|
Activity Number:
|
550
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 3, 2016 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Business and Economic Statistics Section
|
|
Abstract #319123
|
View Presentation
|
|
Title:
|
What Is the Best Model to Predict Real Failure of Hedge Funds?
|
|
Author(s):
|
Jose Faias* and Leila Amorim
|
|
Companies:
|
Catolica Lisbon SBE and UFBA
|
|
Keywords:
|
Survival model ;
prediction ;
Cox ;
Parametric ;
hedge funds
|
|
Abstract:
|
The real failure of hedge funds has increased substantially and it has achieved a rate of 8% at the end of 2012. Past literature has addressed failure of hedge funds using ad hoc models ignoring which one fits the data best. We run a comprehensive evaluation of parametric and semi-parametric models (over 30,000) for a sample constituted by more than 3,000 funds between 1995 and 2013 to study real failure. A model incorporating time-varying covariates and piecewise effects in the risk measure allows a better description of the impact of hedge fund's characteristics on their out-of-sample prediction to failure.
|
Authors who are presenting talks have a * after their name.
