Anmerkungen:
Nach Informationen von SSRN wurde die ursprüngliche Fassung des Dokuments March 23, 2023 erstellt
Beschreibung:
Corporate disclosures increasingly utilize multimedia to assist investors in better comprehending information. This study investigates how nonverbal disclosures, specifically the quality of vocal delivery, influence real-time investment decisions. We measure the quality of vocal delivery using a deep learning algorithm that analyzes the pronunciation, fluency, and diction of speakers to capture the level of cognitive burden imposed on listeners. Using a large sample of earnings conference call audio files, we show that real-time stock market reactions are subdued when vocal delivery is of low quality and that this effect is especially salient in retail trades. We also find that vocal delivery is of lower quality when firms report a decrease in earnings, have high earnings volatility, and present a negative-sentiment transcript. Additionally, firms with lower-quality vocal delivery tend to have less persistent positive earnings. Our research provides evidence that nonverbal disclosure is associated with information processing costs for investors, which could motivate managers to adjust their nonverbal disclosures