COLD vs VXZ: Correlation
Measured on weekly returns over the past three years, Americold Realty Trust, Inc. (COLD) and iPath Series B S&P 500 VIX Mid-Term Futures ETN (VXZ) carry a correlation of -0.25, a negative link.
Data as of 2026-08-27 · refreshed every trading day · weekly returns · methodology
How correlated are COLD and VXZ?
Over the past 3 years, COLD and VXZ moved with a correlation of -0.25, which is negative, meaning they tend to move in opposite directions. The link has loosened recently: the 1-year correlation (-0.36) runs below the 3-year figure (-0.25). Over 5 years the correlation is -0.31, and the annualized covariance of weekly returns is -223.7 %².
VXZ is close to the least connected end of COLD's tracked universe, ranking #18 of 18. Correlation aside, the last 12 months split them widely, with COLD ahead by 24.4 points (+8.3% versus -16.1%).
How is this computed?
Pearson correlation on weekly returns: ρ(A,B) = cov(rA, rB) / (σA · σB), over windows of 52, 156 and 260 weeks. Covariance is annualized (×52) and expressed in %². Full definitions on the methodology page.
COLD vs VXZ: side by side
| COLD (Americold Realty Trust, Inc.) | VXZ (iPath Series B S&P 500 VIX Mid-Term Futures ETN) | |
|---|---|---|
| 1-year return | +8.3% | -16.1% |
| 5-year return | -49.6% | -53.1% |
| Volatility (ann.) | 35.4% | 25.6% |
| Beta vs S&P 500 | 0.70 | -1.31 |
| Max drawdown (3Y) | -67.1% | -36.4% |
| Market cap | $4.3B | – |
| P/E (trailing) | – | – |
| Dividend yield | 6.02% | – |
| Sector / category | US Listed | US Listed |
Year-by-year returns
| Year | COLD | VXZ |
|---|---|---|
| 2022 | -10.9% | +0.5% |
| 2023 | +10.1% | -44.0% |
| 2024 | -26.7% | -12.7% |
| 2025 | -36.2% | +5.7% |
| 2026 | +21.0% | -10.5% |
Calendar-year price returns; the current year is year-to-date as of the data date above.
Are COLD and VXZ good diversifiers for each other?
By historical standards, yes. A correlation of -0.25 means the two rarely move for the same reasons.
FAQ
What is the correlation between COLD and VXZ?
Using weekly returns as of 2026-08-27: -0.25 over 3 years, with -0.36 over the last year and -0.31 over 5 years.
Is VXZ a good diversifier for COLD?
By historical standards, yes. A correlation of -0.25 means the two rarely move for the same reasons.
What does a correlation of -0.25 mean?
A reading of -0.25 sits on a scale from −1 (opposite moves) through 0 (unrelated) to +1 (identical moves). Correlation captures direction, not magnitude or performance.
Use this data
$ curl https://www.pairbook.io/api/v1/pairs/cold-vs-vxz.json
Drop this badge in a README or notebook; it updates with the data:
[](https://www.pairbook.io/pair/cold-vs-vxz/)
The core API is free. Terms and every endpoint in the API documentation.
Related comparisons
Hubs: COLD correlations · VXZ correlations
