TTE intraday high-low range on extreme Brent movement days
Directional oil headlines make for good narratives, but the intraday price data at TTE tells a different story. Across roughly three years of sessions, the expected supply-fear asymmetry simply isn't there.
The question was straightforward: does TTE's high-low range widen more on days when Brent jumps more than 1% than on days when it drops more than 1%? The numbers say no. On 215 up-crude days, the average intraday range was 3.28% of the close; on 218 down-crude days it was 3.33% — a hair wider in the opposite direction, and statistically indistinguishable from noise. A t-statistic of -0.17 with a p-value of 0.87 makes this a clean null result, not an underpowered one.
The full methodology, charts, and distribution comparisons sit below, along with why the market may not behave the way the narrative suggests.
Over the past ~3 years, does TTE's intraday high-low range widen more on days when Brent crude rises more than 1% than on days when it falls more than 1%? I expect a larger range on up-crude days as supply disruption fears trigger aggressive buying and wider price swings, while down-crude days see more orderly selling.
How this was measured
Resampled TTE minute bars to daily OHLC, normalised the daily high-low range by close. Used the global macro frame brent_daily_df to compute daily Brent crude returns. Defined 'Up >1%' as days when Brent rose more than 1%, and 'Down >1%' as days when Brent fell more than 1%. Compared the TTE range distributions under these two regimes via a Welch two-sample t-test (unequal variance). The null hypothesis is that mean TTE range is the same on up and down Brent days.
The key numbers
Reading the numbers
Brent up >1% days (215) averaged a 3.28% intraday range; down >1% days (218) averaged 3.33%. That 0.05 percentage-point gap is negligible, and p=0.868 means it is nowhere near statistically clear.
The charts
The two box plots sit almost on top of each other, with the up-crude group centered around 3.28% and the down-crude group around 3.33%. Their spread tails overlap heavily, reaching roughly 27% on up days and 30% on down days, so there is no visible pattern of up-crude days producing wider swings. The key detail is that the down-crude distribution is if anything slightly broader at the top, which contradicts the expectation of bigger ranges on Brent up days.
This bar comparison makes the near-tie obvious: the down-crude bar is only a hair taller (3.33% vs 3.28%), and it is in the opposite direction from the hypothesis. A gap this small, with t-statistic -0.17 and p=0.868, means the data look like random noise rather than a real supply-fear widening effect. In plain terms, whether Brent jumps up or drops more than 1%, TTE's typical intraday range is about the same.
Regime summary
| Regime | N | Mean range | Std range |
|---|---|---|---|
| Brent UP >+1% | 215 | 0.0328 | 0.0281 |
| Brent DOWN <-1% | 218 | 0.0333 | 0.0327 |
The takeaway
No — over roughly the past three years, TTE’s intraday range is basically the same whether Brent rose or fell more than 1%. On 215 up-crude days the average (high-low)/close was 3.28%, and on 218 down-crude days it was 3.33% — a gap of about 0.05 percentage points, and it points the wrong way versus the supply-fear hypothesis. The t-statistic is -0.17 with a p-value of 0.87, so this is pure noise: if there were truly no directional effect, you’d expect to see a gap this large roughly 87% of the time by chance. That makes this a fairly conclusive null result at this sample size, not just an underpowered one. The practical takeaway is that the direction of a big daily Brent move does not appear to be a reliable driver of TTE’s intraday volatility. If anything, down days look microscopically wider, but it’s a coin flip either way.
The fine print
- The 1% Brent threshold is arbitrary; untested cutoffs like 1.5% or 2% could behave differently.
- TTE range uses all minute bars available, so pre-market/after-hours action may be included and inflate levels.
- No causality: Brent shocks and TTE volatility could both be driven by a common macro event.
- Within each regime ranges vary widely (std ~2.8% up days, ~3.3% down days), so the tiny mean gap is buried in broad overlap.