Overconfidence and Market Efficiency with Heterogeneous Agents

We study financial markets in which both rational and overconfident agents coexist and make endogenous information acquisition decisions. We demonstrate the following irrelevance result when a positive fraction of rational agents (endogenously) decides to become informed in equilibrium, prices are set as if all investors were rational, and as a consequence the overconfidence bias does not affect informational efficiency, price volatility, rational traders’ expected profits or their welfare. Intuitively, as overconfidence goes up, so does price informativeness, which makes rational agents cut their information acquisition activities, effectively undoing the standard effect of more aggressive trading by the overconfident. The main intuition of the paper, if not the irrelevance result, is shown to be robust to different model specifications.We would like to thank Alberto Bisin, Xavier Freixas, Ken French, Moshe Kim, Jose Marín, and Terrance Odean for comments on an early draft, as well as an anonymous referee and seminar participants at HEC Geneva, the 2004 EFA meetings, the 2004 European Econometric Society meetings and the 2005 SAET conference. Diego García and Branko Urošević gratefully acknowledge financial support by SECCF (Belgrade).     

Decomposing World Income Distribution: Does The World Have A Middle Class?

Using the national income/expenditure distribution data from 111 countries, we decompose total inequality between the individuals in the world, by continents and regions. We use Yitzhaki's Gini decomposition which allows for an exact breakdown of the Gini. We find that Asia is the most heterogeneous continent; between-country inequality is much more important than inequality in incomes within countries. At the other extreme is Latin America where differences between the countries are small, but inequalities within the countries are large. Western Europe/North America is fairly homogeneous both in terms of countries' mean incomes and income differences between individuals. If we divide the world population into three groups: the rich (those with incomes greater than Italy's mean income), the poor (those with incomes less than Western countries' poverty line), and the middle class, we find that there are only 11 percent of people who are "world middle class"; 78 percent are poor, and 11 percent are rich.     

THE GINI-TYPE FUNCTIONS: AN ALTERNATIVE DERIVATION

The purpose of this note is to propose an alternative and intuitively simpler derivation of the Gini coefficient (in Section l), to show how it can be generalized and how then a number of coefficients (concentration coefficient, Kakwani's progressivity index) are obtained directly from this generalization (Section 2), and finally to use this approach to obtain some Gini relationships (Section 3).