Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/171866 
Year of Publication: 
2016
Citation: 
[Journal:] Econometrics [ISSN:] 2225-1146 [Volume:] 4 [Issue:] 1 [Publisher:] MDPI [Place:] Basel [Year:] 2016 [Pages:] 1-23
Publisher: 
MDPI, Basel
Abstract: 
We define a dynamic and self-adjusting mixture of Gaussian Graphical Models to cluster financial returns, and provide a new method for extraction of nonparametric estimates of dynamic alphas (excess return) and betas (to a choice set of explanatory factors) in a multivariate setting. This approach, as well as the outputs, has a dynamic, nonstationary and nonparametric form, which circumvents the problem of model risk and parametric assumptions that the Kalman filter and other widely used approaches rely on. The by-product of clusters, used for shrinkage and information borrowing, can be of use to determine relationships around specific events. This approach exhibits a smaller Root Mean Squared Error than traditionally used benchmarks in financial settings, which we illustrate through simulation. As an illustration, we use hedge fund index data, and find that our estimated alphas are, on average, 0.13% per month higher (1.6% per year) than alphas estimated through Ordinary Least Squares. The approach exhibits fast adaptation to abrupt changes in the parameters, as seen in our estimated alphas and betas, which exhibit high volatility, especially in periods which can be identified as times of stressful market events, a reflection of the dynamic positioning of hedge fund portfolio managers.
Subjects: 
nonparametric clustering
Bayesian
cluster
nonparametric alpha and beta
hedge fund performance
JEL: 
C2
C13
C14
Persistent Identifier of the first edition: 
Creative Commons License: 
cc-by Logo
Document Type: 
Article

Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.