| Title |
Multifractality of high frequency data and volatility forecasting / |
| Translation of Title |
Aukšto dažnio duomenų multifraktalumas ir volatilumo prognozavimas. |
| Authors |
Valaitis, Agnius |
| Full Text |
|
| Pages |
46 |
| Keywords [eng] |
multi-fractality, volatility, high frequency data, multi-fractal model |
| Abstract [eng] |
The purpose of this thesis is to emphasize the importance of multi-fractal concept by providing an empirical evidence using intra-day financial time series. The Multifractal Detrended Fluctuation analysis and the Generalized Hurst exponent methods were applied on the price indices of Dow Jones Industrial Average, Australian Securities Exchange, Nikkei-225 and NASDAQ-100. The presence of multi-scaling was detected on prices of each mentioned index. In order to employ high frequency data we calculated realized volatility and applied the Binomial Markov-Switching Multifractal model. The power of prediction accuracy was compared to well-established Heterogeneous Auto-Regressive model. The Binomial Markov-Switching Multifractal model showed better performance on the short horizons while Heterogeneous Auto-Regressive model surpassed the latter on long horizon. |
| Dissertation Institution |
Vilniaus universitetas. |
| Type |
Master thesis |
| Language |
English |
| Publication date |
2018 |
