A marked point process model for intraday financial returns: modeling extreme risk
- 35 Downloads
Forecasting the risk of extreme losses is an important issue in the management of financial risk and has attracted a great deal of research attention. However, little attention has been paid to extreme losses in a higher frequency intraday setting. This paper proposes a novel marked point process model to capture extreme risk in intraday returns, taking into account a range of trading activity and liquidity measures. A novel approach is proposed for defining the threshold upon which extreme events are identified taking into account the diurnal patterns in intraday trading activity. It is found that models including covariates, mainly relating to trading intensity and spreads offer the best in-sample fit, and prediction of extreme risk, in particular at higher quantiles.
KeywordsHawkes process Peaks over threshold Bid-ask spread Extreme risk High frequency
JEL ClassificationC32 C53 C58
Herrera acknowledges the Chilean CONICYT funding agency for financial support (FONDECYT 1180672) for this project.