The Nonlinear Self-excited Threshold Two-component Mixture Multiplicative Error Model
The main purpose of the present paper is to introduce an innovative model for non-negative time series, especially for financial data, which incorporates the mixture multiplicative error models with a self-excited threshold effect. The mixture distributions are applied to the effective handling of the multimodal, asymmetric, and skewed features. The Expectation-Maximization estimation algorithm is developed to estimate the model parameters, including an unknown threshold parameter. Results from the Monte Carlo simulation indicate that the estimates of the unknown parameters are converging towards their true values. Finally, the proposed process is effectively applied to model the producer price index, with goodness-of-fit measures demonstrating its superiority over other relevant models, thereby confirming the validity of the proposed modeling.
Authors
- Mohammad Hassan Behzadi (ORCID: https://orcid.org/0000-0002-1849-5588)
- Ebrahim Bagherian Farahabadi (ORCID: https://orcid.org/0000-0001-5678-5706)
- Einolah Deiri
Institutions
- Islamic Azad University, Tehran (IR)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-18
- DOI
- https://doi.org/10.21203/rs.3.rs-9973966/v1
- Primary Topic
- Financial Risk and Volatility Modeling
- Type
- preprint