Rare Events Analysis for High-Frequency Equity Data

Abstract

In this work we present a methodology to detect rare events which are defined as large price movement relative to the volume traded. We analyze the behavior of equity after the detection of these rare events. We provide methods to calibrate trading rules based on the detection of these events and we exemplify for a particular trading rule. We apply the methodology to tick data for thousands of equities over a period of five days. In order to draw comprehensive conclusions we group the equities into classes and we calculate probabilities of price recovery after these rare events and for each class. The methodology that we have developed is based on non-parametric statistics and makes no assumption about the distribution of the random variables in the study. ______________________________________________________________________________________

Financial Econometrics and Empirical Market Microstructure

“…Engle and Russell (1998) use the Autoregressive Conditional Duration (ACD) model which considers the time between trades as a variable related to both price and volume. Bozdog et al (2011) study the exception of the conclusion presented in the earlier literature, they do not consider models in time but rather make the change in price dependent on the volume directly. Authors present a methodology of detecting and evaluating unusual price movements defined as large change in price corresponding to small volume of trades.…”

Tranched Value Securities

“…See Michael Simkovic (2009). 12 See Dragos Bozdog, Ionut Florescu, Khaldoun Khashanah, and Jim Wang (October 2018). 13 See John Reeves (May 2009).…”

A Slightly Depressing Jump Model: Intraday Volatility Pattern Simulation

“…Lee and Mykland (2008), instead of focusing on realised volatility, propose to observe the magnitude of individual returns relative to a measure of local volatility so as to establish the significance of jump dynamics. In the context of high frequency jump detection, jumps are referred to also as rare events wherein relatively sharp movements in price occur with light volume as in Bozdog et al (2015). Recently, Buckle et al (2017) suggest a new method of running jumps (BCH method) particularly to detect flash-crash type of jumps, utilising an approach that includes a local volatility measure.…”