Scholarlink

<p>This study rigorously evaluates the NHITS model’s forecasting performance against established benchmarks in realized volatility prediction. The findings of this study show that NHITS is superior to the GARCH, LSTM, and TFT models, making it a strong novel benchmark model candidate. However, the HAR and NBEATSx models are still preferred over NHITS for short-term forecasting tasks, while for long-term predicting tasks, NHITS is the best model and ought to be implemented in the financial industry. Lastly, this research also proposes a dynamic variation of the Model Confidence Set (MCS) for financial time series data allows for statistically temporal evaluation of forecasting models.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Evaluating the Efficacy of NHITS for Forecasting Stock Realized Volatility: A Comparative Analysis with Established Models

  • 期刊论文

  • 发表日期：

    2025年3月19日

查看全文
  • Hugo Gobato Souto

摘要

This study rigorously evaluates the NHITS model’s forecasting performance against established benchmarks in realized volatility prediction. The findings of this study show that NHITS is superior to the GARCH, LSTM, and TFT models, making it a strong novel benchmark model candidate. However, the HAR and NBEATSx models are still preferred over NHITS for short-term forecasting tasks, while for long-term predicting tasks, NHITS is the best model and ought to be implemented in the financial industry. Lastly, this research also proposes a dynamic variation of the Model Confidence Set (MCS) for financial time series data allows for statistically temporal evaluation of forecasting models.