No-Transaction Band Network: A Neural Network Architecture for Efficient Deep Hedging

Imaki, Shota; Imajo, Kentaro; Ito, Katsuya; Minami, Kentaro; Nakagawa, Kei

doi:10.2139/ssrn.3797564

Cited by 3 publications

(3 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Horvath et al [15] proposed a deep hedging model in rough volatility models such as the rBergomi model [4], which includes non-Markov price jumps. Imaki et al [18] proposed a deep hedging model considering a no-transaction band [12] into the neural network architecture and achieved fast convergence of learning.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Efficient Learning of Nested Deep Hedging using Multiple Options

et al. 2023

View full text Add to dashboard Cite

Section: Related Workmentioning

confidence: 99%

“…Here, the recursion will become deeper as the time to maturity is longer. Learning of deep recurrent networks is generally difficult because of its slow convergence (see e.g., [18]).…”

Section: B Proposed Solution For Nested Deep Hedgingmentioning

confidence: 99%

Efficient Learning of Nested Deep Hedging using Multiple Options

et al. 2023

View full text Add to dashboard Cite

“…Later in Section 5 we use Entropic Risk Measure(ERM) as a metric to verify our models. Hedging of assets which leverage risk measures output exponential utility indifference price, which is commonly used to benchmark the performance of neural hedging strategies against traditional models [5,6]. Considering exponential utility, u(x) = −exp(λ − x) where the risk aversion coefficient is λ > 0, the ERM is define as Equation 5.…”

Section: Entropic Risk Measurementioning

confidence: 99%

Neural Networks for Delta Hedging

Son¹,

Kim²

2021

Preprint

View full text Add to dashboard Cite

The Black-Scholes model, defined under the assumption of a perfect financial market, theoretically creates a flawless hedging strategy allowing the trader to evade risks in a portfolio of options. However, the concept of a "perfect financial market," which requires zero transaction and continuous trading, is challenging to meet in the real world. Despite such widely known limitations, academics have failed to develop alternative models successful enough to be long-established. In this paper, we explore the landscape of Deep Neural Networks(DNN) based hedging systems by testing the hedging capacity of the following neural architectures: Recurrent Neural Networks, Temporal Convolutional Networks, Attention Networks, and Span Multi-Layer Perceptron Networks In addition, we attempt to achieve even more promising results by combining traditional derivative hedging models with DNN based approaches. Lastly, we construct NNHedge, a deep learning framework that provides seamless pipelines for model development and assessment for the experiments.Keywords Black-Scholes • Neural Derivative Hedging • NNHedge 1. Timely evidence that neural networks score lower profit and loss, when taught traditional hedging strategies.2. Observations that neural networks concentrate more on historical data to calculate present-day delta values.3. The open-sourcing of NNHedge, a deep learning framework for neural derivative hedging. 1

show abstract