This paper investigates the home bias puzzle, which is the tendency of investors to prefer domestic assets over foreign ones despite the potential benefits of diversification. The authors argue that home bias arises not because of restrictions on capital flows, but because of information asymmetry. Home investors have an initial advantage in learning about domestic asset payoffs, and this advantage is not eliminated by global information access. Instead, home investors choose to specialize in what they already know, which amplifies their information advantage and leads to home bias. The paper presents a two-country, rational expectations, general equilibrium model where investors first choose what information to learn before investing. Home investors have a slight advantage in learning about domestic asset payoffs, and this advantage is magnified by their choice to specialize in what they already know. This leads to higher expected returns for home investors and reinforces home bias. The model matches patterns of local and industry bias, foreign investments, portfolio out-performance, and asset prices. The authors propose new avenues for empirical research and show that learning can magnify home bias considerably. When all home investors have a small initial advantage in all home assets, the home bias is between 5 and 46%. When each home investor has an initial advantage in one local asset, the home bias can rise as high as 76%, which is consistent with U.S. portfolio data. The paper also discusses the implications of information asymmetry for other home bias explanations, such as ambiguity aversion. It shows that information asymmetry can amplify home bias even when investors have access to global information. The authors conclude that the home bias puzzle is best explained by information asymmetry, and that learning about what one already knows can lead to higher expected returns and reinforce home bias.This paper investigates the home bias puzzle, which is the tendency of investors to prefer domestic assets over foreign ones despite the potential benefits of diversification. The authors argue that home bias arises not because of restrictions on capital flows, but because of information asymmetry. Home investors have an initial advantage in learning about domestic asset payoffs, and this advantage is not eliminated by global information access. Instead, home investors choose to specialize in what they already know, which amplifies their information advantage and leads to home bias. The paper presents a two-country, rational expectations, general equilibrium model where investors first choose what information to learn before investing. Home investors have a slight advantage in learning about domestic asset payoffs, and this advantage is magnified by their choice to specialize in what they already know. This leads to higher expected returns for home investors and reinforces home bias. The model matches patterns of local and industry bias, foreign investments, portfolio out-performance, and asset prices. The authors propose new avenues for empirical research and show that learning can magnify home bias considerably. When all home investors have a small initial advantage in all home assets, the home bias is between 5 and 46%. When each home investor has an initial advantage in one local asset, the home bias can rise as high as 76%, which is consistent with U.S. portfolio data. The paper also discusses the implications of information asymmetry for other home bias explanations, such as ambiguity aversion. It shows that information asymmetry can amplify home bias even when investors have access to global information. The authors conclude that the home bias puzzle is best explained by information asymmetry, and that learning about what one already knows can lead to higher expected returns and reinforce home bias.