Q4 Sector Seasonality — Which Sector Leads in Q4?
11 sector SPDRs plus SPY, QQQ, IWM, SMH, XBI, GLD; total-return month-end closes (yfinance, auto-adjusted),
1998-12 → 2026-09; 27 calendar Q4s (Oct–Dec), 1999–2025. Excess = ETF minus SPY in the same window; t = mean / std × √n
over the 27 yearly excesses. Evidence level C (group comparison, no control, no costs). Generated 2026-09-24.
In one line: Q4 is the best quarter for almost everything — SPY +4.9% on average,
up in 81% of years — but the sector edge on top of that is small and no positive sector reaches t = 2.
The one robust finding is negative: utilities and staples lag SPY in about two thirds of Q4s, in both halves of the
sample. If a tilt is wanted: semis / Nasdaq into October, industrials and small caps for November–December,
stay out of XLU and XLP. A tilt on top of being long — not a standalone bet.
1. Q4 excess return over SPY, 1999–2025
| ETF | Excess mean (pp) | Years beating SPY | t | 1999–2012 | 2013–2025 | Worst excess |
|---|
| XLB Materials | +2.2 | 70% | +1.5 | +4.9 | −0.6 | −14.7 |
| QQQ | +1.8 | 67% | +0.9 | +3.4 | +0.0 | −25.8 |
| SMH Semis | +1.6 | 59% | +0.8 | −0.6 | +4.1 | −22.7 |
| XLK Tech | +1.4 | 52% | +0.9 | +1.7 | +1.0 | −24.1 |
| XLI Industrials | +1.1 | 59% | +1.2 | +1.5 | +0.6 | −11.0 |
| XLY Discretionary | +0.9 | 56% | +0.8 | +2.5 | −0.9 | −16.6 |
| XLF Financials | +0.5 | 52% | +0.4 | −1.7 | +2.9 | −14.5 |
| IWM Small caps | +0.4 | 48% | +0.4 | +0.6 | +0.3 | −14.4 |
| XLV Health care | −0.2 | 52% | −0.2 | +0.1 | −0.5 | −12.8 |
| XLP Staples | −0.9 | 30% | −0.7 | −0.6 | −1.3 | −12.8 |
| XLE Energy | −1.1 | 37% | −0.6 | +0.4 | −2.6 | −18.0 |
| XLU Utilities | −1.9 | 37% | −1.2 | −1.8 | −2.1 | −19.3 |
Absolute Q4 returns for reference: SPY mean +4.9% / median +7.0% / 81% positive / worst −21.6% (2008).
XLI 85% positive, XLB 81%, XLK 81%, XLU 74%, XLE 67%. XLRE (2015) and XLC (2018) have too few years to read.
2. Month by month (excess over SPY, mean pp and share of years beating SPY)
| Month | Leaders | Laggards |
|---|
| October | QQQ +1.5 (74%), XLK +1.4 (67%), SMH +1.3 (54%), XLF +0.5 (56%) |
XBI −2.2 (40%), XLV −0.8 (33%), XLE −0.6 (37%) |
| November | XBI +2.4 (60%), SMH +1.5 (58%), XLB +1.2 (67%), XLI +1.1 (78%), IWM +0.6 (58%) |
XLU −1.6 (33%), GLD −1.3, XLE −0.6 (37%) |
| December | IWM +0.8 (42%), XLB +0.8 (56%), XLV +0.7 (52%), XLI +0.4 (63%), XLF +0.4 (59%) |
XLK −0.5 (41%), SMH −0.5, QQQ −0.3 (44%) |
Tech is the October story and fades by December. November is broad — small caps, industrials, materials and biotech lead.
December leadership shifts to small caps, materials, health care and financials.
3. What holds up and what does not
- Materials tops the 27-year table, but the entire edge is 1999–2012 (the China commodity cycle); since 2013 it is
flat to negative. Not a reason to buy XLB in Q4.
- Industrials is the only sector positive in both halves with the second-best t-stat — modest, but the most consistent.
- Semis / Nasdaq carry the October edge (QQQ beat SPY in 74% of Octobers). SMH's Q4 edge is all post-2013 (+4.1 pp) —
the AI / semiconductor cycle, not a calendar effect that can be assumed to persist.
- Utilities and staples lag in Q4 in about 65–70% of years, in both halves. The most robust line in the study.
- No positive sector edge is significant (max t 1.5). The last five Q4s each had a different relative winner;
SMH was strong in three of five (2021, 2023, 2025), XLB was the worst in 2024 (−14.7 pp).
Q4 excess over SPY, last five years (pp)
| XLK | XLB | XLI | XLF | XLY | XLU | XLP | XLE | XLV | QQQ | SMH | IWM |
|---|
| 2021 | +5.6 | +4.1 | −2.5 | −6.5 | +3.0 | +1.9 | +1.7 | −3.1 | +0.0 | +0.2 | +10.1 | −9.1 |
| 2022 | −2.5 | +7.4 | +11.6 | +5.8 | −16.6 | +0.9 | +5.0 | +15.1 | +5.1 | −7.7 | +3.3 | −1.3 |
| 2023 | +6.0 | −2.0 | +1.4 | +2.3 | −0.4 | −3.2 | −6.2 | −18.0 | −5.2 | +3.0 | +9.7 | +2.3 |
| 2024 | +0.7 | −14.7 | −4.7 | +4.6 | +9.7 | −8.0 | −7.1 | −4.1 | −12.8 | +2.4 | −3.4 | −2.2 |
| 2025 | −0.4 | −0.9 | −1.7 | −0.6 | −2.8 | −4.0 | −2.8 | −1.7 | +9.0 | −0.2 | +8.0 | −0.6 |
⚠ Limits. 27 observations per sector — a 2 pp average edge with a 10 pp standard deviation needs about
100 years to reach t = 2. Month-end closes only (no intra-quarter path, no costs). Sector composition drifts (XLC carved out of XLK/XLY in
2018, XLRE out of XLF in 2015). Calendar Q4 mixes rate cycles, election years and crashes (2008, 2018) with any true seasonal effect.
4. How to use it
- Tilt, not trade: it only adjusts what you already hold long in Q4 — lean semis/Nasdaq in October, lean industrials and
small caps in November–December.
- The avoid list is the actionable part: don't add XLU / XLP in Q4.
- Don't buy XLB for Q4: its lead is a 1999–2012 artefact.
- Treat SMH's recent Q4 strength as a cycle view (AI capex), sized by that thesis — not by the calendar.
Source notes: ~/research/seasonality.md; reproducible from ~/research/scripts/q4_sectors.py.