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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Algorithmic trading has gained attention due to its potential for generatingsuperior returns. This paper investigates the effectiveness of deepreinforcement learning (DRL) methods in algorithmic commodities trading. Itformulates the commodities trading problem as a continuous, discrete-timestochastic dynamical system. The proposed system employs a noveltime-discretization scheme that adapts to market volatility, enhancing thestatistical properties of subsampled financial time series. To optimizetransaction-cost- and risk-sensitive trading agents, two policy gradientalgorithms, namely actor-based and actor-critic-based approaches, areintroduced. These agents utilize CNNs and LSTMs as parametric functionapproximators to map historical price observations to marketpositions.Backtesting on front-month natural gas futures demonstrates that DRLmodels increase the Sharpe ratio by $83\%$ compared to the buy-and-holdbaseline. Additionally, the risk profile of the agents can be customizedthrough a hyperparameter that regulates risk sensitivity in the reward functionduring the optimization process. The actor-based models outperform theactor-critic-based models, while the CNN-based models show a slight performanceadvantage over the LSTM-based models.

Commodities Trading through Deep Policy Gradient Methods