Portfolio Concentration


Speaker


Abstract

The information available to the investor to build up his portfolio can be of financial as well as non-financial nature. Past performance, volatility and price to earning ration, among other financial data are readily estimated and widely available. Non-financial information like geographic location, industry sector, ownership structure or carbon emission levels are also available and constitute some ex-ante modulators in the investment decision making process. However, two different sample portfolios with different stocks can have the same expected return and the same standard deviation.

In this research we introduce an additional descriptor class for investment portfolios that is estimation error free, namely the concentration and the specialization of an investment portfolio. This information provides the investor with a tool for monitoring and managing his portfolio, as well as an additional insight on the composition of the same. We define a concentrated and a specialized portfolio and we suggest the Hirschman-Herfindahl index and the Gini index as practical measures for concentration and specialization. We create two quarterly portfolios with the top listed 500 stocks of the US market, one equally weighted and the other weighted by the market capitalization of each stock over 21 consecutive years and we describe the variation of the concentration and specialization of each quarterly portfolio over the years.The results obtained show the relevance of the descriptors we introduce in understanding the nature and the constitution of each portfolio with respect to its concentration and specialization.

We conclude with a conjecture on the absence of a neutral portfolio when there are no constraints on the attributes of the stocks constituting, and hence on the importance of the new descriptors in monitoring and managing an investment portfolio.