Local bias of individual investors at Borsa İstanbul Equity Market: Tendency to invest in geographically close firms

Although the advantages of diversification are emphasized by traditional finance theories, individual investors often tend to hold only a few domestic stocks in their portfolios. This phenomenon has been subject to research since the last two decades under the title of home bias. The preference for investing close to home also applies to portfolios of domestic stocks. In other words, asymmetric information between local and nonlocal investors may drive the preference for geographically proximate investments and this phenomenon is called local bias in the terminology of behavioral finance. The dissertation provides new evidence concerning local bias of individual investors in an emerging country's stock market namely, Borsa Istanbul (BIST). More specifically, this study investigates whether individual investors weigh shares of firms, which are geographically close to themselves, more than expected in their portfolios. The results of the analyses indicate that Turkish individual investors are prone to local bias on the average, which is the main finding of the dissertation. This result is robust to alternative calculation methods of local bias, including both city/region-based and distance-based definitions. Further, it is shown that local investments do not generate extra returns with respect to that of non-local investments. Therefore we find evidence that the motivation of local investment is not information advantage. We support this finding further by comparing the returns of local investments between different classifications of the companies' shares. We classify the shares based on whether they are blue-chip or not, have higher corporate governance score, have higher free float rate, and have larger institutional ownership base, all of which imply the difference in the level of information asymmetry. If the reason of local bias is information advantage, we may expect that such advantage is more obvious in shares which have high level of information asymmetry. However, we find that the proxies of high level of information asymmetry do not lead to higher returns for local investments. Since our findings imply that the reason of local bias is behavioral, it may be meaningful to check several individual characteristics of investors which are used in other behavioral studies. We find that gender affects local bias, female investors have lower level of local bias. On the other hand, as the investors get older, their tendency to invest in local stock decreases, which may be explained by increasing level of their financial knowledge and experience. As investors get more sophisticated in tme, they may trade on reliable information more. Portfolio size, which may be a proxy for wealth, affects local bias positively. Last, we check the effects of city-related factors on local bias. We find that local bias is lowe in more populous and more developed cities. Development level of the cities is not found effective on local bias. Last, distance to the economic capital of the country, İstanbul, has negative effect on local bias.