Figure 6 · Cross-firm placebo · 20 donor units

ExxonMobil had the third-smallest gap of any energy firm

If the synthetic-control method is applied to each of the 20 energy firms as if it were the treated unit, ExxonMobil’s announcement-day gap is among the three smallest. A genuine governance shock would sit far out in the tail. This one sits near the bottom.

  1. 1Ovintiv0.017%
  2. 2ONEOK0.020%
  3. 3ExxonMobil0.026%
  4. 4Energy Transfer0.051%
  5. 5Baker Hughes0.257%
  6. 6Chevron0.270%
  7. 7Targa Resources0.542%
  8. 8APA Corporation0.853%
  9. 9EQT Corporation0.867%
  10. 10ConocoPhillips0.974%
  11. 11Devon Energy0.990%
  12. 12Diamondback Energy1.115%
  13. 13Phillips 661.224%
  14. 14Marathon Petroleum1.315%
  15. 15Occidental1.540%
  16. 16EOG Resources1.582%
  17. 17Valero2.005%
  18. 18Williams2.559%
  19. 19Halliburton3.501%
  20. 20Schlumberger3.762%
3 / 20
ExxonMobil’s rank among 20 firms by absolute synthetic-control gap (third smallest)
145×
How much bigger the largest placebo gap is than ExxonMobil’s
0.95
Permutation p-value — the announcement looked like an ordinary firm-day
Sources & methodology notes
Cross-firm placebo permutation p-valuep̂placebo = (1/N) Σi=1N 𝒲[|gapi,0| ≥ |gapExxonMobil,0|]
N = 21 (20 donors + treated). Canonical 250-day pre-window: p = 0.286, rank 6 of 21. Sensitivity 60-day pre: p = 0.500, rank 11 of 21. Per Abadie, Diamond & Hainmueller (2010).

This figure runs the synthetic-control estimator on each donor firm as if it were the treated unit, using the same 10-firm pre-registered donor pool (with the placebo-treated firm itself removed and remaining weights renormalized). The unsigned-rank position of ExxonMobil in the resulting distribution is the cross-firm placebo statistic recommended by ADH (2010) §5.

  1. Alberto Abadie, Alexis Diamond & Jens Hainmueller, Synthetic Control Methods for Comparative Case Studies, 105 J. Am. Stat. Ass’n 493 (2010). Establishes the cross-firm placebo permutation procedure operationalized in this chart. doi.org/10.1198/jasa.2009.ap08746.
  2. Alberto Abadie, Using Synthetic Controls: Feasibility, Data Requirements, and Methodological Aspects, 59 J. Econ. Literature 391 (2021). §6.2 establishes the rank-based inference framework: the empirical p-value is the share of placebo units with an absolute gap at least as extreme as the treated unit’s.
  3. R. A. Fisher, The Design of Experiments 17–21 (Oliver & Boyd 1935). Foundational permutation-inference logic underlying the cross-firm placebo rank test.
  4. Devon Energy Corp. & Coterra Energy Inc., Form 425 (Feb. 2, 2026); Devon Energy Corp., Press Release, Devon Energy and Coterra Energy Complete Merger (May 7, 2026). Documents the merger that retired CTRA from the 21-firm placebo universe; here the universe is 19 peers + ExxonMobil = 20 units (CTRA and LNG both absent from this CIQ pull).
  5. Shane Goodwin, Read the Fine Print: What ExxonMobil’s Proxy Actually Says About Texas Redomiciliation, Columbia Law School Blue Sky Blog (May 2026). Companion paper. Article fn. 28 reports ExxonMobil rank 21 of 22 in the published 22-unit placebo universe (yfinance-based, nested ADH). This CIQ replication on 20 units yields rank 3 of 20 (smaller universe, frozen weights); both place ExxonMobil in the lowest decile of the absolute-gap distribution. Both correspond to empirical p = 0.95.

Data attribution. Day-0 synthetic-control gap computed for each of 20 energy firms as if treated, using the article’s 10-firm pre-registered donor pool with the placebo unit removed (renormalized to 1.0). All daily prices from S&P Capital IQ (IQ_CLOSEPRICE_ADJ feed). CTRA (post-merger delisting) and LNG (CIQ identifier issue) are not in this snapshot; their inclusion in the published 22-unit run leaves ExxonMobil’s rank inside the lower decile of the distribution.

Source: Author’s calculations from S&P Capital IQ daily adjusted closing prices; cross-firm placebo procedure from ADH (2010) §5; methodology in Goodwin (May 2026) fn. 28.