Cash flow forecast for South African firms

Abstract

This paper applies models in the extant literature that have been used to forecast operating cash flows to predict the cash flows of South African firms listed on the Johannesburg Stock Exchange. Out-of-sample performance is examined for each model and compared between them. The reported results show that some accrual terms, i.e. depreciation and changes in inventory do not enhance cash flow prediction for the average South African firm in contrast to the reported results of studies in USA and Australia. Inclusion of more explanatory variables does not necessarily improve the models, according to the out-of-sample results. The paper proposes the application of moving average model in panel data, and vector regressive model for multi-period-ahead prediction of cash flows for South Africa firms.