Explaining bank stock performance with crisis sentiment

Abstract

a b s t r a c tUsing search volume data on crisis-related queries from Google Trends, we estimate three different measures of market-level and individual crisis sentiment. We find that the stock performance of international banks during the period Q1 2004 to Q4 2012 was significantly driven by investors' irrational market-wide crisis sentiment. Our empirical analysis shows that irrational market-wide crisis sentiment leads investors to devalue bank stocks irrespective of idiosyncratic or macroeconomic fundamentals. Comparing this finding with results for a sample of non-financial companies, we find evidence in support of the notion that the effect of crisis sentiment on stock returns is strongest in the absence of implicit bailout guarantees.

Crisis Sentiment in the U.S. Insurance Sector

“…Similar to the General CSI, we find that FEARS as a proxy of household‐level negative sentiment (not necessarily tied to this specific crisis) is strongly negatively correlated with weekly stock returns, which is in line with previous findings (see Da, Engelberg, and Gao, ; Irresberger, Mühlnickel, and Weiß ). Interestingly, FEARS is negatively correlated with insurers’ stock return volatility.…”

“…We then consider the estimate of to be our primary proxy for the general crisis sentiment. We follow Irresberger, Mühlnickel, and Weiß () and define as the General CSI…”

“…To avoid a possible look‐ahead bias in the estimation of the correlations, we estimate using rolling windows of length 52 weeks using data up to week t (the rolling windows are skipped ahead 1 week for each estimation). Finally, we construct the Crisis Sentiment Index (CSI) by combining the dynamic correlation between the first principal component and a firm's , multiplying the estimated correlation with the sum and then dividing the resulting term by 200 (see Irresberger, Mühlnickel, and Weiß (): …”

The role of internet in stock market

“…Following the intuition that investors react to uncertainty by intensifying their search behavior, Dzielinski (2012) expounds the search frequency from Google Trends as the economic uncertainty and find this indicator is significantly correlated with stock return and volatility. Irresberger et al (2015) construct a market-level and individual crisis sentiment with the search volume from Google Trends and find this sentiment can lead investor to devalue the bank stocks. Da et al (2015) also construct the investor sentiment by Google Trends and find this FEARS (Financial and Economic Attitudes Revealed by Search) index can predict temporary increases in volatility, mutual fund flow and short-term price reversals.…”

“…Recent studies have interpreted the search frequency of Google Trends as some reflections of individual investors' sentiment (Joseph et al, 2011;Dzielinski, 2012;Da et al, 2015 andIrresberger et al, 2015). The rationale behind this interpretation can be traced back to the literature in economic psychology consistently support that the agents will react to their inner state of mood with increasing searching behavior (Stigler, 1961;Rothschild, 1974 andLoibl andHira, 2009).…”