Summer 2026

This research paper establishes that, among commodity futures markets, industrial metals stand out as uniquely informative about subsequent real economic activity. Specifically, returns in this sector forecast future industrial production growth and revisions in producers’ expectations across a broad cross-section of countries. This predictive relationship is particularly pronounced in economies that are more integrated into global markets, is primarily driven by demand-side shocks, and remains robust after accounting for trade exposure and the increasing financialization of commodity markets. Moreover, the information embedded in industrial metals prices is also reflected in international equity markets. By contrast, while energy and agricultural futures convey useful signals, their relevance is largely confined to countries with strong trade linkages to these commodities. Overall, the evidence highlights the informational content of commodity futures markets, and positions industrial metals as a consistent leading indicator of global economic conditions.

This paper examines the asset pricing implications of sustainability in commodity futures markets. Focusing on metals and agricultural goods, the authors collect production-based environmental footprint data covering greenhouse gas emissions and water consumption for 41 commodities. Examining the performance of green-minus-brown portfolios, the authors find no evidence that sustainability is priced in the cross-section of commodity futures returns. However, the paper documents substantial welfare benefits when diversifying traditional equity-bond allocations with low-carbon commodities, with investor welfare increasing by up to 22% when the commodity share reaches 20%. These findings reveal dual opportunities — both financial and environmental — arising from investments in low-footprint commodities.

The paper reports a significant downward shift in the marginal convenience yield for holding agricultural commodity inventory, attributing this shift to speculative demand shocks. The authors additionally show that the positive response of agricultural convenience yields to hedging demand shocks is higher compared to commodity supply shocks. Overall, the paper’s SVAR analysis verifies the Keynesian theory of normal backwardation by showing a long-lasting positive response of agricultural convenience yields to a hedging demand shock, thereby leaving the hedging demand as the main factor explaining the less frequently observed backwardations in agricultural futures markets.

The authors propose a stochastic spanning approach to assess whether a traditional portfolio of stocks and bonds spans augmented portfolios including commodities, foreign exchange, and real estate. The authors empirically show that in all seven portfolio combinations, the augmented portfolio is not spanned by the traditional one. Both parametric and non-parametric tests further confirm the paper’s results in an out-of-sample setting. Traditional investors face an opportunity cost of up to 0.77% per month by not switching to an augmented portfolio that includes alternative asset classes. Additional analysis demonstrates that diversification benefits are stronger in times of economic decline and financial market weakness.

This paper develops a flexible Sustainable Style-Integrated (SSI) framework to decarbonize commodity futures portfolios through a tilting function that dynamically decreases (increases) allocations to relatively high (low) carbon emitting commodities. Using 26 commodity futures from five sectors, the SSI framework is implemented with a range of tilts to obtain monthly long-short portfolios from December 1989 to May 2024. The portfolio evaluation shows that small sustainability tilts deliver the best tradeoff between financial performance and carbon efficiency. SSI investors can, at a moderate Sharpe ratio loss, hold portfolios with substantially lower emissions than the traditional style-integrated benchmark, while also enjoying greater resilience to climate news shocks. A panel regression analysis, grounded in the limits-to-arbitrage and hedging framework of Acharya et al. (2013), provides evidence consistent with a hedging-cost channel through which commodity futures portfolio decarbonization can contribute to reducing global emissions.

The shipping sector underpins global trade yet faces unprecedented disruption from geopolitical shocks and the demands of a rapid clean-energy transition. With well-to-wake emissions from shipping representing 2.1% of the global total and projected trade growth set to raise the sector’s carbon footprint, stringent decarbonization measures are accelerating. These policies are expected to increase freight costs, with potential inflationary consequences. This article investigates how international shipping costs influence U.K. inflation volatility using a Generalized Autoregressive Conditional Heteroscedasticity-in-Mean (GARCH-MX) framework incorporating sector-specific freight rate indices and key macroeconomic indicators. The authors show that higher shipping costs are a statistically significant driver of both the level and volatility of U.K. inflation, with the strongest effects arising from containerized freight. Scenario simulations, however, indicate that these inflationary effects are predominantly transitory, with impacts decaying over time even under high shipping cost increases. For instance, even a doubling of container freight costs raises U.K. inflation by approximately 0.20 percentage points above baseline over a two-year horizon, declining to 0.03 percentage points after a decade. The paper’s results offer evidence-based guidance for policymakers seeking to balance short-term inflation risks with long-run economic stability in the transition to net-zero shipping.

Prediction products in the form of binary options that settle based on the occurrence of discrete events have experienced a rapid expansion on U.S. derivatives exchanges. Once limited to conventional economic variables such as weather or macroeconomic indicators, these products now encompass elections, sporting events, and elements of popular culture. This overview provides an account of the regulatory challenges posed by this expansion, focusing on (a) jurisdictional issues between gambling and derivatives activity, (b) manipulation, and (c) insider trading.

The paper applies the recently developed quadratic normal model (QNM) to oil options. This three-parameter model assumes parabolic local volatility and extends the Bachelier model to a much broader class of “fat-tailed” distributions with two additional parameters related to skewness and kurtosis. The primary focus of this paper is to demonstrate how this model can be efficiently calibrated to option prices and used by volatility arbitrageurs in managing their portfolios. The authors calibrate model parameters daily to market prices of WTI options over an extensive twenty-five year period and analyze the dynamics and stability of parameters over time. The authors show two primary applications of the model for delta-hedging of market-making portfolios and for pricing digital (binary) options, which are also becoming increasingly popular in prediction markets.

The authors study the effect of supply and demand-induced oil price uncertainty on the cost of debt of U.S. loans. The paper estimates oil price uncertainty as the unpredictable component of crude oil price fluctuations. Interestingly, the paper’s findings reveal that supply-driven oil price uncertainty increases the cost of credit, while oil price uncertainty driven by demand shocks decreases the cost of bank loans. Additionally, the positive association between supply induced oil price uncertainty and interest loan spread is more pronounced for major users of oil, while the negative effect of demand driven oil price uncertainty on the cost of bank loans is stronger for oil-producing firms.

In this article, the author explores how the exponential growth of AI has unexpectedly reversed the declining trajectory of nuclear energy. A decade ago, many countries planned nuclear phase-outs amid safety concerns and the promise of renewables, yet today many more governments are extending nuclear reactor lifetimes while big tech companies invest in advanced nuclear technologies to satisfy AI’s voracious demand for reliable electricity. Drawing on recent projections, the author examines the rapid expansion of hyperscale data centers, the superior safety and low-emission profile of nuclear power compared to fossil fuels, and the emergence of small modular reactors as a solution for stable, carbon-free power. The author also contrasts U.S. leadership in nuclear-AI integration with Europe’s more cautious approach. Ultimately, AI is catalyzing a pragmatic nuclear revival essential for achieving decarbonization alongside the demands of the digital era.

This article examines operational water intensity in AI computing using the industry metric Water Usage Effectiveness (WUE) and the expanded metric WUE Source, which incorporates the water intensity of electricity generation. A regional comparison across major U.S. data center markets shows that differences in grid composition can drive substantial variation in total water intensity even when cooling architectures are identical. The analysis further demonstrates that water risk enters project economics primarily through siting and permitting uncertainty rather than direct operating cost. As hyperscale campuses scale toward hundreds of megawatts and municipal governance frameworks evolve, water is likely to emerge as a material determinant of the geography and pace of AI infrastructure deployment.

This paper applies machine learning (ML) methods to commodity futures return prediction, adapting the cross-sectional framework commonly used in equity research. Using gradient boosted trees trained on 126 theory-grounded features across 41 commodity contracts over a 30-year period, the study demonstrates that ML can effectively forecast cross-sectional returns and construct profitable long-short portfolios. Time-series momentum emerges as the dominant predictive signal, while skewness-based features contribute meaningfully at shorter horizons. An ensemble approach aggregating predictions across eight forecast horizons (5 to 200 days) achieves a Sharpe ratio of 2.4, with reduced volatility and generally improved drawdown control compared to individual models. These findings establish commodities as a viable and underexplored domain for systematic ML-based investing.

This issue features an in-depth conversation with Dr. Angelos Gkanoutas-Leventis, a senior energy executive with extensive experience across global low-carbon markets. Drawing on his career spanning economics, geopolitics, and large-scale renewable energy, Dr. Gkanoutas-Leventis shares his perspectives on offshore wind, energy storage, digitalization, and the evolving role of sustainable finance. In specific, the energy executive generously offers candid insights on what it really takes to deliver the energy transition at scale, from market design and capital allocation to policy stability and system resilience. The interview also highlights Dr. Gkanoutas-Leventis’ engagement with Bayes Business School as a Guest Industry Expert on the MSc Sustainable Management & Finance program, where he works closely with the next generation of sustainable finance professionals. Together, these experiences offer readers a sharp, practical and forward-looking view of where clean energy investment is heading, and what will matter most in the decade ahead.

Colin Waugh is a seasoned commodities trading practitioner and researcher with extensive experience in global macro investing, carbon markets, and commodity research. The conversation with Colin Waugh covers his career journey, insights into carbon and commodity markets, and perspectives on climate finance, technology, and regulation. The discussion also highlights his engagement with Bayes Business School and the MSc in Sustainable Management & Finance program.