Double deep Q-learning for optimal execution
From MaRDI portal
Abstract: Optimal trade execution is an important problem faced by essentially all traders. Much research into optimal execution uses stringent model assumptions and applies continuous time stochastic control to solve them. Here, we instead take a model free approach and develop a variation of Deep Q-Learning to estimate the optimal actions of a trader. The model is a fully connected Neural Network trained using Experience Replay and Double DQN with input features given by the current state of the limit order book, other trading signals, and available execution actions, while the output is the Q-value function estimating the future rewards under an arbitrary action. We apply our model to nine different stocks and find that it outperforms the standard benchmark approach on most stocks using the measures of (i) mean and median out-performance, (ii) probability of out-performance, and (iii) gain-loss ratios.
Recommendations
- A reinforcement learning approach to optimal execution
- Deep reinforcement learning for the optimal placement of cryptocurrency limit orders
- Optimal execution in high-frequency trading with Bayesian learning
- Adaptive execution: exploration and learning of price impact
- Optimal liquidation through a limit order book: a neural network and simulation approach
Cites work
Cited in
(13)- Deep reinforcement learning for the optimal placement of cryptocurrency limit orders
- Adaptive execution: exploration and learning of price impact
- Applying deep reinforcement learning in automated stock trading
- Learning a functional control for high-frequency finance
- A reinforcement learning approach to optimal execution
- QuantNet: transferring learning across trading strategies
- Deep differentiable reinforcement learning and optimal trading
- The QLBS Q-Learner goes NuQLear: fitted Q iteration, inverse RL, and option portfolios
- On parametric optimal execution and machine learning surrogates
- Robust Q-learning algorithm for Markov decision processes under Wasserstein uncertainty
- Deep learning in finance: a review of deep hedging and deep calibration techniques
- Reinforcement Learning for Optimal Execution When Liquidity Is Time-Varying
- Reinforcement learning for trade execution with market and limit orders
This page was built for publication: Double deep Q-learning for optimal execution
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5093248)
