# Sector Correlation Matrix: 20-Day Return Correlations, as of Oct 22, 2026

Over the 20 sessions from Sep 25, 2026 to Oct 22, 2026, the most correlated pair of sector ETFs was XLF and XLRE (+0.76) and the least correlated was XLK and XLP (-0.80).

As of the Oct 22, 2026 session close. Page: https://squawkflow.com/correlation-matrix

## 20-session correlations, Sep 25, 2026 to Oct 22, 2026

| | XLK | XLY | XLC | XLF | XLRE | XLB | XLI | XLU | XLV | XLP | XLE |
|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
| **XLK** | +1.00 | +0.52 | +0.24 | -0.17 | -0.26 | -0.13 | +0.39 | -0.21 | -0.42 | -0.80 | -0.24 |
| **XLY** | +0.52 | +1.00 | +0.49 | +0.35 | +0.35 | +0.47 | +0.46 | -0.02 | +0.07 | -0.19 | -0.33 |
| **XLC** | +0.24 | +0.49 | +1.00 | +0.22 | +0.15 | +0.36 | +0.18 | -0.08 | -0.04 | -0.06 | +0.03 |
| **XLF** | -0.17 | +0.35 | +0.22 | +1.00 | +0.76 | +0.72 | +0.54 | +0.49 | +0.44 | +0.48 | +0.04 |
| **XLRE** | -0.26 | +0.35 | +0.15 | +0.76 | +1.00 | +0.59 | +0.44 | +0.61 | +0.45 | +0.45 | +0.01 |
| **XLB** | -0.13 | +0.47 | +0.36 | +0.72 | +0.59 | +1.00 | +0.53 | +0.24 | +0.24 | +0.42 | +0.11 |
| **XLI** | +0.39 | +0.46 | +0.18 | +0.54 | +0.44 | +0.53 | +1.00 | +0.35 | -0.16 | -0.05 | +0.17 |
| **XLU** | -0.21 | -0.02 | -0.08 | +0.49 | +0.61 | +0.24 | +0.35 | +1.00 | +0.12 | +0.35 | +0.11 |
| **XLV** | -0.42 | +0.07 | -0.04 | +0.44 | +0.45 | +0.24 | -0.16 | +0.12 | +1.00 | +0.54 | -0.38 |
| **XLP** | -0.80 | -0.19 | -0.06 | +0.48 | +0.45 | +0.42 | -0.05 | +0.35 | +0.54 | +1.00 | +0.18 |
| **XLE** | -0.24 | -0.33 | +0.03 | +0.04 | +0.01 | +0.11 | +0.17 | +0.11 | -0.38 | +0.18 | +1.00 |

## 10-session correlations, Oct 9, 2026 to Oct 22, 2026

| | XLK | XLY | XLI | XLC | XLF | XLRE | XLU | XLB | XLV | XLP | XLE |
|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
| **XLK** | +1.00 | +0.68 | +0.59 | +0.31 | -0.11 | -0.18 | -0.50 | -0.14 | -0.38 | -0.91 | -0.25 |
| **XLY** | +0.68 | +1.00 | +0.33 | +0.44 | +0.12 | +0.16 | -0.52 | +0.36 | +0.07 | -0.54 | -0.68 |
| **XLI** | +0.59 | +0.33 | +1.00 | +0.18 | +0.52 | +0.31 | +0.08 | +0.29 | -0.09 | -0.63 | -0.07 |
| **XLC** | +0.31 | +0.44 | +0.18 | +1.00 | +0.08 | +0.10 | -0.29 | +0.37 | -0.04 | -0.26 | -0.17 |
| **XLF** | -0.11 | +0.12 | +0.52 | +0.08 | +1.00 | +0.85 | +0.65 | +0.60 | +0.48 | +0.23 | -0.34 |
| **XLRE** | -0.18 | +0.16 | +0.31 | +0.10 | +0.85 | +1.00 | +0.66 | +0.49 | +0.54 | +0.25 | -0.21 |
| **XLU** | -0.50 | -0.52 | +0.08 | -0.29 | +0.65 | +0.66 | +1.00 | +0.24 | +0.20 | +0.47 | +0.26 |
| **XLB** | -0.14 | +0.36 | +0.29 | +0.37 | +0.60 | +0.49 | +0.24 | +1.00 | +0.16 | +0.10 | -0.38 |
| **XLV** | -0.38 | +0.07 | -0.09 | -0.04 | +0.48 | +0.54 | +0.20 | +0.16 | +1.00 | +0.55 | -0.44 |
| **XLP** | -0.91 | -0.54 | -0.63 | -0.26 | +0.23 | +0.25 | +0.47 | +0.10 | +0.55 | +1.00 | -0.08 |
| **XLE** | -0.25 | -0.68 | -0.07 | -0.17 | -0.34 | -0.21 | +0.26 | -0.38 | -0.44 | -0.08 | +1.00 |

## Biggest changes over the last 5 sessions

20-session correlation ending Oct 22, 2026 against the same window ending Oct 15, 2026. The two windows share 15 of 20 sessions, so changes are damped.

| Pair | Before | Now | Change |
|---|---:|---:|---:|
| XLY / XLU | +0.50 | -0.02 | -0.52 |
| XLF / XLB | +0.27 | +0.72 | +0.45 |
| XLY / XLRE | +0.74 | +0.35 | -0.39 |
| XLK / XLRE | +0.05 | -0.26 | -0.31 |
| XLK / XLU | +0.06 | -0.21 | -0.27 |
| XLF / XLP | +0.22 | +0.48 | +0.26 |
| XLK / XLE | -0.50 | -0.24 | +0.26 |
| XLY / XLP | +0.06 | -0.19 | -0.25 |

## Method

- Scope: only the 11 sector ETFs listed below, over 10 and 20 sessions. There are no individual stocks and no 60-session window on this page.
- Inputs: completed-session closing prices of the 11 SPDR sector ETFs (XLK, XLF, XLE, XLV, XLI, XLY, XLP, XLU, XLB, XLC, XLRE). A session counts once its close has settled, about 15 minutes after the market closes.
- Returns, not prices: each series is the daily close-to-close log return, ln(close today / close the previous session). Two funds that both drifted upward would correlate near +1 on prices with no relationship between their days, so prices are never correlated here.
- Aligned on one calendar of trading sessions: a return exists only when a fund has a close on that session and on the session before it. A missing close removes that return; it is never stretched into a two-day return, even if every fund misses the same session.
- Full window or blank: the window is the most recent N sessions, the same for every cell. A pair is shown only when both funds have a return on all of them; a fund with a missing close is left blank rather than computed on an older span.
- Pearson correlation of the two return series, from -1 (moved opposite) through 0 (no linear relationship) to +1 (moved together).
- Order: average-linkage clustering on 1 minus correlation, so funds that moved together sit next to each other.
- Limits: 10 and 20 sessions are short windows, so single large days move a coefficient a lot. A 60-session window is not offered because the close history held is up to 30 sessions per fund. A correlation describes the past window only.