Abstract
A model which simultaneously accounts for long memory and regime switching has been applied to the German stock returns. The parameters estimation assumes that the transition probabilities are known. We overcome this assumption by using the estimated transition probabilities of Markov switching model as ex-ante information. The results show that the simultaneous model outperforms Markov switching model, and give better realization of ex-post forecasting.
| Original language | English |
|---|---|
| Pages (from-to) | 7-17 |
| Number of pages | 11 |
| Journal | European Journal of Economics, Finance and Administrative Sciences |
| Issue number | 15 |
| Publication status | Published - Jan 2009 |
Keywords
- Long memory
- Persistence
- Regime switching
- Simultaneous model
Fingerprint
Dive into the research topics of 'Regime switching long memory model for German stock returns'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver