By isolating stock-specific momentum from earnings announcements, the study shows it predicts returns across the US, Europe, and Japan. This approach avoids factor momentum bias, supports underreaction theories, and revives interest in the post-earnings drift.