# Reinforcement Learning for Systematic FX Trading

arXiv: https://arxiv.org/abs/2110.04745 · HTML: https://mlsanity.com/paper/2110.04745/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 5 of its 52 references that are in the map.
- First posted on arXiv 10 Oct 2021.

## Builds on (most cited first, 5 of 5)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Playing Atari with Deep Reinforcement Learning](https://mlsanity.com/paper/1312.5602/) (2013)
- [Asynchronous Methods for Deep Reinforcement Learning](https://mlsanity.com/paper/1602.01783/) (2016)
- [Deep Reinforcement Learning for Trading](https://mlsanity.com/paper/1911.10107/) (2019)
- [Online Learning with Radial Basis Function Networks](https://mlsanity.com/paper/2103.08414/) (2021)

## Built on by (most cited first, 3 of 3)

- [The Evolution of Reinforcement Learning in Quantitative Finance: A Survey](https://mlsanity.com/paper/2408.10932/) (2024)
- [The Recurrent Reinforcement Learning Crypto Agent](https://mlsanity.com/paper/2201.04699/) (2022)
- [Sequential asset ranking in nonstationary time series](https://mlsanity.com/paper/2202.12186/) (2022)

## Abstract (arXiv)

We explore online inductive transfer learning, with a feature representation transfer from a radial basis function network formed of Gaussian mixture model hidden processing units to a direct, recurrent reinforcement learning agent. This agent is put to work in an experiment, trading the major spot market currency pairs, where we accurately account for transaction and funding costs. These sources of profit and loss, including the price trends that occur in the currency markets, are made available to the agent via a quadratic utility, who learns to target a position directly. We improve upon earlier work by targeting a risk position in an online transfer learning context. Our agent achieves an annualised portfolio information ratio of 0.52 with a compound return of 9.3%, net of execution and funding cost, over a 7-year test set; this is despite forcing the model to trade at the close of the trading day at 5 pm EST when trading costs are statistically the most expensive.