DYNAMIC STYLE ANALYSIS IN RECOVERY OF BRAZILIAN INVESTMENT FUNDS EXPOSURES: AN APPLICATION OF RESTRICTED KALMAN FILTERING / ANÁLISE DINÂMICA DE ESTILO NA RECUPERAÇÃO DAS EXPOSIÇÕES DE FUNDOS DE INVESTIMENTOS BRASILEIROS: UMA APLICAÇÃO DO FILTRO DE KALMAN RESTRITO
AUTOR(ES)
CAIO OLIVEIRA DE AZEVEDO
DATA DE PUBLICAÇÃO
2009
RESUMO
This dissertation aims to investigate and interpret the investment style of Brazilian investment funds - exchange funds and stock funds with active management, in the period ranging from January 2004 to August 2008, with the central objective of verifying if, in fact, these funds pursue the investment style promised to their customers. Accordingly, we used the methodology of returnbased dynamic style analysis, using the reduced restricted Kalman filtering with exact initialization applied to state space models. For this purpose, new indexes were created with the intention of covering the eventual lack of indexes that could appropriately represent the asset classes of financial market. The main conclusions were: (1) even in the midst of an adverse scenario that marked part of the analyzed period, the exchange funds kept the investment strategies announced to the public, demonstrating transparency in their actions; and (2) stock funds were indeed predominantly exposed to the stock market, but also allocated considerable part of their resources on federal public long-term bonds indexed by inflation indexes.
ASSUNTO(S)
linear restrictions filtro de kalman kalman filter analise dinamica de estilo restricoes lineares dynamic style analysis
ACESSO AO ARTIGO
Documentos Relacionados
- FILTRO DE KALMAN RESTRITO NA ANÁLISE DE ESTILO SEMI-FORTE DE FUNDOS ATUARIAIS
- RESTRICTED KALMAN FILTERING: THEORY, METHODS AND APPLICATIONS
- STATE SPACE MODELS WITH RESTRICTIONS IN COMPONENTS OF INTEREST: APPLICATIONS IN DYNAMIC STYLE ANALYSIS FOR BRAZILIAN INVESTMENT FUNDS
- INVESTMENTS POLICIES ANALYSIS OF THE BRAZILIAN PENSION FUNDS
- Brazilian Hedge Funds Dynamics: econometric Modelling with Time-Varying Parameters