Modeling right-skewed financial data streams: A likelihood inference based on the generalized Birnbaum--Saunders mixture model

Mehrdad Naderi*, Farzane Hashemi, Andriette Bekker, Ahad Jamalizadeh

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

9 Citations (Scopus)

Abstract

Finite mixture models have recently been considered for analyzing positive support economical data streams with non-normal features. In this paper, a new mixture model based on the novel class of generalized Birnbaum–Saunders distributions is proposed to enhance strength and flexibility in modeling heterogeneous lifetime data. Some characteristics and properties of this mixture model are outlined. By presenting a convenient hierarchical representation, a mathematically elegant and computationally tractable EM-type algorithm is adopted for computing maximum likelihood estimates. Theoretical formulae of well-known risk measures referring to the class of generalized Birnbaum–Saunders distributions are derived. Finally, the utility of the postulated methodology is illustrated with some real-world data examples.
Original languageEnglish
Article number125109
Number of pages16
JournalApplied Mathematics and Computation
Volume376
Early online date25 Feb 2020
DOIs
Publication statusPublished - 1 Jul 2020
Externally publishedYes

Cite this