This paper studies carbon price variation using a generalized present value decomposition. We treat the carbon net convenience yield as a dividend-like cash flow analogue that measures the value of current allowance ownership relative to deferred ownership. Using EUA futures data from August 2008 to April 2026, we find that both carbon returns and net convenience yield growth contribute to carbon price variation, with net convenience yield growth playing a substantially larger role. The result is robust to funding cost adjustments, EU ETS phase subsamples, and several other robustness checks. Thus, allowances behave more like commodities, but with distinctive features.
This study examines extreme joint price movements in agricultural futures through the lens of the coexceedance framework. Focusing on two key substitute pairs, corn-soybean and winter wheat-spring wheat, we introduce supply-side information shocks from USDA reports as a novel determinant of coexceedances. Using partial ordered logit and quantile regression models, we analyze both the occurrence and the intensity of coexceedances in response to information shocks and macro-financial factors. The results show that USDA surprises significantly increase the likelihood and degree of coexceedances, with wheat markets exhibiting stronger and more persistent joint extremes than corn and soybeans. Effects are asymmetric: price-bearish surprises amplify downside risk, while price-bullish surprises intensify upside extremes. Macro-financial shocks further reinforce these patterns, particularly in wheat markets. These findings extend the coexceedance literature to agricultural futures, highlighting the systemic relevance of public information releases for risk management and market stability.
This study investigates excise tax base management in the oil industry. Drawing on economic theory concerning tax pass-through rates and demand elasticity, we hypothesize that firms manage the excise tax base to mitigate the threat of an excise tax increase. This study uses a quasi-experimental difference-in-differences research design, exploiting the exogenous shock created by the impending insolvency of the U.S. Highway Trust Fund and the ensuing 2005 SAFETEA-LU legislative debate, which raised a credible threat of an excise tax increase. Analyzing quarterly data from 2002 to 2007, we find that, relative to non-U.S. oil firms, U.S. firms accelerate sales and increase production and inventories during the debate window. The evidence suggests that firms manage the excise tax base and build inventories in anticipation of a potential rate hike. The effects are stronger among petroleum refiners than among exploration and production firms. Overall, the study contributes to research on consumption taxes by showing how firms manage non-income tax bases through operational decisions.
This study examines how stringent environmental regulations and associated firms' environmental impacts influence profitability, using a profit-persistence model. This model distinguishes between core and non-core profit, corresponding to permanent and transitory earnings, respectively. This approach allows for an analysis of how environmental variables and profit components jointly affect the persistence of profits. The empirical case is Norwegian salmon farming, where a new environmental regulation (traffic light system-TLS) was recently introduced. Using an index that reflects environmental impact, the empirical results suggest that the new environmental regulation strengthens the negative impact of non-core profit on the overall profit, leading to increased profit volatility. Additionally, the impact of the new regulation on profit and its combined effect with profit components vary across production regions depending on their environmental impacts. This study presents a novel method for comprehensively evaluating the impact of environmental regulations and the associated firms' environmental impacts on financial performance.
This study introduces a novel framework for modeling extreme spillovers in commodity markets, with a particular focus on the energy sector. The framework combines a factor Hawkes model with the peaks-over-threshold method to capture both the temporal clustering and cross-sectional propagation of tail risk. To address the high dimensionality of commodity markets, we employ factors and clusters as complementary tools for dimensionality reduction. Together, these reduce model complexity while preserving interpretability and forecasting accuracy. Empirical results based on daily losses from S&P GSCI components reveal that the energy sector is both the dominant source and the most vulnerable recipient of extreme spillover effects. By concentrating solely on the tail of the distribution, our approach complements but differs fundamentally from volatility-based models, avoiding biases tied to volatility specifications and offering a robust tool for analyzing systemic risk.
The relationship between oil and equity market volatility remains an important yet unsettled question in financial economics, with prior studies reaching conflicting conclusions. This study revisits this issue by replicating and extending the study of Maghyereh et al. (2016) on volatility spillovers between oil and global equity markets. Our findings provide a contrasting perspective: rather than acting as a dominant transmitter of volatility shocks, the oil market is more often a net receiver of shocks originating in major equity markets. In particular, leading equity markets such as the US, UK, and Germany are key sources of volatility transmission within the global system. We show that this divergence in results is driven by the choice of VAR forecast error variance decomposition method. A comprehensive set of robustness checks confirms the stability of our findings across alternative data frequencies, sample periods, oil proxies, market specifications, and estimation methodologies. Overall, our results contribute to the ongoing debate by providing new evidence on the directionality of oil-equity volatility spillovers and by underscoring the need for careful methodological consideration in empirical studies of financial connectedness.
Existing studies have documented substantial extreme connectedness between crude oil and agricultural commodity markets, with the potential to amplify cross-market shock transmission and intensify price instability. In the context of commodity financialization, this paper further examines the dynamic associations among speculative pressure, extreme connectedness, and commodity price bubble risk. Drawing on nearly two decades of data for WTI crude oil and nine representative agricultural futures, we quantify extreme connectedness within an improved tailevent driven network framework, construct a market-wide speculative pressure measure using Commitments of Traders (COT) position data, and identify bubble states via the GSADF approach. The results show that both speculative pressure and extreme connectedness exhibit episodic surges around major historical stress events, and that lagged speculative pressure is positively associated with subsequent extreme connectedness. Crucially, extreme connectedness continues to provide additional information about bubble risk beyond speculative conditions, suggesting that bubble risk is also closely related to the market's network structure in extreme states. Further evidence from a broader sample of financialized commodities shows that stronger extreme connectedness is also linked to more pronounced cross-commodity clustering of bubble episodes. These findings deepen our understanding of systemic instability in crude oil and agricultural commodity markets under commodity financialization and have practical implications for crossmarket risk monitoring and hedging.
This study uses Shanghai Crude Oil Futures (SC) as a proxy for the upstream segment of China’s petrochemical industry and investigates how its market efficiency influences five key downstream product markets. Considering that markets differ in how they absorb information and in their structural features, we employ the Feasible Exact Local Whittle (FELW) estimator to construct a continuous market efficiency index. To capture efficiency dynamics across different time horizons, the study applies the Maximal Overlap Discrete Wavelet Transform (MODWT) to decompose the efficiency series into short-, medium-, and long-term components. These are then examined by Quantile-on-Quantile (QQ) regression to trace the varying marginal effects across different efficiency states. The results reveal strong state dependence and structural differences in the efficiency transmission from SC to downstream markets. Among the five markets, Low-Sulfur Fuel Oil and Asphalt exhibit the most stable transmission patterns, with the former showing a “saddle-shaped” structure and the latter following a “dual-path” pattern. In contrast, the links between SC and the markets for Linear Low-Density Polyethylene and Polypropylene are highly nonlinear and less predictable. Purified Terephthalic Acid demonstrates a dual mechanism of efficiency resonance and long-term anchoring. These findings deepen our understanding of information efficiency within industrial value chains. They also offer practical insights for managing market risk, guiding price policies, and designing regulatory frameworks in the energy sector.
Excessive stress of electricity markets has posed a serious concern for the overall economic activities in the recent years. Measuring the stress is always a challenge, in this context, this study introduces an Electricity Market Stress Index (EMSI). This novel index is constructed using key market variables such as price volatility, supply-demand imbalance, bid-ask spread, and volume shocks. Our analysis utilizes a comprehensive daily dataset spanning from 2012, to 2024. To validate EMSI, we employ Quantile Vector Autoregression (QVAR) and Local Projection (LP) estimation, analysing its impact on electricity prices and broader financial markets, including sectoral indices such as Auto, Energy, and Metal. The Nifty Energy and Nifty Metal indices are the most impacted by electricity market stress, showing strong and statistically significant responses, particularly at high stress levels. In contrast, the Nifty Commodity Index is the least affected, with minimal sensitivity across quantiles, indicating relative insulation from electricity market fluctuations. The index successfully captures market distress during critical global events such as COVID-19, the Russia-Ukraine war, and the Israel-Palestine conflict. Our findings highlight EMSI’s predictive capability in anticipating electricity market disruptions and spill over into related financial sectors, offering a powerful tool for policymakers, market participants, and investors.
This study investigates the volatility connectedness between the international crude oil and 10 Chinese commodity sectors, with a particular emphasis on the contributions of common and idiosyncratic factors. We first decompose market returns into common and idiosyncratic components using the Generalized Dynamic Factor Model, and then apply GARCH models to estimate the volatility of each. Then the TVP-VAR-DY model (Time-Varying Parameter Vector Autoregression-DY Spillover Model) is employed to investigate the raw, common and idiosyncratic volatility linkages across markets. On this basis, pattern causality and quantile regression are adopted to identified the sources of volatility spillovers. Empirical results indicate that volatility spillovers between international crude oil and various Chinese commodity sectors are not only heterogeneous in nature but also driven by distinct factors. Common factors primarily govern the dynamics of these spillovers, whereas idiosyncratic factors determine their magnitude. Furthermore, during periods of market turbulence, idiosyncratic factors become the dominant driver, in contrast to stable periods where common factors prevail.
This paper employs wavelet quantile regression on daily data for rare earth elements (REE) to investigate the non-linear and horizon-specific effects of global and US-specific indicators of trade policy uncertainty (TPU) on herding behavior in the REE market during the US-China trade war. Spanning the period from January 2015 to April 2025, our sample covers the pre-Trump stability phase and subsequent tariff escalations phases. Our analysis reveals that REE investors generally exhibit anti-herding behavior which is driven by the herding patterns of the trade war Phase I, in contrast to heightened herding observed during Trump's early presidency. Among the horizonspecific herding effects associated with TPU indicators, US-based TPU exerts the most persistent and extensive non-linear influence on herding in the REE market, especially across short- and medium-term horizons. Importantly, the full-sample herding dynamics are largely shaped by market responses during Phase II of the US-China trade war. These findings highlight the dominant role of the US-origin TPU in shaping horizon-specific correlated trading in REE market. The findings carry important implications for investors, portfolio managers, and policymakers, suggesting the necessity to incorporate TPU dynamics into both investment strategies and regulatory frameworks to enhance resilience and responsiveness under heightened trade policy uncertainty.
This study examines asymmetric pass-through from oil to gasoline prices in the United States using weekly data from January 2000 to May 2023. Employing a smooth transition vector error correction model with generalized impulse response functions, we test three sources of asymmetry: the existence of an inaction band-a range of oil price changes too small to trigger any gasoline price adjustment, differences in error-correction speeds, and nonlinear responses to shocks of varying sign and size. The results reveal a clear inaction band with asymmetric thresholds: gasoline prices adjust quickly to small oil price increases but require much larger decreases before price cuts occur. Adjustment speed asymmetry across regimes is not statistically significant. Instead, asymmetry arises primarily through nonlinear impulse responses, where large positive shocks are passed through more rapidly and completely than negative shocks. These effects intensify during crisis episodes such as the COVID-19 pandemic and the early phase of the Russia-Ukraine conflict.
Accurate identification of dominant predictors remains pivotal for crude oil returns forecasting amidst complex multi-timescale interactions. Utilizing 2008-2023 monthly data from 56 predictors across macroeconomic, technical, climate and financial groups, we develop a multi-timescale framework combining wavelet decomposition with group shrinkage models. Three key findings emerge. First, climate and macroeconomic predictors demonstrate stable forecasting power during major crisis periods, while technical indicators show marked volatility. Second, group shrinkage models such as Group-QLASSO consistently outperform conventional approaches by capturing within-group predictor dynamics through group sparsity, finding that climate predictors maintaining strength across both short and long forecasting horizons. Third, timescale decomposition reveals the superior performance of predictors’ short-to-medium-term components, with medium-term components of climate predictors proving more informative than their macroeconomic counterparts. Methodologically, quantile-based regularization techniques retain robust under extreme market conditions. Notably, the substantial economic gains further validate the superior forecasting performance of our comprehensive forecasting framework, not only statistically but also economically. These findings offer actionable guidance for policymakers and market participants by highlighting the value of integrating climate and macroeconomic predictors’ short-to-medium-term dynamics into energy forecasting and early-warning systems.
This study develops an integrated ARMAX-GARCH anomaly-detection framework for identifying potential strategic bidding and manipulation in day-ahead electricity and other energycommodity markets. By combining econometric modeling of fundamentals with conditional volatility dynamics, the framework distinguishes between normal, fundamentals-driven price movements and anomalies indicative of strategic behavior. Using Turkiye's day-ahead electricity market as a detailed case study of an emerging liberalized system, the analysis demonstrates how the method can uncover hidden irregularities within auction-based commodity exchanges. The results show that price anomalies-defined as statistically significant deviations unexplained by fundamentals-cluster systematically during low-liquidity hours and intensify during periods of external stress, such as pandemic disruptions or global energy crises. These findings challenge the view that commodity-price volatility is purely random, revealing that market design and participant behavior jointly shape observed dynamics. The model further indicates that greater renewable generation and higher system capacity mitigate the probability of anomalous outcomes, underscoring the stabilizing role of supply diversification. Beyond its national context, the study contributes to the broader literature on commodity-market volatility, market power, and regulatory surveillance, showing how econometric diagnostics can serve as early-warning instruments for policymakers and market operators. By linking volatility modeling with enforcement analytics, the proposed methodology offers a replicable blueprint for monitoring strategic conduct and strengthening integrity across global energy-commodity markets.
How do climate-policy uncertainty and climate shocks affect systemic risk within clean and fossil-fuel energy markets? Could more advanced connectedness models help reduce this risk? This study investigates these concerns through the dynamic interconnections between climate risk and the leading energy commodities: natural gas, crude oil, and clean energy. Employing innovative text-based climate uncertainty indices capturing natural disasters, global warming, international summits, and U.S. climate policy, we apply the frequency-Quantile VAR model to unveil the asymmetric spillovers across time horizons and return quantiles. Our results show that events like the U.S. withdrawal from the Paris Agreement and the 2024 U.S. presidential election, significantly enhance interconnections among both renewable and conventional fossil-fuel energy markets. We also find that climate-policy uncertainty stemming from U.S. climate policy and international summits, consistently transmits risk in the high quantiles, driving both short-term volatility and long-term structural repricing in the energy commodity market. Our findings are useful guidelines for traders, portfolio managers, and policymakers aiming to hedge against tail risks and adapt to a decarbonizing global economy.
We explore the dynamic effects of the global common volatility (GCV) on selected precious metals and energy commodities by employing the two advanced econometric methods: Fourier quantileon-quantile regression and Fourier quantile regression. GCV induces volatility across gold, silver, and platinum markets in the short-, medium-, and long-term. However, short-term volatility in the silver market exhibits a negative relationship with GCV under both bearish and bullish conditions. Additionally, the oil, gas, and heating oil markets experience substantial losses due to GCV, with the impact intensifying from the short-to long-term across various market states. Moreover, the COVID-19 crisis and the ongoing Russia-Ukraine conflict have markedly strengthened volatility in precious metal and energy markets, reflecting an elevated level of GCV. Nonetheless, the natural gas markets exhibit a negative long-run relationship with GCV during the Russia-Ukraine conflict. Overall, our results underscore a strong interconnectedness between GCV and precious metals and energy markets, highlighting significant risks that global financial market volatility poses to these sectors.
We adopt a data-driven approach to examine uranium price explosiveness. We detect explosive episodes across varying durations and apply a LASSO-Logit framework to uncover key variables associated with price explosiveness. Our findings reveal that uranium price explosiveness is persistent, with positive explosiveness dominating and lasting an average of ten months. Variables such as dividend growth, monetary conditions, and expansion in the uranium sector significantly increase the likelihood of explosiveness. Additionally, uncertainty and geopolitical risks shape market dynamics. A local projections approach highlights that monetary tightening and uranium price momentum can sustain upward price pressures, while economic activity and sovereign debt risks exert downward forces. As uranium becomes increasingly vital to the transition toward a net-zero economy, our findings help bring greater transparency to a traditionally opaque commodity market.
Oil price changes have been considered a good (negative) predictor of stock market returns. In this study, we show via predictive regressions, for an extensive dataset of 44 developed and developing stock markets, that this negative relationship is present only up to the global financial crisis and has largely disappeared ever since. We document an evident shift in the predictive behavior of oil price changes after the 2008 global financial crisis, especially in developed stock markets, a finding that is robust to several additional tests we perform. A possible explanation of the change in the oil-stock return relationship post 2008 could be the increased significance of industrial metals (mainly copper and aluminum). We show that after 2008, industrial metals price changes have gained significance as stock market predictors, mainly in recessions, a finding that is partly consistent with the existing literature on stock market predictability.
This research analyzes the association between gold-mining stock returns and climate policy uncertainty (CPU) and examines whether CPU moderates the relationship between gold returns and gold-mining stock returns. Using monthly data for 68 gold-mining companies from nine countries over the period 2011-2022, we report that CPU exerts a significant and adverse effect on gold-mining stock returns, diminishing the positive impact of gold returns on gold stock performance. In contrast, Global Economic Policy Uncertainty (GEPU), Monetary Policy Uncertainty (MPU), and Fiscal Policy Uncertainty (FPU) are positively associated with gold-mining stock returns and strengthen the relationship between gold returns and mining stock performance, whereas Local Economic Policy Uncertainty (LEPU) does not exhibit a significant association. The results remain robust after correcting for endogeneity using an instrumental variable approach. Extending the analysis to energy transition metals, including copper, lithium, nickel, and cobalt, we find that climate policy uncertainty is positively associated with stock returns in these sectors and depending on time windows we find a negative impact of CPU on the link between metal returns and metal-stock returns.
This paper develops a continuous-time framework for pricing electricity future contracts that addresses key limitations of traditional models, particularly their inability to capture price spikes and shifts in hedging behavior. The proposed model incorporates both jump components and a time-varying drift to reflect dynamic changes in supply and demand for hedging. Additionally, correlated Brownian motions are included to capture common shocks across contracts with different delivery periods. Model parameters are estimated using the generalized method of moments (GMM) on daily settlement data from the Norwegian power market. Monte Carlo simulations confirm the consistency and robustness of the estimators. Out-of-sample forecasting exercises demonstrate superior predictive performance relative to standard ARMA-GARCH benchmarks. The results underscore the model’s practical relevance for traders and risk managers engaged in electricity portfolio management.
