Energy's Oil Beta Is a Fair-Weather Friend
If there's one thing the market thinks it knows about energy stocks, it's that they move with oil. Crude goes up, producers go up; crude goes down, they get hit. The actual research coming out of the platform this week suggests that simple story is mostly fiction. Across several different regimes and signal definitions, oil beta turns out to be stubbornly stable—or, in some cases, it behaves the opposite of what you'd expect. The data leans toward a more complicated reality: the headline correlation is real, but it's far from constant, and the moments where it seems to matter most are exactly the moments where it fades.
The beta that doesn't budge
Take TotalEnergies. A recent study looked at TTE's daily-return beta to Brent crude across tight and slack labor markets. The conventional guess would be that when the economy is running hot and unemployment is low, energy stocks become more sensitive to oil—more demand, more pricing power, more correlation. The data says no. In the 17 months with unemployment below 4.2%, the average monthly beta was 0.2635; in the 19 high-unemployment months, it was 0.2360. That is a difference of barely 0.03, well within noise. In other words, the macro backdrop doesn't rewire the oil-stock link.
Baker Hughes tells a similar story from a different angle. When oil volatility spikes, you'd expect an oilfield services firm to become more sensitive to Brent. Over the past three years, on days when Brent's 20-day realized volatility was above its median, BKR's mean 60-day beta was 0.263, versus 0.255 in calmer periods. The t-stat of 0.46 and p-value of 0.646 put that gap squarely in coin-flip territory. High oil vol doesn't amplify BKR's oil exposure. It's not that these stocks are insensitive to oil; it's that the sensitivity is a constant, not a dial.
The volume twist that flips the script
The most interesting finding, though, is where the relationship actually weakens. A study on FANG (the ticker, not the tech group) measured its beta to crude, using USO as the proxy, split by volume regimes. On days when USO volume was in the top quartile—meaning oil traders were especially active—FANG's beta dropped to about 0.52, versus roughly 0.65 on lower-volume days. The difference of -0.13 is not just large; it's statistically meaningful, with only a 2-in-100 chance of arising by luck across 753 trading days. So exactly when oil is moving hardest and the crowd is betting on energy stocks to ride along, the correlation weakens. That looks less like a market that's pricing oil risk efficiently and more like a sentiment-driven overreaction on high-volume days.
Backtests: the market's two faces
Our backtest suite sharpens the picture further. Buying APA when it closes in the bottom 20% of its 20-day range returned +156.10% on $100,000 across 37 trades, with a 70% win rate, beating SPY by 87.80 points. That's a strong mean-reversion signal: beaten-down energy names bounce hard. But a similar contrarian trade on VLO—buying when Brent drops more than 2% but VLO closes up—was a disaster, losing -9.88% with a 35% win rate and trailing SPY by 78.18 points. Even the range-compression idea on XLE, buying when the ETF's daily high-low range is in the bottom decile of its 20-day range, only returned +12.66% with a 56% win rate, underperforming SPY by 55.64 points.
The common thread is that oil sensitivity isn't a reliable shortcut for trading energy stocks. The underlying relationship is real but flat; the moments that feel like obvious signals often are not. For anyone who treats energy as a simple leverage play on crude, these findings suggest the trade is more subtle. Context—volume, range, specific stock behavior—matters far more than the daily oil headline. The data leans toward one conclusion: in energy, the market's surest instincts are often the ones worth questioning.