Use this URL to cite or link to this record in EThOS:
Title: Bayesian methods in analysis of fund management performance
Author: Fan, Yun
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 2008
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
This thesis develops a model of fund management performance that incorporates market behaviour, beliefs, opinions and economic index into measurement. Given that some of the measure will be subjective, it is inappropriate to use Bayesian methodology. Traditionally modelling of fund management performance has been based on the Capital Assets Pricing Model (CAPM), which has been revised and modified over time. The simple CAPM can be described as a linear regression model of the return of the fund against a benchmark. It can be extended to include other factors to aid modelling. A review of the literature, as well as to interviews and a survey, reveals 12 macro-economic variables as significant in predicting fund performance. Two estimation methods, Ordinary Least Square (OLS) and Bayesian Modelling (BM) were employed on a monthly data for 26 equity funds from the USA market over 15 year period. After several processes of filtering, five factors include Standard & Poor 500 and four other macro-economic factors: US federal funds rate, US federal funds rate target, US monetary base and US money supply 1 were left. These key variables were then used in subsequent modelling of fund performance. A range of models were considered for the modelling including: dependent and independent models. Results from different models are consistent. Five factors model consistently score a quite high adjusted R square which proves good tracking ability of the model on fund performance. Overall, funds do not have a superior performance compared to the benchmark and do have similar risk preference to the market portfolio. Intermarket effect has been investigated as well in the study and it is shown empirically that no such effects exist. Empirical Bayesian models are explored using Bootstrap Re-sampling method. The results obtained are similar previously.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available