EOG capex-decline + positive Brent tape: EOG vs XOP over next 30 trading days (N=2 conditional events)
A thesis only a quant could love: when EOG trims capex just as Brent is breaking higher, disciplined free-cash-flow generation should beat the broad E&P basket. Over the past three years, that setup occurred only twice — and EOG lagged XOP both times, trailing by about 44 basis points on average over the next 30 trading days. That is the opposite of the expected reward for restraint.
Two events is a coin flip, not a signal. The gap versus all other EOG report dates is not statistically distinguishable, which makes the honest read more useful than the headline: the market may not pay up for capex discipline into a firm tape, or the sample is simply too thin to say.
The full run is below — all ten EOG releases, the capex screen, the Brent filter, and how EOG vs XOP behaved across every report window.
Over the past ~3 years, when EOG reports a quarter-over-quarter decline in capital expenditures while Brent crude's trailing 30-day return is positive, does EOG outperform XOP over the next 30 trading days? I expect capex discipline in a firm crude tape to be rewarded as investors price free-cash-flow generation and buybacks rather than production growth.
How this was measured
EOG quarterly capex was read from EOG_fundamentals and matched to EOG_earnings release dates on fiscal_date_ending. A quarter was classified as a capex decline when capex was below the prior quarter's capex. Brent crude's trailing 30-trading-day return was computed from brent_daily_df using pct_change(30) and sampled one business day before the report date to respect T+1 macro publication lag. For every report date, the first trading day strictly after the report date was taken as the anchor, and the next-30-trading-day return was computed for EOG and XOP; outperformance is defined as EOG return minus XOP return over that same window. The conditional set is the intersection of capex-decline quarters and a positive Brent trailing 30d return; other EOG report dates form the comparison baseline.
The key numbers
Reading the numbers
Of 10 EOG report dates, only 2 had a capex cut plus rising Brent, and EOG trailed XOP in both—averaging -0.44% versus +0.23% over all report dates, with 0% of events above XOP. The p-value of 0.36 says this negative gap is not statistically distinguishable from zero.
The charts
This chart lines up the two earnings dates that matched the condition. On 2024-11-08 EOG fell about 9.8% versus 9.1% for XOP, and on 2026-02-24 EOG rose 13.4% versus 13.5% for XOP, so the EOG-minus-XOP gap was negative both times. The small but consistent lag is the visual opposite of what the capex-discipline thesis predicted.
Each dot is one of the two qualifying events, with Brent's trailing return on the x-axis and EOG's forward gap on the y-axis. Brent was up 3.7% and 9.9%, yet the EOG-vs-XOP gap was -0.7% and -0.2%. With just two dots, no slope is meaningful; what matters is that both dots sit below zero.
The conditional-event group on the left averages -0.44%, below zero, while the other eight EOG report dates average +0.4%—and even though that other group includes a -8.2% outlier, its mean is still positive. The conditional group's small sample of two observations makes this comparison descriptive. It says the specific setup in the user's question did not produce a positive EOG edge in this data.
Conditional events: capex-decline quarters with positive Brent 30d tape
| report_date | capex_qoq_mm | brent_30d_ret | EOG_30d | XOP_30d | EOG_minus_XOP |
|---|---|---|---|---|---|
| 2024-11-08 | -168 | 0.0365 | -0.0983 | -0.0912 | -0.0071 |
| 2026-02-24 | -120 | 0.0994 | 0.1336 | 0.1352 | -0.0016 |
The takeaway
No — in the data tested, EOG did not outperform XOP after reporting a capex decline into a positive Brent tape; it actually lagged. Only 2 of the 10 EOG report dates fit both conditions, and EOG trailed XOP over the next 30 trading days both times, by an average of about -0.44 percentage points (EOG +1.8%, XOP +2.2%). That sits below EOG's +0.23 percentage-point average edge over XOP across all 10 report dates, so the expected "discipline gets rewarded" pattern is not showing up. Statistically this is a coin flip, not a signal: with just two events and a p-value near 0.36, the result is far too thin to trust, and the gap versus other report dates is also indistinguishable (p≈0.67). The practical takeaway is that the thesis remains untested rather than disproven — you need many more qualifying quarters before treating this as a reliable edge.
The fine print
- Only 2 conditional events in 10 report dates; with n=2, the -0.44% average is descriptive, not a reliable estimate.
- Returns are measured from the first trading day after the report, so any same-day reaction to the earnings release is excluded.
- Brent's return uses the prior business day's close (T+1), avoiding look-ahead but potentially missing same-day macro moves.
- XOP is a broad E&P ETF; the comparison does not control for EOG-specific beta, factor exposures, or trading costs.