An explorative analysis of sentiment impact on S&P 500 components returns, volatility and downside risk

The main contribution of this study is to assess whether investor sentiment, as measured through textual analysis of newspaper articles or social media posts, does have an effect on the mean returns and on the variance of financial stocks. The analysis is carried on a basket of the S &P 500 components where stock returns and volatility are modeled within the GARCH family augmented, both in the mean and the variance equation, with an exogenous variable representing the investor sentiment; the latter is measured through specific Bloomberg proprietary scores based on News or Twitter feeds. Empirical results support the hypothesis that these indicators do have a positive impact on stock prices: the Twitter based index positively affects the components returns, confirming the outcomes of existing studies, whereas the news based index has a significant impact on their volatility. We also contribute the literature by performing the same analysis across the 11 sectors of the index, evidencing that investor sentiment has a significant impact on Industrials, Health Care, Consumer Discretionary, Consumer Staples, Information Technology and Communication Services. As a further contribution, we perform an out-of-sample analysis to assess the potential effect of Bloomberg sentiment scores on downside risk measures, such as Value at Risk and Expected Shortfall. This subject is relevant to regulators in order to conceive suitable policy interventions during turmoil periods around specific market sectors or stocks.