BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Deep Hedging
DTSTART:20180308T120000
DTEND:20180308T130000
DTSTAMP:20260406T171940Z
UID:a543932b7127a292513ca55a8ecf18b9bec84fc3f0ad403d0df98102
CATEGORIES:Conferences - Seminars
DESCRIPTION:Lukas GONON (ETHZ)\nWe present a framework for hedging a portf
 olio of derivatives in the presence of market frictions such as transactio
 n costs\, market impact\, liquidity constraints or risk limits using moder
 n deep reinforcement machine learning methods.\nWe discuss how standard re
 inforcement learning methods can be applied to non-linear reward structure
 s\, i.e. in our case convex risk measures. As a general contribution to th
 e use of deep learning for stochastic processes\, we also show that the se
 t of constrained trading strategies used by our algorithm is large enough 
 to ϵ-approximate any optimal solution.\nOur algorithm can be implemented 
 efficiently even in high-dimensional situations using modern machine learn
 ing tools. Its structure does not depend on specific market dynamics\, and
  generalizes across hedging instruments including the use of liquid deriva
 tives. Its computational performance is largely invariant in the size of t
 he portfolio as it depends mainly on the number of hedging instruments ava
 ilable.\nWe illustrate our approach by showing the effect on hedging under
  transaction costs in a synthetic market driven by the Heston model\, wher
 e we outperform the standard "complete market" solution.
LOCATION:UNIL\, Extranef\, room 118 https://planete.unil.ch/plan/?local=EX
 T-118.1
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR