The Return Power of Analyst Reports: Definition, Factor and Strategy

Analyst reports are crucial in financial markets, providing information that guides investment decisions. However, existing research on their profitability and impact, especially over different time horizons, remains inconclusive. This study introduces a novel metric, ‘return power’, to assess the performance of analyst reports in both short‐ and long‐term contexts. We find that analyst reports serve as the short‐term ‘news’ and the long‐term ‘information’, each influencing stock returns differently. Using a dataset of 249,576 analyst reports from the Chinese stock market (2007–2022), we empirically examine key factors—such as broker's status, broker's prior performance, and investors' or analysts' attention—that affect return power at different time horizons. Based on these findings, we design investment strategies utilising both regression models and eight other well‐known machine learning techniques. Our results show that the regression‐based strategy outperforms other machine learning models and a market index, consistently providing superior risk‐adjusted returns across various transaction costs. This study confirms the profitability of analyst reports and offers a practical framework for report‐driven investment strategies.