A comparative goodness-of-fit analysis of distributions of some Lévy processes and Heston model to stock index returns

•We analyze goodness-of-fit of three Lévy processes and Heston model to index returns.•We identify normal and turbulent periods for twenty developed and emerging markets.•We observe Lévy processes perform better than Heston model in developed markets.•In most cases, VG and NIG distributions provide...

Full description

Saved in:
Bibliographic Details
Published in:The North American journal of economics and finance 2016-04, Vol.36, p.69-83
Main Authors: Göncü, Ahmet, Karahan, Mehmet Oğuz, Kuzubaş, Tolga Umut
Format: Article
Language:English
Subjects:
Emerging markets
Generalized hyperbolic model
Heston model
Markov analysis
Markov regime-switching model
Normal-inverse gaussian model
Probability distribution
Rates of return
Securities markets
Stochastic models
Stock exchanges
Studies
Variance-gamma model
Volatility
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•We analyze goodness-of-fit of three Lévy processes and Heston model to index returns.•We identify normal and turbulent periods for twenty developed and emerging markets.•We observe Lévy processes perform better than Heston model in developed markets.•In most cases, VG and NIG distributions provide better fit than GH distribution. In this paper, we investigate the goodness-of-fit of three Lévy processes, namely Variance-Gamma (VG), Normal-Inverse Gaussian (NIG) and Generalized Hyperbolic (GH) distributions, and probability distribution of the Heston model to index returns of twenty developed and emerging stock markets. Furthermore, we extend our analysis by applying a Markov regime switching model to identify normal and turbulent periods. Our findings indicate that the probability distribution of the Heston model performs well for emerging markets under full sample estimation and retains goodness of fit for high volatility periods, as it explicitly accounts for the volatility process. On the other hand, the distributions of the Lévy processes, especially the VG and NIG distributions, generally improves upon the fit of the Heston model, particularly for developed markets and low volatility periods. Furthermore, some distributions yield to significantly large test statistics for some countries, even though they fit well to other markets, which suggest that properties of the stock markets are crucial in identifying the best distribution representing empirical returns.
ISSN:1062-9408
1879-0860
DOI:10.1016/j.najef.2015.12.001