12 October – from 1:15pm to 2:15pm
How attentional are retail investors to medium-sized stock returns? A Machine learning approach based on Internet search queries
Sarra Ghaddab, PhD student at University Lyon 1, ISFA research center
This paper fills the gap in the investor attention theory and considers mid-cap companies, often overlooked in the literature, to predict their stock market returns using the Google search volume index (GSVI). Indeed, mid-cap markets are less efficient than large-cap markets, and inefficiency may be predicted using the GSVI. To conduct this study, we work with 8-minute frequency data and join the rare studies working with intraday data in this financial application. Within this framework, a Generalized Autoregressive Conditional Heteroskedastic model with a Generalized Error Distribution (GED-GARCH) is used as a benchmark model to predict stock returns. The proposed model is then compared to Machine Learning models. The results show that the returns of mid-sized companies are indeed influenced by the GSVI. Therefore, the study of investor attention should not be limited to large market capitalization companies and increased attention should be paid to smaller companies.
Registration, please contact email@example.com
Room 235 Ecully campus