Recognition-Based and Familiarity-Based PortfolioStrategies – An Experimental Study

Empirical evidences show that investors tend to be biased toward investing indomestic (home bias) and local (local bias) stocks. Familiarity is considered to be one of thereasons. A similar concept was proposed by Goldstein and Gigerenzer (1999, 2002), known asthe recognition heuristic: “when choosing between two objects, of which only one is recognized,choose the recognized. Investors recognize or are familiar with local stocks, expect them toprovide higher returns and, therefore, invest more in such stocks”. We conducted an experimentin Jena, Germany to test whether subjects show local bias and use recognition-based andfamiliarity-based portfolio strategies. We categorized them into an experienced and aninexperienced group; in addition, we used two data periods, i.e., bull market and bear market, tosee if they behave differently in the two markets. Results show that all participants invested moreof their endowments in the stock market in bull rather than bear market. All participants showedgreater familiarity with local stocks. However, the experienced participants only invested morein local rather than recognized and familiar stocks; on the contrary, the inexperiencedparticipants invested more in recognized and familiar but not local stocks. Our experiment showsno evidence that familiarity is a reason for local bias.