Neural Networks in the Capital Markets: An Application to Index Forecasting

1995/01/01 by Christian Helmenstein, Helmenstein, Christian, Christian Häfke +1
Decision Sciences · Engineering · #Energy Load and Power Forecasting #Stock Market Forecasting Methods

paper · pdf · doi:10.57938/8d439ac2-60e6-4f1f-a539-c7d1c0ba39a6

openalex publication_date 1995/01/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/01

Abstract

In this article we construct an Index of Austrian Initial Public Offerings (IPOX) which is isomorph to the Austrian Traded Index (ATX). Conjecturing that the ATX qualifies as an explaining variable for the IPOX, we investigate the time trend properties of and the comovement between the two indices. We use the relationship to construct a TJ.eural network and a linear error-correction forecasting model for the IPOX and base a tracling scheme on either forecast. The results suggest that trading based on the forecasts significantly increases an investor's return as compared to Buy and Hold or simple Moving Average trading strategies.

Related

• "It Looks All the Same to Me": Cross-index Training for Long-term Financial Series Prediction
• High-performance stock index trading: making effective use of a deep LSTM neural network
• Forecasting VIX using interpretable Kolmogorov-Arnold networks
• S&P 500 Stock's Movement Prediction using CNN
• Stock Price Forecasting and Hypothesis Testing Using Neural Networks