English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Optimized trade execution is to sell (or buy) a given amount of assets in agiven time with the lowest possible trading cost. Recently, reinforcementlearning (RL) has been applied to optimized trade execution to learn smarterpolicies from market data. However, we find that many existing RL methodsexhibit considerable overfitting which prevents them from real deployment. Inthis paper, we provide an extensive study on the overfitting problem inoptimized trade execution. First, we model the optimized trade execution asoffline RL with dynamic context (ORDC), where the context represents marketvariables that cannot be influenced by the trading policy and are collected inan offline manner. Under this framework, we derive the generalization bound andfind that the overfitting issue is caused by large context space and limitedcontext samples in the offline setting. Accordingly, we propose to learncompact representations for context to address the overfitting problem, eitherby leveraging prior knowledge or in an end-to-end manner. To evaluate ouralgorithms, we also implement a carefully designed simulator based onhistorical limit order book (LOB) data to provide a high-fidelity benchmark fordifferent algorithms. Our experiments on the high-fidelity simulatordemonstrate that our algorithms can effectively alleviate overfitting andachieve better performance.

Towards Generalizable Reinforcement Learning for Trade Execution