大宗商品周期与结构变化的特征（英）

USA | ChemicalsEquity ResearchJune 17, 2025 Laurence Alexander *  | Equity Analyst(212) 284-2553 | lalexander@jefferies.comChristopher LaFemina, CFA *  | Equity Analyst(212) 336-7304 | clafemina@jefferies.comDaniel Rizzo *  | Equity Analyst(212) 336-6284 | drizzo@jefferies.comKevin Estok *  | Equity Associate(212) 778-8516 | kestok@jefferies.comXianrao Zhu *  | Equity Associate+1 (212) 778-8742 | xzhu@jefferies.comCarol Jiang *  | Equity Associate+1 (212) 284-1714 | cjiang@jefferies.comPatricia Hove, CFA *  | Equity Analyst(212) 707-6362 | phove@jefferies.comCommodity Cycles & the Fingerprints ofStructural ChangeWe sketch a range of frameworks for commodity prices, integrating cycleanalysis and ML. Demand shocks could drive upside for copper and natural gas,and supply shocks could support oil prices this summer. Risks of air pocketsappear highest for lithium and methanol, with a more balanced framework forcrop prices compared to last year. So near-term a relative risk for Buy-ratedALB and MEOH. For more structural shifts, our favorites are Buy-rated FCX andCTVA.Where Angels Fear To Tread: Commodity prices shift depending on shifts in the value of theirapplications and in how hard they are to bring to market. This primer sketches the degree that"reading the tape" and attempting to decipher price signals can illuminate the outlook for basiccommodities--what a bot sees, as it were, if it does not have any commodity-specific data on thescale or difficulty of supply additions, or if it does not discount novel structural changes in demand.A Diverse Tool Kit For A Range Of Questions: We deploy linear trend analysis to establishnaive benchmarks for each commodity cycle, lowess smoothers to put the spotlight on structuralchanges, regression and random forest models to highlight the direction and degree of volatilityexpected due to shifts in aggregate demand, and Fourier models to tease out key underlying cyclesburied within the observable data.Fingerprints Of Structural Change: Integrating the analysis of both recurring cycles and macrosensitivities provides, in our view, a better "base case" which we can then adjust depending onhow the current environment is fundamentally different. We highlight (Table 1) where our forecastslean more heavily on structural changes in regulatory policy, the cadence of capacity additions, andcustomer objectives, whereas the pure statistical and top-down frameworks lean the other way. Ourmacro-based models do a decent job, in particular, forecasting periods of unusual volatility (OOBAUC 85%) and the evolution of spot prices (Avg. R2 83%). When commodity prices are moving insync with the macro models, we recommend focusing on predicting macro conditions. One othertakeaway from these studies: when structural divergences emerge due to structural shocks, theyalmost always last only 2-4 years before new capacity, substitution effects, technology innovationsor policy shifts bring the commodity back into alignment with the longer-t