Forecasting Revenue Surprises using Time Series Forecasting models in the context of an active investment strategy

Details

Supervisors

Vande Kerckhove, Corentin

Faculty

Louvain School of Management

Degree label

Master [120] : ingénieur de gestion, à finalité spécialisée

Abstract

In the context of active investing, many traders have attempted to forecast stock values using machine learning techniques. However, it has always been seen as a hard and challenging task. In this thesis, we aim to predict revenue surprises prior to quarterly revenue publications using time series forecasting models based on brokers' estimates and macro-economic indicators. The main objective of these forecasts is to capitalize on the direct impact of revenue surprises on stock prices through an active investment strategy. Studies cited and exploited in this research showcase the significance of forecasting surprises on companies’ fundamentals, rather than focusing directly on these fundamentals. Our results, however, do not conclusively establish our model's superiority over random investment decisions. In summary, the results of this thesis suggest that enhancing the quality of datasets used in predictive models, along with leveraging more powerful forecasting models, could improve the quality of our predictions.