Figure 4 · Synthetic ExxonMobil — composition

A made-up ExxonMobil, built from 10 other oil firms

To know what ExxonMobil would have done on announcement day in a parallel universe where it didn’t redomicile, we weight ten energy peers into a synthetic ExxonMobil. The composition is pre-registered.

10 firms · pre-registered weights
  1. Chevron41.7%
  2. Diamondback Energy14.7%
  3. EOG Resources13.5%
  4. Schlumberger7.8%
  5. Williams5.6%
  6. Baker Hughes4.5%
  7. Occidental3.4%
  8. ConocoPhillips2.2%
  9. Marathon Petroleum*1.7%
  10. Phillips 661.7%
Sources & methodology notes
Synthetic-control donor weights (Abadie, Diamond & Hainmueller 2010)ω̂ = argminω ∈ Ω ‖ Y1,pre − Σj ωj Yj,pre ‖2
subject to   ωj ≥ 0,   Σj ωj = 1
SLSQP solver on the unit simplex; 11 non-zero weights from the 20-firm S&P Capital IQ oil & gas donor pool.

The synthetic-control donor pool composition shown above rests on a methodological tradition with three historical pillars: the original synthetic-control framework for political-economy case studies, its formalization for comparative case-study estimation with statistical inference, and the event-study research design within which the announcement-day reading is taken. Each is sourced below, followed by the firm-specific provenance (SEC filings, merger documentation, and the companion paper).

  1. Alberto Abadie & Javier Gardeazabal, The Economic Costs of Conflict: A Case Study of the Basque Country, 93 Am. Econ. Rev. 113 (2003). Original synthetic-control paper. Introduces the weighted-donor counterfactual construction for a single treated unit; predates the formal statistical-inference framework of Abadie, Diamond & Hainmueller (2010). Method directly applied here to ExxonMobil as the treated unit.
  2. Alberto Abadie, Alexis Diamond & Jens Hainmueller, Synthetic Control Methods for Comparative Case Studies: Estimating the Effect of California’s Tobacco Control Program, 105 J. Am. Stat. Ass’n 493 (2010). Foundational synthetic-control methodology; establishes the donor-pool weighting protocol and pre-period RMSPE fit-quality discipline used here. doi.org/10.1198/jasa.2009.ap08746.
  3. Stephen J. Brown & Jerold B. Warner, Using Daily Stock Returns: The Case of Event Studies, 14 J. Fin. Econ. 3 (1985). Foundational empirical-event-study paper establishing the daily-returns research design within which the announcement-day synthetic-control gap is taken. Documents small-sample inference properties for the parametric tests reported in fn. 27.
  4. Alberto Abadie, Using Synthetic Controls: Feasibility, Data Requirements, and Methodological Aspects, 59 J. Econ. Literature 391 (2021). Establishes feasibility criteria for synthetic-control inference; the §6.2 post/pre RMSPE ratio threshold of 2.0 referenced in the right-card “fit quality” callout. The 0.88 reported here is well below the rejection threshold.
  5. ExxonMobil, Preliminary Proxy Statement (Form PRE 14A), SEC Accession No. 0001193125-26-098908 (Mar. 10, 2026). Establishes T0 = March 10, 2026 as the event date; this is the SEC filing date of the preliminary proxy disclosing the New Jersey–to–Texas redomiciliation that is the subject of the announcement-day event study.
  6. ConocoPhillips and Marathon Oil Corp., Joint Press Release, ConocoPhillips Completes Acquisition of Marathon Oil (Nov. 22, 2024); SEC Form 8-K, Accession No. 0001193125-24-258711 (filed Nov. 22, 2024). Documents the Marathon Oil (MRO) merger closing that retired the ticker from public trading prior to the event-window estimation period. The pre-registered donor pool originally specified MRO at 1.7%; Marathon Petroleum (MPC) substituted as the closest available proxy. A 9-firm sensitivity dropping the slot entirely yields Day-0 = −0.003% (Δ 2.4 bp from published +0.021%).
  7. A. Craig MacKinlay, Event Studies in Economics and Finance, 35 J. Econ. Literature 13 (1997). Canonical reference for the event-study research design adopted here: 240-day estimation window, Day-0 abnormal-return framework, and the standardized residual test (Patell-z) used for parametric inference on AR.
  8. James M. Patell, Corporate Forecasts of Earnings Per Share and Stock Price Behavior: Empirical Tests, 14 J. Acct. Res. 246 (1976). Introduces the standardized abnormal-return test statistic used in the multi-window battery; accounts for the additional out-of-sample prediction error variance not captured by the raw residual variance.
  9. Shane Goodwin, Read the Fine Print: What ExxonMobil’s Proxy Actually Says About Texas Redomiciliation, Columbia Law School Blue Sky Blog (May 2026); replication kit on file with the SMU Corporate Governance Initiative. Author’s companion paper from which the donor weights and event-date specification are taken; (May 2026) incorporates the MRO/MPC substitution disclosure and the CTRA cross-firm-placebo merger note. Data source: S&P Capital IQ daily adjusted closing prices (IQ_CLOSEPRICE_ADJ feed); kit’s full_results.json reproduces published numbers within ±0.5pp tolerance.

Logo attribution. Corporate logos used to identify donor pool firms are reproduced from each firm’s corporate website via Google’s S2 favicon service at chart-element size, for editorial identification only. All marks are trademarks of their respective owners; no endorsement is implied.

Source: Author’s calculations from S&P Capital IQ daily adjusted closing prices; weights pre-registered in Goodwin (May 2026), event-study methodology footnote. Marathon substitution documented in audit memo 06.