Author(s): Aldila Rizkiana
The development of information technology affects the investor's behavior in stock investment. Thus data in Indonesia Social Media, hold important information to predict the movement of the stock price. Unfortunately, the research about investor sentiment in social media repeatedly using the same methods that focus on the amount of stock return and mainly analyzes the stock market in the US. So, we will use logistic regression to model stock buy/ sell decision using investor sentiment as the predictor. We find that the logistic regression model we developed is not fit, and thus, investor sentiment in Social Media alone cannot be used as the predictor of stock buy/ sell decision. We argue there are five possible reasons why we did not find a significant relationship between investor sentiment and stock return. There are the speed of information diffusion on stock prices, data sources, market structure, market capitalization, and investor sentiment measurement methods.