Patriotic Behaviour: Behavioural Finance and an Empirical Investigation of the Patriotic Bias in the U.S. Equity Market

Abstract

Identifying patriotism in the financial market can, to some extent, be difficult, because the concept is based on feelings rather than rationality. However, several scholars have already proofed the existence of patriotic bias in the equity market, which is not consistent with the efficient market hypothesis. Patriotic bias occurs because agents feel patriotic and want to support their home country. One of the possible ways, to support their country, is by investing in companies with patriotic names in them, thus referring to the agent’s home country. This kind of investor behaviour, of preferring patriotic equities, all else equal, is not rational and therefore challenges the traditional economic theories that are developed on investor rationality and perfect markets. In this study the patriotic bias is investigated in the US equity market around patriotic holidays. A surge of patriotism around those days in the US, together with previous findings by other scholars, contributes to the foundation of this thesis. In order to investigate whether there exists a patriotic bias or not around the patriotic holidays, two patriotic portfolios are constructed, one equally weighted and one value weighted. The portfolios consist of 97 American equities with the company names including either of the four names America(n), or US(A). Thirteen American patriotic holidays have been identified and a dummy variable has been constructed to capture these. Several regression analyses with the equally weighted and the value weighted portfolios, as the dependent variables have been performed to see if there is any significance in the variable of the patriotic holidays. The results did, however, not show any significance of increased returns in the patriotic portfolios around patriotic days. In fact, each estimated coefficient on the patriotic holidays dummy was slightly negative for both the equally weighted portfolio, and the value weighted portfolio in all the regressions, after controlling for economic market news, and other calendar anomalies such as pre-holiday effect, week day effect and month effect. These results indicate a return in the patriotic portfolios of approximately 0.026%-0.05% less, around patriotic holidays compared to any other “regular” days. The results proofed to be valid, as same evidence of negative estimated coefficients of the patriotic holidays dummy were found when making robustness checks of increased time periods. Thus, no patriotic bias was detected around the patriotic holidays.