METHODOLOGY FOR VALUING COMMODITY-SECTOR COMPANIES ON THE EXAMPLE OF SHELL PLC: A RETROSPECTIVE TEST

Abstract

This paper develops a methodological framework for valuing commodity-sector companies, with a focus on extractive firms, applied to Shell plc with a 13 June 2025 valuation date. The classical DCF model with a perpetual-growth assumption is shown unsuitable for a mature commodity firm whose primary asset is a finite natural resource, and requires methodological adaptation. Four approaches are proposed and empirically applied: (1) commodity-neutral DCF combined with Black-Scholes real options on undeveloped reserves, (2) cycle-normalized fundamentals using medians of revenue, operating margin and reinvestment rate over the 2010–2024 cycle, (3) commodity-price normalization via revenue-to-Brent regression, and (4) Monte Carlo simulation with correlated oil-price and operating-margin distributions. The cost of capital is derived bottom-up from a peer set of 300 oil-and-gas companies, with a country-risk premium weighted across Shell's disclosed asset geography and a sovereign-CDS-adjusted risk-free rate. Intrinsic-value ranges are computed for each method on a consistent share-count basis and synthesised into a single range. A distinguishing feature of the study is retrospective back-testing: the valuation date fortuitously preceded a major geopolitical shock in early March 2026 (escalation around the Strait of Hormuz, with Brent moving from roughly 71 to over 130 USD per barrel). Splitting the observation window into pre-shock and post-shock regimes, the authors measure the mean absolute gap between the realized price and each intrinsic-value estimate. In the pre-shock regime Monte Carlo demonstrates the tightest fit; in the post-shock regime commodity-price normalization becomes the most accurate method, while Monte Carlo loses accuracy because the empirical oil-price distribution did not contain observations above 130 USD per barrel; realized prices nonetheless remain within the bulk of the Monte Carlo distribution, confirming the robustness of the range-based design. The practical value of the study is an applied framework for domestic valuation of extractive enterprises relevant to Ukraine's post-war reconstruction and prospective privatization of mining assets, emphasizing probability-distribution outputs over single-point estimates and a real-options insurance layer against scenarios that point-estimate models cannot forecast.