This research investigates whether articial neural networks which make use of rm- specic fundamental and technical factors can accurately predict the returns of a sample of several large-cap stocks from various markets across the globe. This study also explores which hidden layer conguration leads to the best network predictive performance. Fur- thermore, this research identies which rm-specic factors predominantly in uence the predictions made by the articial neural networks. Five articial neural networks are designed, trained and tested on a sample of 161 stocks from the Russell 1000 and the S&P International 700 stock indices. The investigation period extends over a 166-month period from January 2001 to October 2014 with a 70:30 split for training and testing subsamples respectively. Eighteen rm-specic factors, based on prior research about the presence of style eects or anomalies on the cross-section of global equity returns, are used as the input variables of the articial neural networks to forecast one-month forward returns of all the stocks in the sample. The ve articial neural networks investigated in this research diered in hidden layer size. Specically, the number of hidden neurons examined were three, nine, 13, 18 and 30. All ve networks train signicantly well, with each network's training error indicating a good model t. Each network also achieves the desirable information coecient of 0.1 between its predicted returns and the actual returns in the training sample. It is interestingly discovered that network performance generally improves as the number of hidden neurons in the hidden layer increases until a specific point, after which network performance weakens. In the context of avoiding overtting, the best-trained network in this research is that with 13 neurons in its hidden layer. This is the primary network used for the out-of- sample testing analysis. This network achieves an average prediction error magnitude of approximately 7% and an information coecient of 0.05 during out-of-sample testing. These results underperform their respective benchmarks moderately. However, further analyses of the network's performance suggest an overall poor out-of-sample predictive ability. This is illustrated by a signicant bias and a considerably weak relationship between the network's predicted returns and the actual returns in the testing sample. Global sensitivity analysis reveals that growth style eects, particularly, the capital expenditure ratio, return on equity, sales growth, 12-month percentage change in non-current assets and six-month percentage change in asset turnover were the most persistent factors across all the ANN models. Other signicant factors include the 12-month percentage change in monthly volume traded, three-month cumulative prior return and one-month prior return. An unconventional result of this analysis is the relative insignicance of the size and value style effects.

show more