XLE daily return vs SPY high‑low range (equity turmoil proxy)
Three years of overlapping trading data just did not produce the pattern the energy‑turmoil thesis expects. XLE daily returns and SPY's intraday high‑low range—used here as a real‑time volatility proxy—come in at a Pearson correlation of -0.061, and the rank-based Spearman correlation is essentially flat at -0.015. The effect, if any exists, tilts slightly negative rather than positive.
The analysis below covers the full 753-day sample, the p-values behind those coefficients, and a 60-day rolling correlation that shows how unstable the relationship is over time. The short takeaway: on days when SPY's range widens, XLE does not reliably follow. The detailed evidence is in the body of the report.
For XLE over the past ~3 years, does its daily return correlate positively with SPY's daily high-low range (volatility proxy), suggesting energy stocks rally on days of heightened equity market turmoil? I expect a positive correlation because geopolitical supply disruptions spike both oil prices and equity volatility, pulling XLE higher.
How this was measured
Resampled minute bars for XLE and SPY to daily OHLCV. Computed SPY's intraday high‑low range as a percentage of close, a real‑time volatility proxy. XLE daily return is close‑to‑close percent change. Aligned both on common trading days. Measured correlation with Pearson and Spearman coefficients, including a 60‑day rolling Pearson to track time‑variation.
The key numbers
Reading the numbers
Across 753 trading days, XLE's return and SPY's high-low range show no meaningful positive link: Pearson r = -0.061 (p ≈ 0.095) and Spearman rho = -0.015 (p ≈ 0.673). The expected "turmoil lifts energy" pattern does not show up.
The charts
This scatter piles 753 daily points with SPY's intraday range on the horizontal axis and XLE's daily return on the vertical. The cloud is roughly circular and flat, not tilted upward: the center sits near XLE return 0.0006 and SPY range 0.056, and the points spread out with no tendency for high-range days to sit higher on the y-axis. The flat cloud is the visual version of the near-zero negative correlation, so big equity-range days don't systematically come with XLE gains.
This line tracks the 60-day rolling Pearson correlation from late 2023 through mid-2026. It starts near +0.29 and ends near +0.09, but it dips as low as -0.38 and peaks at about +0.30, with an average of roughly -0.06. So the relationship is not only tiny; it is also unstable, flipping sign over time and never showing a sustained positive stretch that would support the expectation.
The takeaway
The short answer is no: over 753 trading days, XLE's daily return does not move with SPY's high-low range in the way you expected—if anything, the relationship is slightly negative. The linear correlation is -0.061 with a p-value around 0.095, meaning there's roughly a 1-in-10 chance that a correlation this extreme is just luck. The rank-based correlation is even closer to zero at -0.015 (p≈0.67), which is essentially no signal at all. So this is not a case where the data are too thin to tell; with more than two years of daily observations, the evidence fairly convincingly rules out a meaningful positive link. The occasional oil-shock/equity-turmoil episode you're thinking of does not show up as a consistent pattern. In practice, SPY's intraday range has almost no standalone value as a predictor of XLE's daily return.
The fine print
- Correlation isn't causation: XLE and SPY's range may both respond to a common shock, like a geopolitical event or macro news, without one driving the other.
- SPY's high-low range is a noisy proxy for market volatility; a cleaner measure like the VIX could tell a slightly different story.
- The near-zero correlation is sensitive to outliers and to the specific Aug 2023–Jul 2026 window; the rolling correlation wanders around over time.
- The rolling 60-day correlation uses overlapping windows, which inflates the effective sample size, so pointwise moves in that chart shouldn't be over-interpreted.