Abstract

The purpose of this paper is to identify elements of intentional herd behavior (HB), differentiating it from spurious, or unintentional HB. Using a panel of 50 stocks belonging to 18 Chilean equity mutual funds between December 2002 and October 2009, with manually collected data regarding physical positions of monthly purchases and sales, the authors calculate the level of HB and, by applying panel regressions with fixed and random effects, analyze the factors that determine this behavior, classifying them as agency, information, efficiency and behavioral problems. The research establishes that among Chilean equity mutual funds, there is a herding of 2.8 percent, implying that for 100 funds trading a certain stock, 53 go in the same direction and 47 in another. This effect increases during widespread market dips and when stocks become fashionable, attracting market attention. This behavior is not merely spurious, associated with variables that predict returns, but also has an intentional component, related to agency problems and information, and a behavioral component, related to investors’ biases and beliefs. The paper is original because, despite existing evidence of herding in international markets, it has been little quantified or studied in emerging markets. In addition, the literature does not distinguish between spurious and intentional HB, nor does it test different hypotheses jointly to explain the phenomenon.