Paper: arXiv 2409.08356

Authors: Zian Wang, Xinyi Lu

Abstract

This paper investigates the forecasting performance of COMEX copper futures realized volatility across various high-frequency intervals using both econometric volatility models and deep learning recurrent neural network models. The econometric models considered are GARCH and HAR, while the deep learning models include RNN (Recurrent Neural Network), LSTM (Long Short-Term Memory), and GRU (Gated Recurrent Unit). In forecasting daily realized volatility for COMEX copper futures with a rolling window approach, the econometric models, particularly HAR, outperform recurrent neural networks overall, with HAR achieving the lowest QLIKE loss function value. However, when the data is replaced with hourly high-frequency realized volatility, the deep learning models outperform the GARCH model, and HAR attains a comparable QLIKE loss function value. Despite the black-box nature of machine learning models, the deep learning models demonstrate superior forecasting performance, surpassing the fixed QLIKE value of HAR in the experiment. Moreover, as the forecast horizon extends for daily realized volatility, deep learning models gradually close the performance gap with the GARCH model in certain loss function metrics. Nonetheless, HAR remains the most effective model overall for daily realized volatility forecasting in copper futures.

Complexity vs Empirical Score

  • Math Complexity: 4.0/10
  • Empirical Rigor: 6.5/10
  • Quadrant: Street Traders — practical and empirical, lighter on theory

Why this score: The paper provides a standard empirical comparison between traditional econometric models (GARCH, HAR) and deep learning models (RNN, LSTM, GRU) applied to COMEX copper futures volatility. While it presents a solid practical evaluation across daily and intraday high-frequency data using robust loss functions like QLIKE, it relies on standard, well-established theoretical architectures with limited methodological novelty.

Research Flowchart

  flowchart TD
    A[Research Goal: Forecast COMEX Copper Futures Realized Volatility] --> B[Data Inputs: High-Frequency Copper Futures Data]
    B --> C[Data Processing: Construct Daily & Hourly Realized Volatility]
    
    C --> D[Econometric Models: GARCH & HAR]
    C --> E[Deep Learning Models: RNN, LSTM, & GRU]
    
    D --> F[Rolling Window Forecasting & Evaluation via QLIKE Loss]
    E --> F
    
    F --> G[Outcomes: HAR is best overall for Daily RV; DL Models outperform GARCH & match HAR on Hourly RV]

Cited References (26)

  1. Forecasting copper prices with dynamic averaging and selection models (2015)
  2. A component model for dynamic correlations (2011)
  3. Two are better than one: Volatility forecasting using multiplicative component GARCH-MIDAS models (2020)
  4. Volatility of copper prices and the effect of real interest rate changes (2018)
  5. Economic drivers of commodity volatility: The case of copper (2021)
  6. Impact of macroeconomic news on metal futures (2012)
  7. On the economic sources of stock market volatility (2008)
  8. Stock market volatility and macroeconomic fundamentals (2013)
and 18 more
  1. Dynamic conditional correlation: A simple class of multivariate generalized autoregressive conditional hetroscedasticity models (2002)
  2. The MIDAS touch: Mixed data sampling regression models (2004)
  3. Copper price determination: fundamentals versus non-fundamentals (2018)
  4. Metal volatility in presence of oil and interest rate shocks (2008)
  5. Asymmetric effects of oil shocks on carbon allowance price: evidence from China (2021)
  6. The impact of macroeconomic news on Chinese futures (2019)
  7. Modeling volatility and correlations between emerging market stock prices and the prices of copper, oil and wheat (2014)
  8. Forecasting changes in copper futures volatility with GARCH models using an iterated algorithm (2003)
  9. The macroeconomic variables impact on commodity futures volatility: A study on Indian markets (2021)
  10. Research on the Influence of Economic Policy Uncertainty on the Supply Chain Finance (2020)
  11. Economic policy uncertainty and commodity market volatility: implications for economic recovery (2022)
  12. COVID-19 impact on commodity futures volatilities (2022)
  13. Can LSTM outperform volatility-econometric models? (2022) — link
  14. Deep learning stock volatility with google domestic trends (2015) — link
  15. Machine learning for realised volatility forecasting (2020)
  16. Realized volatility forecasting with neural networks (2020)
  17. Volatility forecasting with machine learning and intraday commonality (2022) — link
  18. Volatility forecast comparison using imperfect volatility proxies (2011)

0 of 26 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.