Efeito índice: determinantes da ocorrência no BOVA11 e IGCX

Abstract

In this research, evidence of the occurrence of the index effect for the periods between 2005 and 2015 were analyzed in relation to the hypotheses present in literature, which explain such anomaly in the financial market. The objective of the research was to present statistical evidence that sustain the occurrence of the effect in Ibovespa to, later, trace a set of characteristics that suit the companies involved in the movements of inclusion and exclusion of the theoretical portfolios of the index. The issue questions were: Which hypotheses would be more adequate to explain the Index Effect in the Brazilian financial market? Which is the patrimonial structure of the included and excluded companies considering the financial and market indicators? The theoretical approach was based on finance fundaments over portfolio composition and price change of the actives in the market, as well as on the available literature on the different observation reports of the index effect in various markets. Different publications of the BM&FBOVESPA were consulted, concerning the elaboration, disclosure and standardization of the different indexes it provides the market. Data collection used quotation series, patrimonial and financial data and banks of news transmitted by BM&FBOVESPA, during the analyses period. After elaborating an event study, a few of the methodologies were employed with the objective of providing statistically proven conclusions regarding the compiled data. The intervention analysis presented significant oscillations in the quotation series. The abnormal returns were calculated based on different market returns for differentiated performance evaluation between included and excluded companies and with the creation of BOVA11.The results were evaluated by means of the hypothesis test for a more thorough comparison of means. A higher amount of event with coefficient presenting expected signal were observed, between the periods of precedent analysis and portfolio changes, concerning the events with signal inverted from the expected. Non-persistent abnormal returns were verified when observing the hypotheses tests in eighty-day negotiation windows. Both results suggest an explanation by the Price-Pressure Hypothesis. It was possible to estimate a logistic regression model presenting the positive (Net Profit, Market Value Variation and Negotiability) and negative (Size) coefficient characteristics from the explanation of the inclusion and exclusion movements of the index. This characterization of the companies suggest an informational aspect in the disclosure of the compositions. Despite not intended by BM&FBOVESPA, this is consistent with the Informational Content Hypothesis. With the presentation of the patrimonial and market characteristics of the included companies, the possibility of other factors influencing the indexation process can be inferred, notwithstanding the index presenting transparent criteria for listing. A few r esults obtained from the companies were shown to influence movements of input or no of the theoretical portfolio of the index, such as higher profitability, higher market value and a relatively smaller size.