AI Research COPCOP_newsmacro:brent_daily

COP same-day lead on Brent next-day returns, conditional on OPEC headline intensity

728
Total valid trading days

The front-running story made sense on paper: an OPEC headline hits the tape, a large-cap E&P like COP reprices first because its liquidity attracts the fastest information, and Brent catches up a day later. That is not what the data show. Over 728 trading days, the lead slope on high-OPEC-headline days was 0.248 versus 0.187 on quiet days — a gap that statistical testing says is indistinguishable from noise.

The sharper relationship, if anything, lives on the quiet days, where the slope is highly significant but still modest. After roughly three years of daily alignment, headline intensity does not separate COP's same-day move from Brent's next-day reaction in any meaningful way. The full breakdown, charts, and methodology are below.

The research question

Over the past ~3 years, when OPEC-related headline intensity (quota, production cut/increase, OPEC meeting) spikes into its top quintile, does COP's same-day return lead Brent crude's next-day return by more than after low-intensity headline days? I expect COP to front-run OPEC supply decisions, repricing large-cap E&P before the crude market fully digests the headline.

How this was measured

COP minute bars were resampled to daily close-to-close returns. OPEC-related headline intensity was constructed from the COP news feed by counting daily headlines whose title/summary/topics matched OPEC, quota, production cut, production increase, output cut, output increase, or OPEC meeting. Trading days were split into high-intensity (count >= 80th percentile, floored at 1 headline) and low-intensity (count <= 20th percentile) buckets. The lead relationship is measured as the slope of Brent next-day return on COP same-day return within each bucket, plus an interaction test of whether the slope differs between the two buckets. Brent daily values were aligned to the COP trading calendar by forward fill and the next-day return was computed as close[t+1]/close[t]-1.

The key numbers

Total valid trading days
728
Days with both COP same-day return and Brent next-day return available
High-intensity days
55
OPEC headline count >= 1
Low-intensity days
673
OPEC headline count <= 0
High group lead slope
0.2484
units: Brent next-day return per 1.0 COP same-day return
Low group lead slope
0.1867
units: Brent next-day return per 1.0 COP same-day return
Slope difference (high - low)
0.0617
Positive means COP leads Brent by more on high-intensity days
Interaction p-value
0.7238
p=0.7238 >= 0.05 -> no statistically-clear slope difference
High group Pearson r
0.207
N=55
Low group Pearson r
0.133
N=673
High group slope p-value
0.1290
High-day slope p=0.1290 -> slope is not statistically clear
Low group slope p-value
0.0006
Low-day slope p=0.0006 -> slope is statistically clear
Mean Brent next-day return after high
-0.0190%
N=55 high-intensity days
Mean Brent next-day return after low
0.0473%
N=673 low-intensity days

Reading the numbers

Across 728 valid trading days, only 55 had a high OPEC-headline spike versus 673 quiet days. The lead slope was 0.2484 on high-intensity days versus 0.1867 on low-intensity days, but that gap is not statistically clear (p=0.7238), so the data do not support COP front-running OPEC headlines.

The charts

High OPEC-intensity days: COP same-day return vs Brent next-day return
What this chart says

This scatter of the 55 high-intensity days shows a wide cloud of points around the center, with no tight pattern. The fitted slope says Brent next-day return moves about 0.25% for every 1% move in COP on the same day, but the p-value of 0.1290 means this relationship is too noisy to be statistically clear. For the question at hand, this is the key chart: if COP reliably front-ran OPEC news on spike days, you would expect a much tighter, more consistent positive tilt.

Low OPEC-intensity days: COP same-day return vs Brent next-day return
What this chart says

The low-intensity scatter uses all 673 quiet days and again shows a loose positive relationship, with a slope of 0.1867. This slope is statistically clear (p=0.0006), meaning even on ordinary days COP's same-day move has some predictive link to Brent's next-day move. The visual point is that the quiet-day pattern is not dramatically weaker than the spike-day pattern in chart 0.

Lead slope of Brent next-day return on COP same-day return
What this chart says

The bar chart directly compares the two lead slopes: 0.2484 for high-intensity days versus 0.1867 for low-intensity days. High is numerically larger by 0.0617, which is in the direction the question expected, but the interaction p-value of 0.7238 says that difference is well within the range of random noise. With only 55 high-intensity days, the sample is too small to declare a real difference.

Lead relationship summary by OPEC headline intensity

bucketnslopepearson_rslope_pmean_brent_nextmean_cop_ret
High intensity550.24840.20720.129-0.0002-0.002
Low intensity6730.18670.13270.00060.00050.0004

The takeaway

Bottom line: no, the data do not support the idea that COP same-day moves front-run Brent more after OPEC headline spikes. On high-OPEC days the Brent-next-day-on-COP-same-day slope was 0.248, versus 0.187 on quiet days — a 0.062 difference that is statistically indistinguishable from noise (interaction p=0.72, meaning a gap this large would appear roughly 72 times out of 100 if there were no real difference). The high-day slope itself is not clearly nonzero (p=0.13), so the apparent extra front-running is a weak lean at best, not a signal. Interestingly, the cleaner lead shows up on ordinary low-headline days: slope 0.187 with p=0.0006, though the correlation is still modest (r=0.13). Average Brent next-day returns were slightly negative after headline spikes and slightly positive after quiet days, but that does not rescue the core hypothesis. Takeaway: after roughly three years of data, OPEC headline intensity does not separate out a meaningful COP lead — this is essentially a coin flip, and the front-running story remains unproven.

The fine print