An Analysis of Neural Network Architectures for Deep Quadratic Hedging
2024
- 1Citations
- 319Usage
Metric Options: CountsSelecting the 1-year or 3-year option will change the metrics count to percentiles, illustrating how an article or review compares to other articles or reviews within the selected time period in the same journal. Selecting the 1-year option compares the metrics against other articles/reviews that were also published in the same calendar year. Selecting the 3-year option compares the metrics against other articles/reviews that were also published in the same calendar year plus the two years prior.
Example: if you select the 1-year option for an article published in 2019 and a metric category shows 90%, that means that the article or review is performing better than 90% of the other articles/reviews published in that journal in 2019. If you select the 3-year option for the same article published in 2019 and the metric category shows 90%, that means that the article or review is performing better than 90% of the other articles/reviews published in that journal in 2019, 2018 and 2017.
Citation Benchmarking is provided by Scopus and SciVal and is different from the metrics context provided by PlumX Metrics.
Example: if you select the 1-year option for an article published in 2019 and a metric category shows 90%, that means that the article or review is performing better than 90% of the other articles/reviews published in that journal in 2019. If you select the 3-year option for the same article published in 2019 and the metric category shows 90%, that means that the article or review is performing better than 90% of the other articles/reviews published in that journal in 2019, 2018 and 2017.
Citation Benchmarking is provided by Scopus and SciVal and is different from the metrics context provided by PlumX Metrics.
Paper Description
This paper considers the deep stochastic optimal control methodology and examines the model uncertainty resulting from the choice of neural network architecture for quadratic hedging of a European Call option with transaction cost, under a discrete trading schedule. Researchers and quantitative finance practitioners have introduced various bespoke neural network architectures for the problem of deep hedging; however, we notice that methodology benchmarking in the literature has been limited to high-level performance analysis. Recognizing the intimate relationship between the deep stochastic optimal control and certain solution methods for trajectory optimization problems, we examine the interaction between model training and the choice of neural network architecture. We show that parameterizing the hedge ratio policies at each time step by an independent neural network is more consistent with the dynamics of the gradients in the ADAM optimization and results in better training and higher performance.
Bibliographic Details
Provide Feedback
Have ideas for a new metric? Would you like to see something else here?Let us know