A carregar...
Publicação
Sobre a avaliação da performance de fundos de investimento
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.2 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
A avaliação da performance dos gestores de carteiras tem sido uma questão
amplamente debatida na literatura financeira. O desenvolvimento da Moderna Teoria do
Mercado de Capitais proporcionou o surgimento, nos anos 60, de medidas de avaliação
ajustadas ao risco: as de Jensen, Treynor e Sharpe, também denominadas medidas
tradicionais de avaliação da performance. As críticas de que têm sido alvo têm a ver, por
um lado, com problemas conceptuais e econométricos que lhes estão associados, e por
outro, com a impossibilidade destas medidas detectarem as componentes timing e
selectividade que contribuem para a performance global.
Neste contexto, e após a revisão da literatura, na qual se identificam, comparam e
criticamente se discutem as principais escolas de pensamento na área da avaliação da
performance de carteiras, procede-se a uma análise empírica, na base de uma amostra de
fundos de investimento portugueses, no sentido de (1) se estimarem as medidas
tradicionais de avaliação da performance, tendo em consideração vários cenários
(diferentes horizontes temporais, diferentes índices de mercado e a possibilidade de
heteroscedasticidade) e assim chamar a atenção para algumas questões ao nível da sua
aplicação e (2) se aplicar, dadas as limitações das abordagens tradicionais, o modelo
proposto por PFLEEDERER e BHATTACHARYA [1983] de forma a obter medidas
individuais de timing e selectividade.
Os resultados da análise sugerem que os fundos não evidenciam capacidades quer
ao nível da selectividade quer de timing. Algumas possíveis explicações para estes
resultados são adiantadas, nomeadamente ao nível da utilização dos "benchmarks",
existência de elevados custos de transacção, restrições de ordem legal e, eventualmente
incapacidade de previsão por parte dos próprios gestores. No entanto, e em face de
recentes desenvolvimentos nesta área, sugere-se também que outras razões (inclusivé, a
nível teórico), poderão ajudar a compreender estes mesmos resultados, por exemplo, a não
consideração da composição das carteiras e de modelos condicionados.
iv
Portfolio performance evaluatíon has received considerable attention in the past decades. Modem Capital Market Theory led to the development of risk-adjusted measures for ranking investment performance: the so called traditional measures of portfolio peformance of Jensen, Treynor and Sharpe. However, they have been subject to a continuous debate, mainly related either to conceptual and econometric problems, either to the impossibility of these measures in detecting timing and selectivity abilities of investment managers. In such a context, and after reviewing the literature, where the main schools of thoughts are identifíed, compared and critically evaluated, an empirical analysis is carried out, on the basis of a sample of Portuguese investment funds, in order to (1) estimate the above mentioned traditional measures, taking into account various scenarios (different time horizons, different market indexes and the possibility of heteroscedasticity), this way calling the attention to some relevam issues still being debated and (2) to apply, given the limitation of the traditional methodologies, the model suggested by PFLEIDERER and BHATTACHARYA [1983] in order to obtain separate measures of timing and selectivity. The results suggest that the funds do not show any ability in terms of either selectivity or timing. Some possible explanations for these results are advanced, namely in relation to the benchmark chosen, the existence of high transaction costs, constraints at the fiscal levei and, eventually, nonexistent forecasting ability at both leveis from the investment managers. However, in the light of very recent developments in this area, other kind of reasons (inclusive, at the theoretical levei) may help to understand such results, for example, the fact that the portfolio composition as well as conditional models are not being considered v
Portfolio performance evaluatíon has received considerable attention in the past decades. Modem Capital Market Theory led to the development of risk-adjusted measures for ranking investment performance: the so called traditional measures of portfolio peformance of Jensen, Treynor and Sharpe. However, they have been subject to a continuous debate, mainly related either to conceptual and econometric problems, either to the impossibility of these measures in detecting timing and selectivity abilities of investment managers. In such a context, and after reviewing the literature, where the main schools of thoughts are identifíed, compared and critically evaluated, an empirical analysis is carried out, on the basis of a sample of Portuguese investment funds, in order to (1) estimate the above mentioned traditional measures, taking into account various scenarios (different time horizons, different market indexes and the possibility of heteroscedasticity), this way calling the attention to some relevam issues still being debated and (2) to apply, given the limitation of the traditional methodologies, the model suggested by PFLEIDERER and BHATTACHARYA [1983] in order to obtain separate measures of timing and selectivity. The results suggest that the funds do not show any ability in terms of either selectivity or timing. Some possible explanations for these results are advanced, namely in relation to the benchmark chosen, the existence of high transaction costs, constraints at the fiscal levei and, eventually, nonexistent forecasting ability at both leveis from the investment managers. However, in the light of very recent developments in this area, other kind of reasons (inclusive, at the theoretical levei) may help to understand such results, for example, the fact that the portfolio composition as well as conditional models are not being considered v
Descrição
Instituto Superior de Economia e Gestão
Palavras-chave
Contexto Educativo
Citação
Cortez, Maria do Céu Ribeiro (1993). " Sobre a avaliação da performance de fundos de investimento". Dissertação de Mestrado, Universidade de Lisboa. Instituto Superior de Economia e Gestão.