Does the real economy still forecast technology stock returns in the United States listed companies? Evidence in an age of digital economy and crises

Abstract

This thesis investigates whether the real economy continues to provide predictive signals for technology stock returns in U.S. companies during the digital era and economic crises. The study updates and extends previous research by focusing on modern economic indicators and their forecasting power amidst the evolving market landscape influenced by technological advancements and recent economic disruptions. The research explores the viability of historical economic indicators in forecasting stock returns, particularly in the U.S. technology sector. It questions whether these indicators, which proved useful in past decades, still hold predictive power in a radically transformed economic and technological environment. The thesis adopts a mixed-methods approach, combining quantitative analyses with econometric modelling. The data spans several decades, focusing on periods marked by significant economic events, such as the dot-com bubble, the 2008 financial crisis, and the COVID-19 pandemic, to assess the robustness of forecasting models under different economic conditions. The study replicates and extends previous methodologies using RStudio, allowing for a comparison between past and current forecasting abilities of various economic indicators under different economic conditions. Leveraging Google Colab and Python, the thesis incorporates machine learning techniques to examine the predictive power of traditional and newly proposed economic indicators. This analysis aims to uncover complex nonlinear relationships that might be missed by conventional econometric models. The findings suggest that the traditional indicators have diminished in predictive power. The discussion delves into the implications of the findings for investors and policymakers, emphasising the need for adaptive strategies that account for the rapid technological changes and their impact on the economic landscape. The thesis concludes that the real economy continues to provide valuable insights into technology stock returns, even if the predictive power has diminished. It calls for ongoing research to refine these indicators and adapt forecasting models to the changing economic and technological environment. Suggestions for future research include exploring additional digital economy indicators and extending the analysis to global technology markets to validate the findings and enhance the generalizability of the forecasting models. This study contributes to the literature by updating forecasting models with contemporary economic indicators and by demonstrating the evolving relationship between the real economy and technology stock returns in the face of digital transformation and economic crises.