Table of Contents

Recurrent Neural Networks in Forecasting S&P 500 Index

Samuel Edet, African Institute for Mathematical Sciences


"Recurrent Neural Networks in Forecasting S&P 500 Index" Free Download

SAMUEL EDET, African Institute for Mathematical Sciences

The objective of this research is to predict the movements of the S&P 500 index using variations of the recurrent neural network. The variations considered are the simple recurrent neural network, the long short term memory and the gated recurrent unit. In addition to these networks, we discuss the error correction neural network which takes into account shocks typical of the financial market. In predicting the S&P 500 index, we considered 14 economic variables, 4 levels of hidden neurons of the networks and 5 levels of epoch. From these features, relevant features were selected using experimental design. The selection of an experiment with the right features is chosen based on its accuracy score and its Graphical Processing Unit (GPU) time. The chosen experiments (for each neural network) are used to predict the upward and downward movements of the S&P 500 index. Using the prediction of the S&P 500 index and a proposed strategy, we trade the S&P 500 index for selected periods. The profit generated is compared with the buy and hold strategy.


About this eJournal

This eJournal distributes working and accepted paper abstracts focused on research where economic outcomes are the product of many individual decisions, constrained by scarcity, and equilibrium forces that simultaneously shape a person's social networks and the institutionally defined rules of the game. Decisions are made by computations in the brain which produce action-choices that directly affect the homeostatic wellbeing of the individual and choices that indirectly change wellbeing by changing an individual's future constraints, the scope of their social networks, and their message sending rights within the institutions they participate. Neuroeconomics broadly speaking is interested in the study of these computations and the resulting choices they produce. This includes experiments that attempt to understand the mechanisms of neuronal computations that produce action-choices, theories which predict how neuronal computations in socio-economic environments produce decisions, outcomes and wellbeing, and policy which use our understanding of neuoroeconomic behavior to either build or defend better solutions to societal problems.

Editors: Michael C. Jensen, Harvard University, and Kevin A. McCabe, George Mason University


To submit your research to SSRN, sign in to the SSRN User HeadQuarters, click the My Papers link on left menu and then the Start New Submission button at top of page.

Distribution Services

If your organization is interested in increasing readership for its research by starting a Research Paper Series, or sponsoring a Subject Matter eJournal, please email:

Distributed by

Economics Research Network (ERN), a division of Social Science Electronic Publishing (SSEP) and Social Science Research Network (SSRN)



SSRN, Harvard Business School, National Bureau of Economic Research (NBER), European Corporate Governance Institute (ECGI), Harvard University - Accounting & Control Unit

Please contact us at the above addresses with your comments, questions or suggestions for ERN-Sub.

Advisory Board

Neuroeconomics eJournal

Harris & Harris Group Professor, Massachusetts Institute of Technology (MIT) - Sloan School of Management, National Bureau of Economic Research (NBER), Principal Investigator, Massachusetts Institute of Technology (MIT) - Computer Science and Artificial Intelligence Laboratory (CSAIL)

Professor, Baylor University - Department of Neuroscience

Professor of Economics and Law, Chapman University - Economic Science Institute, Chapman University School of Law