Revenue forecasting for European capital market-oriented firms : A comparative prediction study between financial analysts and machine learning models

This study uses publicly available information for European firms and recent machine learning algorithms to predict future revenues in an IFRS context, examining the benefits of predictive analytics for both preparers and users of these financial projections. For this purpose, the study evaluates the prediction quality of the forecasting models applied and compares them with each other and with the prediction quality of sell-side financial analysts’ forecasts. Our empirical results, based on 3,000 firm-year observations from 2010 to 2019, demonstrate that machine learning provides comparably accurate or even more accurate revenue forecasts than financial analysts. Therefore, the study highlights the considerable potential of machine learning and predictive analytics for improving the forecasting process in general and, in particular, to increase the accuracy, transparency, and objectivity of the forecasts. Since the latter also reduce information asymmetry between firms and investors, machine learning and predictive analytics contribute to capital market efficiency