Show simple item record

Distinguish the indistinguishable: a Deep Reinforcement Learning approach for volatility targeting models

hal.structure.identifierLaboratoire d'analyse et modélisation de systèmes pour l'aide à la décision [LAMSADE]
dc.contributor.authorBenhamou, Éric
dc.contributor.authorSaltiel, David
dc.contributor.authorTabachnik, Serge
dc.contributor.authorWong, Sui Kai
dc.contributor.authorChareyron, François
dc.date.accessioned2021-11-15T15:12:28Z
dc.date.available2021-11-15T15:12:28Z
dc.date.issued2021
dc.identifier.urihttps://basepub.dauphine.psl.eu/handle/123456789/22204
dc.language.isoenen
dc.subjectFeatures sensitivityen
dc.subjectWalk forwarden
dc.subjectPortfolio allocationen
dc.subjectModel-freeen
dc.subjectModel-baseden
dc.subjectDeep Reinforcement learningen
dc.subject.ddc006.3en
dc.titleDistinguish the indistinguishable: a Deep Reinforcement Learning approach for volatility targeting modelsen
dc.typeDocument de travail / Working paper
dc.description.abstractenCan an agent efficiently learn to distinguish extremely similar financial models in an environment dominated by noise and regime changes? Standard statistical methods based on averaging or ranking models fail precisely because of regime changes and noisy environments. Additional contextual information in Deep Reinforcement Learning (DRL), helps training an agent distinguish different financial models whose time series are very similar. Our contributions are four-fold: (i) we combine model-based and modelfree Reinforcement Learning (RL). The last model-free RL allows us selecting the different models, (ii) we present a concept, called "walk-forward analysis", which is defined by successive training and testing based on expanding periods, to assert the robustness of the resulting agent, (iii) we present a method based on the importance of features that looks like the one in gradient boosting methods and is based on features sensitivities, (iv) last but not least, we introduce the concept of statistical difference significance based on a two-tailed T-test, to highlight the ways in which our models differ from more traditional ones. Our experimental results show that our approach outperforms the benchmarks in almost all evaluation metrics commonly used in financial mathematics, namely net performance, Sharpe ratio, Sortino, maximum drawdown, maximum drawdown over volatility.en
dc.publisher.cityParisen
dc.relation.ispartofseriestitlePreprint Lamsadeen
dc.subject.ddclabelIntelligence artificielleen
dc.description.ssrncandidatenon
dc.description.halcandidatenonen
dc.description.readershiprechercheen
dc.description.audienceInternationalen
dc.date.updated2021-11-15T15:11:39Z
hal.author.functionaut
hal.author.functionaut
hal.author.functionaut
hal.author.functionaut
hal.author.functionaut


Files in this item

Thumbnail
Name:
DTU-VolatilityTargetting.pdf
Size:
1.287Mb
Format:
PDF
View/Open

This item appears in the following Collection(s)

Show simple item record