Intraday Value-at-Risk Estimation for Directional Change Events and Investment Strategies

Rui Jorge Almeida*, Nalan Bastürk, Robert Golan

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingAcademicpeer-review

Abstract

Accurate risk measurement is important for making and assessing investment decisions. Recently, directional change representations of returns are proposed as a new method for analyzing and forecasting intra-day price movements and for creating investment strategies. This paper presents an FGARCH model for intraday Value-at-Risk (IVaR) estimation, for assessing risk properties in time series represented as directional change events, and for predicting the market risk for investment strategies based on directional changes. We apply the proposed method to 5-minute intraday data and report the time-varying risk based on IVaR estimates. We study the accuracy of these estimates and report the robustness of the risk estimates to the choice of the threshold parameter for directional change representations. Furthermore, we apply the proposed methodology to compare the risk properties of two investment strategies based on directional changes with a baseline moving window investment strategy. For these data, we find that the directional change strategies lead to higher returns but no substantial increase in risk compared to the baseline strategy. The proposed methodology is applicable to other intra-day data frequencies and it is generalizable to other risk metrics.
Original languageEnglish
Title of host publication2017 IEEE SYMPOSIUM SERIES ON COMPUTATIONAL INTELLIGENCE (SSCI)
PublisherIEEE
Pages588-595
Number of pages8
ISBN (Print)9781538627266
DOIs
Publication statusPublished - 2017
EventIEEE Symposium Series on Computational Intelligence : IEEE SSCI - Honolulu, United States
Duration: 27 Nov 20171 Dec 2017
https://ewh.ieee.org/conf/ssci/2017/#:~:text=The%202017%20IEEE%20Symposium%20Series,the%20IEEE%20Computational%20Intelligence%20Society.

Symposium

SymposiumIEEE Symposium Series on Computational Intelligence
Country/TerritoryUnited States
CityHonolulu
Period27/11/171/12/17
Internet address

Cite this