: 23
Keywords: Equity premium, Forecasting, CAPM, Neural networks
JEL Classification: C45, G1, G2, G3Accepted Paper Series
Date posted: August 8, 2008 ; Last revised: June 15, 2009
© 2014 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo3 in 0.672 seconds