A structural hidden Markov model for forecasting scenario probabilities for portfolio loan loss provisions

作者：

Highlights：

• Scenario probabilities are forecasted via a Hidden Markov model.

• An adaptation to hidden Markov models which imposes a structure is proposed.

• The proposed adaptation outperforms the parameter-intensive benchmark.

摘要

•Scenario probabilities are forecasted via a Hidden Markov model.•An adaptation to hidden Markov models which imposes a structure is proposed.•The proposed adaptation outperforms the parameter-intensive benchmark.

论文关键词：Scenario forecasting,Hidden Markov model,Credit risk,IFRS 9

论文评审过程：Received 31 December 2021, Revised 3 April 2022, Accepted 27 April 2022, Available online 5 May 2022, Version of Record 14 May 2022.

论文官网地址：https://doi.org/10.1016/j.knosys.2022.108934