Sujet 6: Intraday risk measures

Conventional probability based risk measures, such as Expected Variance, Value at risk, Expected Shortfall operate under the assumption of fixed time between the potential outcomes, usually multiples of a day. However, with the raise of High Frequency and algorithmic trading, empirical evidence has shown that there are profitable intraday trading strategies because asset prices on this time framework deviate significantly from the randomness implied by the random walk model. One major reason is that events do not appear in fixed time intervals, such as 5’. Due to the irregularly spaced timing of events new types of risk emerge and need to be managed. They are linked to the fundamental question of waiting times:

a) How long do I need to wait for an order to be executed (execution risk)

b) How long am I exposed to a particular type of risk (time at risk)

c) How does the intensity of risk (VaR, ES, etc.) changes with time (time variant risks)

Potential research topics could evolve around the identification of types of intraday risks, the measurements of existing (or newly identified) risk measures, potential strategies for intraday risk management, etc.

Pre-requisites: a) Fluency in English; b) Solid knowledge of econometrics and statistics.


Return >>  Offres de thèses


Coordinateurs scientifiques du projet :

François LANGOT (francois.langot @

Directeur du GAINS (Groupe d'Analyse des Itinéraires et Niveaux Salariaux)

Laurent DENIS (laurent.denis @

LMM (Laboratoire Manceau de Mathématiques)

Mélanie COUELLIER (panorisk-ecodroit @

Chargée de gestion administrative et d'aide au pilotage

Tel. (33) (0)2 43 83 31 11