Journal of Futures Markets最新文献

We investigate the impact of heterogeneous investor attention on carbon prices. Although previous research has demonstrated the influence of investor attention on asset prices, how investors attribute attention toward the carbon-market, emission, and environment issues, and their consequences on carbon prices remain unknown. Leveraging data from the European Union (EU) and Chinese (Guangdong and Hubei pilots) carbon markets, our regression results reveal that investor attention toward the carbon-market and environment issues negatively affects carbon prices in the EU Emissions Trading System, whereas attention toward the carbon-market and emission issues exhibits positive impacts in the Guangdong pilot. Interestingly, these effects are reversed in the subsequent month, implying a lasting effect of investor attention. Additional analyses suggest that the different reactions in the EU and China can be attributed to investor types and green assets diversity. Finally, we demonstrate the economic usefulness of heterogeneous investor attention in improving carbon trading investors' trading strategies.

Implied covariance asymmetry is a market-wide measure defined as the average of the absolute difference between the downside and upside pairwise co-movements of individual stocks, estimated from options data. Its risk premium is linked to improved long-term economic conditions and significantly forecasts excess market returns from 1 month to 2 years. This predictive power persists at horizons beyond 6 months after controlling for popular financial and economic predictors in in-sample analyses. It also translates into superior out-of-sample forecasts and substantial economic gains for a mean-variance investor, particularly over medium and long horizons.

We examine trading behavior in commodity futures markets in the United States during the 2008 Global Financial Crisis (GFC) and the COVID-19 pandemic, focusing on absolute changes and relative exposure dynamics. These crises led to distinctly different trading patterns. During the 2008 GFC, speculators rapidly closed long positions while producers facilitated these trades, shifting risk from speculators to producers. In contrast, during the COVID-19 crisis—characterized by milder financial stress and an early commodity market rally—there was not meaningful risk transfer from speculators. The impact on traders' relative exposures was minimal in both crises. However, speculators generally showed greater sensitivity to changing financial conditions than hedgers throughout the study period. These findings highlight the varying impacts of financial stress on commodity futures markets and the importance of crisis-specific context in understanding trader behavior.

This paper evaluates the association between carbon emissions and tail-risk spillovers in European futures markets. We propose an innovative quantile LSTM-KAN model to capture the time-varying, nonlinear dynamics of tail-risk spillover networks. Using data from 29 EU futures markets, we find that tail-risk spillovers increase significantly during key events, including the 2016 Brexit referendum and the 2020 COVID-19 pandemic. Oil, natural gas, and EU allowance futures play central roles as recipients of tail risk, whereas bond and low-carbon futures exert tail-risk spillovers on other markets. In addition, we analyze the impact of $��\; ��{\text{CO}}_2��$ emissions on tail-risk spillovers. Higher $��\; ��{\text{CO}}_2��$ emissions significantly increase the tail-risk spillovers received by EU allowance futures and low-carbon equity futures. In low-volatility periods, $��\; ��{\text{CO}}_2��$ emissions increase the spillovers transmitted from oil and gas sector futures to other markets. In high-volatility periods, they intensify the tail-risk spillovers received by crude oil futures.

This study innovatively develops an interpretable mixed-frequency feature interaction deep learning network (IMF-FIDNet) to improve high-frequency grain futures price prediction via effective multi-frequency data integration, with a focus on ensuring robustness amid market uncertainty. By refining advanced mixed-frequency processing methods, proposing a new deep learning model, and integrating multiple modules, IMF-FIDNet enhances feature interaction modeling between low-frequency uncertainty indicators and high-frequency grain prices. Experiments show it outperforms traditional models in accuracy and robustness, and effectively supports investment decisions; further, its interpretability quantifies uncertainty indices' contributions, confirming macro-indicators' role in high-frequency price forecasting.

In this paper, we investigate the tail dependence and risk spillovers between International Energy Exchange (INE) crude oil futures and global crude oil benchmarks (WTI and Brent), as well as its underlying spot markets, by integrating the ARMA–GARCH-skewed-$��\; ��t��$ model with the Copula-CoVaR framework. Using high-frequency data with synchronized trading windows, we find consistently strong tail dependence across all sessions, supporting the role of INE as an emerging Asian benchmark. Risk spillovers are asymmetric, with downside risk dominating. INE functions as an information sender during daytime trading, characterized by notable volatility transmission, whereas nighttime spillover is more stable and symmetric. Moreover, INE is more sensitive to extreme events such as COVID-19 pandemic and the Russia–Ukraine conflict during its domestic trading hours. Our findings offer practical implications for market regulation, emphasizing the need to improve nighttime liquidity and enhance systemic risk monitoring under time-varying uncertainty.

We leverage the capacity of large language models such as Generative Pre-trained Transformer (GPT) in constructing factor models for Chinese futures markets. We successfully obtained 40 factors to design single-factor and multi-factor portfolios through long-short and long-only strategies, conducting backtests during the in-sample and out-of-sample periods. Comprehensive empirical analysis reveals that GPT-generated factors deliver remarkable Sharpe ratios and annualized returns while maintaining acceptable maximum drawdowns. Notably, the GPT-based factor models also achieve significant alphas over the IPCA benchmark. Moreover, these factors demonstrate significant performance across extensive robustness tests, particularly excelling after the cutoff date of GPT's training data.