An Empirical Investigation of Portfolios with Little Idiosyncratic Risk

摘要: The objective of this study is to answer the following research question: How large a diversified portfolio? Although previous work abundant, very little progress has been made in answering question since seminal Evans and Archer (1968). This proposes two approaches address question. first approach measure rate risk reduction as diversification increases. For approach, I identify kinds risks: (1) that portfolio returns vary across time (Evans (1968), Campbell et al. (2001)); (2) portfolios same size (Elton Gruber (1977), O'Neil (1997)). show times series reaches an asymptote Cross sectional risk, on other hand, does not appears reach second consists comparing portfolios' performance benchmark assumed be (Statman (1987)). develop index. index calculated, for any given test portfolio, ratio Sharpe-like diversified. based intuition increase reduces cross increases transaction costs. Portfolio worth increasing long marginal from decrease greater than Diversification attained when value one. results my simulations indicate well at least 30. number can substantially higher if, example, investment horizon length, and/or cost investing are changed. active strategy considered study, which optimizing randomly selected portfolios, seem produce smaller portfolios. result supports market efficiency hypothesis.