COP daily-return beta to Brent by Treasury curve regime
A steeper Treasury curve is supposed to be a macro tell: better growth, firm oil demand, and energy names trading like leveraged crude. On paper that makes COP an interesting test — does its daily beta to Brent rise when the 10-year minus 2-year spread runs above its own trailing average? In practice, the question barely got off the ground.
Over 729 classifiable sessions, COP spent only 3 days in the above-average regime; the other 726 define the result. Beta to Brent on those ordinary days was 0.3516, essentially the full-sample 0.3520. The model reports a +0.2758 slope bump for those three days with a tiny p-value — a statistical curiosity, not evidence.
The steep-curve/high-beta thesis is untested here, not confirmed: the curve essentially never ran above its trailing average. The full regression breakdown follows below.
Over the past ~3 years, does COP's daily-return beta to Brent crude increase when the 10-year minus 2-year Treasury spread is above its trailing 12-month average? I expect a steeper yield curve to signal stronger growth and oil demand, making the E&P trade more like a high-beta play on crude.
How this was measured
Using daily closes for COP, I regress COP's daily return on same-day Brent spot returns separately for days where the prior day's 10Y-2Y spread sat above versus below its trailing 252-trading-day average. The spread is built from constant-maturity Treasury yields; its moving average uses a full 252-day window. The regime flag is lagged one session so the classification only uses data available at the start of the trading day being classified. A pooled regression with an interaction term between Brent return and the regime dummy tests whether the slope difference is statistically distinguishable, using heteroscedasticity-robust standard errors.
The key numbers
Reading the numbers
Of 729 days, only 3 had the 10Y-2Y spread above its trailing 12-month average, versus 726 at or below it. The pooled beta to Brent was about 0.352, but with only 3 'above' days the above-regime beta cannot be estimated, so the data cannot answer the question.
The charts
Chart 0 traces the 10-year minus 2-year Treasury spread from -0.87 percentage points at the start to 0.47 at the end, with a peak of 0.74 and an average near 0.13, while the trailing 12-month average line sits around -0.27. The detail worth your eye is how rarely the spread is above that trailing average: only 3 days in the whole sample. That extreme imbalance is why the regime comparison that follows is lopsided.
Chart 1 is a one-bar picture: the beta on days at or below the trailing-average spread is 0.3516, and there is no bar for the above-average regime because no beta could be estimated. The missing bar is the key detail — three days is far too few to produce a reliable slope, so we cannot tell whether COP's beta would have been higher. The full-sample beta of 0.352 is essentially the same as the below/at regime simply because that regime dominates the sample.
Regime-specific Brent betas
| Regime | N | Beta | SE | t-stat | p-value |
|---|---|---|---|---|---|
| Below/at 12m avg | 726 | 0.3516 | 0.0322 | 10.925 | 0 |
| Above 12m avg | 3 | ||||
| Difference (Above - Below) | 0.2758 | 0.0514 | 5.363 | 0 |
The takeaway
Short answer: no — this window doesn't give the yield-curve regime test anything to work with, because COP almost never traded above the trailing 12-month average spread. Of 729 classifiable days, just 3 were in the above-average regime; 726 sat below or at it. On those ordinary days, COP's daily beta to Brent was 0.3516, essentially the same as the full-sample beta of 0.3520. The interaction model does report a +0.2758 slope bump for the three above-average days with a tiny p-value, but with only 3 observations that is a statistical curiosity, not a real signal — a couple of sessions can drive the whole result. So the "steeper curve makes COP a high-beta oil play" idea is untested here rather than confirmed, and practically the curve barely ever ran above its own trailing average, making it an unusable regime for trading COP's oil sensitivity. The solid takeaway is that on the vast majority of days, COP's daily oil beta sits right around 0.35.
The fine print
- Only 3 sessions were classified 'above average' — the beta difference rests on 3 data points, so the tiny p-value is not meaningful evidence.
- The 252-day trailing average consumes the first year, so the regime comparison effectively covers roughly the final two years of the window.
- Treasury and Brent series are forward-filled onto COP's trading calendar, which can insert flat prices on market holidays and slightly alter returns.
- Single-ticker OLS estimate; it ignores asymmetric up/down oil moves and time-varying volatility within each regime.