Prediction of Operating Cash Flows: Further Evidence from Australia

This paper examines empirically the relative abilities of current operating cash flows (hereafter OCF) and earnings in predicting future operating cash flows in Australia. It extends prior Australian research on cash flow prediction (Percy and Stokes 1992; Clinch, Sidhu and Sing 2002; Farshadfar, Ng and Brimble 2009) by examining future cash flow predictions for one-, two- and three-year-ahead forecast horizons; incorporating additional contextual variables likely to affect the predictive association between current cash flows or earnings and future cash flows; and comparing cross-sectional versus time series-based prediction models to ascertain the relative superiority of one approach over the other. Regression results reveal that the cash flow-based models are more accurate in predicting future operating cash flows than earnings-based models. This result, however, is moderated by firm-specific contextual factors like firm size, negative versus positive cash flow pattern, cash flow variability and firm operating cycle. Finally, a comparison between cross-sectional and time series approaches reveals that the cross-sectional model outperforms the time series model for both the operating cash flows and earnings models in most of the forecast years.