Aligning Asset Pricing Models and Neural Networks for Predicting Portfolio Returns in Frontier Markets

Forecasting Portfolio returns is a challenging task, and conventional forecasting models have partially succeeded in dealing with the nonlinear and complex nature of Equity Markets. Artificial neural networks are a mathematical modeling approach that is resilient enough to forecast portfolio returns in volatile and nonvolatile markets and act like the human brain to simulate the behavior of stock prices. This research documents the predictive ability of Artificial Neural Networks (ANN) by using the constructs of Fama and French three-factor and five-factor models. A comprehensive methodology of neural networks is applied to achieve the purpose of forecasting. This methodology includes the declaration of the internal layers, the hidden layer neurons, and varying parameters for an effective ANN system. A rolling window scheme is applied to forecast the errors among the competing asset pricing models. The predictive performance of ANN is measured by the metric of mean squared error, and the accuracy of ANNs under both pricing models is evaluated by the Diebold Mariano test. The significant findings of the study include the identification of the optimum architecture of the ANN under both asset pricing models, the non-overfitting phenomenon of the networks, and the abnormal returns for the investors for holding high-risk portfolios.