Candidate point selection using a self-attention mechanism for generating a smooth volatility surface under the SABR model
作者：
Highlights：
• We propose two models to generate a smooth volatility surface under the SABR.
• We utilize a transformer as a backbone network of the two models.
• The combined two models perform practitioner’s candidate point selection task.
• We test the models on the S&P500 and KOSPI200 market data.
• The combined models can be applied in other stochastic volatility models.
摘要
•We propose two models to generate a smooth volatility surface under the SABR.•We utilize a transformer as a backbone network of the two models.•The combined two models perform practitioner’s candidate point selection task.•We test the models on the S&P500 and KOSPI200 market data.•The combined models can be applied in other stochastic volatility models.
论文关键词：Candidate point selection,Self-attention mechanism,Transformer network,SABR model,Smooth implied volatility surface
论文评审过程：Received 3 August 2020, Revised 18 January 2021, Accepted 18 January 2021, Available online 9 February 2021, Version of Record 24 February 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.114640
