Heterogeneous Stock Market Modelling and Market Risk Evaluation using High Frequency Information
Efficient markets hypothesis (EMH) has been a debatable topic among market practitioners and researchers since 1970s. The widespread availability of high frequency financial data together with a series of new empirical findings imposes challenges to the definition of financial markets and it also le...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2016
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Efficient markets hypothesis (EMH) has been a debatable topic among market practitioners and researchers since 1970s. The widespread availability of high frequency financial data together with a series of new empirical findings imposes challenges to the definition of financial markets and it also leads to the proposal of various alternative market hypotheses. Inspired by the EMH and the heterogeneous markets hypothesis (HMH), this study first examines the dynamic behaviour of global stock markets using daily and high frequency information. The identified predictable components of stock market dynamics are then captured using econometric models available in the literature with the aim to produce accurate risk forecasts. One of the main applications in conditional volatility modelling and forecasting of financial assets is the value-at-risk (VaR) estimation, which is used by financial institutions to report on the daily capital in risk. It remains as a question, whether realized volatility (RV) models that incorporate intraday data produce better VaR forecasts compared to methodologies which solely based on daily returns. Thus, this study provides extensive comparison of the out-of-sample volatility and the VaR forecast performance on global equity market indices: S&P500, FTSE100, and DAX30 by using 21 risk models which consist of 5 conventional GARCH specifications and 16 RV (ARFIMA and HAR) specifications. |
|---|