Application Of Cascade-Correlation Neural Networks In Developing Stock Selection Models For Global Equities

Abstract

We investigate the potential of artificial neural networks (ANN) in the stock selection process of actively managed funds. Two ANN models are constructed to perform stock selection, using the Dow Jones (DJ) Sector Titans as the research database. The cascade-correlation algorithm of Fahlman and Lebiere (1990/1991) is combined with embedded learning rules, namely the backpropagation learning rule and the extended Kalman filter learning rule to forecast the cross-section of global equity returns. The main findings support the use of artificial neural networks for financial forecasting as an active portfolio management tool. In particular, fractile analysis and risk-adjusted return performance metrics provide evidence that the model trained via the extended Kalman filter rule had greater strength in identifying future top performers for global equities than the model trained via the backpropagation learning rule. There is no distinguishable difference between the performances of the bottom quartiles formed by both ANN models. The zero-investment portfolios formed by longing the top quartiles and simultaneously shorting the bottom quartiles or the market proxy exhibit statistically significant Jensens alpha and continues to accumulate positive returns over the out-of-sample period for both ANN models. On the other hand, the zero-investment portfolios formed by longing the bottom quartiles and simultaneously shorting the market proxy exhibit statistically significant Jensens alpha and continues to accumulate losses over the out-of-sample period for both ANN models. The implementation of the extended Kalman filter rule in training artificial neural networks for applications involving noisy financial data is recommended.

Nonlinearities In Stock Return Prediction: Evidence From South Africa

“…This allows direct comparisons to be made between the performances of the linear expected return factor models and their nonlinear counterparts presented in this paper. We also intend to draw a comparison with the results of Hodnett and Hsieh (2012) When developing a model to project the expected return of each share over the in-sample period, the objective is to maximize the accuracy of the estimation over the in-sample period. As is with the factor selection procedure of the linear factor models, nonlinear models with differential style factors are constructed.…”

“…The extended Kalman filter, based on state-space theory, is an alternative learning algorithm designed to update/adjust the network weights. Both mechanisms are illustrated in detail in Hodnett and Hsieh (2012) Activation is passed forward from the input layer to the first hidden neuron (first hidden layer). Activation then passes to the second hidden neuron (second hidden layer) and finally to the output neuron (output layer).…”

“…All ANN models in this research are trained within the cascade-correlation learning algorithm. This algorithm trains the weights using an embedded backpropagation or extended Kalman filter learning rule described in Hodnett and Hsieh (2012).…”

“…There is also no significant difference between the CAPM and the three-factor model, both in the linear and the nonlinear case. Hodnett and Hsieh (2012) investigate the potential of ANNs in the stock selection process of global actively managed funds over the examination period from 2004 to 2009, using the Dow Jones Sector Titans as the research database. Two ANN models are constructed within the cascade-correlation architecture, with 17 style http://www.cluteinstitute.com/ © 2012 The Clute Institute attributes as candidate model inputs in the cross-section of global stock returns.…”

Application Of Resilient Long-Short Strategies For Taiwanese Hedge Funds

“…The optimal makeup of the genes (style attributes) for each chromosome (investment style) is estimated using a genetic algorithm (GA) proposed in Hodnett and Hsieh (2012). Thus, the final candidates (of stocks) representing the "survival of the fittest" (based on evolutionary theory and natural selection) are retained in the fund.…”