---

Influence of huge sets of information on financial markets has become almost instantaneous. Every new peace of information influences prices of assets and correlations among them. Many risk measures including value-at-risk or hedge ratios are based on variance-covariance forecasts. Moreover, future covariances are key risk measures themselves. Large portfolios of assets demand new methods of forecasting correlations that take into account constantly arriving high-frequency (HF) information. Recent developments in risk forecasting of both individual volatilities and large covariance matrices are the focus of this work. Theoretical part overviews latest modelling approaches to volatility forecasting, whereas in the empirical part selected volatility models are implemented and compared.

---

Boiko, Oleg
Oleg Boiko, BSc. in International Economics and Management at Vadym Hetman Kyiv National Economic University, MSc. in Quantitative Economics at the University of Konstanz, work as Junior Financial Engineer at Solvians GmbH. Fields of interest are financial engineering, programming and cryptocurrencies/blockchain.

---