Utilities (XLU) vs 10Y Treasury Yield

XLU traded $45.74 to $46.30 range on April 24, 2026. 10Y Treasury yield 4.31 percent (April 2026). XLU dividend yield approximately 3.0 percent.

Top holdings April 2026: NextEra Energy (NEE) 14.01 percent; Southern Company (SO) 7.23 percent; Duke Energy (DUK) 6.96 percent; Constellation Energy (CEG) 6.38 percent; American Electric Power (AEP) 5.11 percent. Combined top 5 approximately 40 percent of XLU.

XLU has rallied substantially in 2024-2026 era. NextEra +26 percent in last 6 months alone. Duke Energy committing $103 billion record capex on growth. Constellation Energy benefiting from data center power agreements with Microsoft, Amazon, Meta.

The combined April 2026 reading: XLU rallying despite elevated 10Y because AI data center electricity demand surge has overridden traditional duration drag. Configuration is unusual historically. XLU typically falls when 10Y rises. The 2024-2026 era saw XLU rally with elevated 10Y due to: (1) data center demand growth narrative; (2) electrification capex tailwind; (3) defensive characteristics during equity volatility.

Utilities are classic bond proxies with regulated returns and stable cash flows. Utilities operate under regulatory frameworks (state public utility commissions) that set allowed rates of return on equity (ROE) typically 9-10 percent. Predictable revenue + cost-recovery + ROE structure = bond-like cash flow profile.

Dividend yield: utilities typically yield 3-4 percent (XLU 3.0 percent April 2026). Compares to 10Y Treasury 4.31 percent. Yield gap (10Y minus utility) typically -0.5 to +1.0 percent. April 2026 yield gap +1.3 percent (10Y above utility yield).

Empirical sensitivity (pre-2024): 100 basis point rise in 10Y typically associated with 8-12 percent XLU decline (over 60-90 day windows). Compare to XLF (financials) +5-10 percent on same shock. XLU and XLF have inverse rate sensitivity profiles (XLU bond proxy; XLF NIM beneficiary).

The 2022 hiking cycle illustrated. 10Y rose 1.5 percent to 5.0 percent peak (October 2023). XLU fell 18 percent peak-to-trough. Classical duration framework worked.

The 2024-2026 era anomaly: data center demand catalyzed structural growth narrative for utilities. NextEra benefiting from AI-related capex. Constellation Energy nuclear power deals with hyperscalers. Duke Energy data center customer Amazon/Microsoft/Google/Meta. Electricity demand growth from 2 percent to 5 percent annual range fundamentally changed utility outlook.

AI data center electricity demand has fundamentally changed utility sector outlook. Data centers consumed 200 TWh in 2022 (~5 percent of US power consumption). Projected 260+ TWh by 2026 (~6 percent of US power). Some estimates suggest 400-600 TWh by 2030 (~8-10 percent of US power).

Key utility companies positioning for AI demand. NextEra Energy (14.01 percent of XLU): plans 15 GW power for data center hubs by 2035. Federal backing for 10 GW new natural gas power capacity. $90-100 billion clean energy investment through 2032. Stock +26 percent last 6 months on AI narrative.

Duke Energy (6.96 percent of XLU): announced industry record $103 billion capex on growth (April 25, 2026 announcement). Data center customers include Amazon, Microsoft, Google, Meta. Southern + Midwestern service area among fastest-growing.

Southern Company (7.23 percent of XLU): projects 8 percent annual demand growth through 2029 (data center load + residential).

Constellation Energy (6.38 percent of XLU): nuclear power leader. Multiple direct power purchase agreements with hyperscalers (Microsoft, Amazon, Meta) for data center supply. Stock +50 percent in 2024-2025 era.

American Electric Power (5.11 percent of XLU): also data center exposure plus electrification capex.

The practical implication: traditional utility-rate sensitivity has weakened due to growth narrative. XLU has decoupled from 10Y in 2024-2026.

XLU and 10Y traditionally inversely correlated. The 2024-2026 era has weakened the relationship.

Pre-2024 correlation: XLU and 10Y had -0.50 to -0.65 correlation. Strong inverse relationship. 100bp 10Y rise = 8-12 percent XLU decline.

2024-2026 correlation: weakened to -0.20 to -0.35. AI capex narrative overriding rate sensitivity. 100bp 10Y move now produces only 3-5 percent XLU response.

The practical implication: XLU has gained dual exposure to (1) defensive bond proxy characteristics; (2) growth tailwind from AI data center demand. Combined effect: XLU outperforms in both rate-falling regimes (bond proxy benefit) and growth-driven regimes (data center demand).

April 2026 setup: 10Y at 4.31 percent (elevated) + XLU at $46 (near 52-week highs). Configuration suggests data center narrative dominant. Watch for sustainable 10Y above 4.5 percent: would test whether duration framework reasserts. Sustainable 10Y below 3.8 percent: would catalyze additional XLU rally combining duration + growth tailwinds.

Three rate-cycle examples.

2018-2019: 10Y rose 2.4 percent to 3.2 percent (Q4 2018 peak), then fell to 1.5 percent (mid-2019). XLU fell 8 percent peak-to-trough Q4 2018, then rallied 27 percent through 2019. Pattern: classical bond proxy.

2020 COVID and recovery: 10Y fell 1.9 percent to 0.5 percent (March 2020) then rose. XLU fell 32 percent peak-to-trough COVID flash crash. Recovered with rate stability. Bond-proxy sensitivity dominated.

2022 hiking: 10Y rose 1.5 percent to 5.0 percent peak. XLU fell 18 percent peak-to-trough. Classical duration framework worked.

2024 AI awakening: 10Y elevated 4-5 percent. XLU rallied 30 percent (NEE +30%, CEG +50%, DUK +20%). Data center narrative emergence broke duration framework.

2025 continuation: 10Y stable. XLU continued rally. Constellation Energy nuclear deals with hyperscalers catalyzed sector.

2026 stabilization: 10Y 4.31 percent. XLU near $46 (52-week highs). NextEra +26 percent past 6 months.

The pattern: pre-2024 duration framework dominated. 2024-2026 AI data center demand narrative shifted regime. XLU now offers growth + defensive characteristics.

Stress history.

2008-09 GFC: 10Y fell 4.5 percent to 2.0 percent. XLU -42 percent peak-to-trough. Pattern: pure liquidity crisis hit even defensive utilities.

2011 European debt crisis: 10Y fell modestly. XLU rallied modestly. Pattern: defensive bid emerged.

2018 Q4 Fed pivot: 10Y rose 2.4 percent to 3.2 percent. XLU fell 8 percent. Pattern: classical duration framework.

2020 COVID flash crash: 10Y fell 1.9 percent to 0.5 percent. XLU fell 32 percent peak-to-trough. Both rates and risk-off hit XLU.

2022 hiking: 10Y rose 1.5 percent to 5.0 percent. XLU fell 18 percent. Duration framework.

2023 March SVB: 10Y briefly fell. XLU rose 4 percent. Defensive bid.

2024-2026 AI era: 10Y range-bound 3.6-5.0 percent. XLU rallied 30+ percent. Data center demand override.

2026 Iran war: 10Y stable. XLU near highs. Defensive plus growth characteristics.

The pattern: pre-2024, duration framework dominated. 2024-2026 era has AI data center demand catalyst providing growth narrative. XLU has dual character now (defensive + growth).

XLU realized volatility approximately 13-18 percent annualized. Lower than other equity sectors due to defensive bond-proxy characteristics. 60-day rolling correlation between XLU and 10Y: -0.20 to -0.35 in 2024-2026 (weakened from -0.50 to -0.65 pre-2024).

For pair-trade implementation: XLU exposure through XLU ETF (most liquid utility sector ETF, AUM approximately $20 billion) or VPU (Vanguard Utilities, broader). 10Y yield exposure through TLT (long Treasury) or 10Y futures (TY). Long XLU / long TLT captures defensive positioning. Long XLU / short TLT captures unique 2024-2026 setup (XLU outperforming despite higher yields).

The pair has produced cyclical returns. 2024-2026 long XLU / short TLT gained substantially (XLU +30%, TLT relatively flat). 2008-09 GFC long TLT / short XLU gained dramatically (XLU -42%, TLT +37%). 2022 hiking long short-duration / short XLU gained substantially (XLU -18%).

Most actionable: monitor data center demand growth signals. Hyperscaler capex announcements (MSFT, META, GOOGL, AMZN) drive utility narrative. NextEra Energy investor day announcements catalyze sector.

For macro allocators, XLU-vs-10Y provides utility cycle classification.

10Y rising + XLU rallying: AI data center narrative dominant (current 2024-2026). Long XLU benefits.

10Y rising + XLU falling: classical duration framework. Pre-2024 typical pattern. Long short-duration positioning.

10Y falling + XLU rallying: bond-proxy regime. Long XLU + long TLT.

10Y falling + XLU falling: rare. Often signals utility-specific concerns (nuclear safety, regulatory tightening).

April 2026 setup: 10Y 4.31 percent + XLU $46 (near 52-week highs). AI data center narrative dominant. Long XLU continues to work.

Key watches. Hyperscaler capex announcements (MSFT, META, GOOGL, AMZN earnings April-May 2026) for data center demand confirmation. NextEra Energy quarterly results (late April 2026). Constellation Energy nuclear deal announcements. Duke Energy capex execution (data center projects). 10Y trajectory for any breakdown above 4.5 percent.

For positioning: long XLU through XLU ETF or individual names (NEE, CEG, DUK). Long Constellation Energy specifically benefits from nuclear power data center deals. Hedge through long TLT for downside duration protection.

XLU/10Y captures bond-proxy + AI data center demand. Compared to other sector-vs-rates pairs.

Vs XLF/10Y: XLF positive rate sensitivity (NIM expansion). Opposite directional preference. April 2026 both XLF and XLU at highs (different drivers).

Vs XLK/10Y: XLK negative rate sensitivity (duration drag) but AI capex narrative override. Both XLK and XLU benefit from AI narrative through different channels.

Vs XLRE/10Y: XLRE negative rate sensitivity. Similar to XLU but with REIT-specific dynamics. XLRE has data center REITs (EQIX, DLR) but not as concentrated as XLU's utility names.

Vs XLE/10Y: XLE primarily commodity-driven. Modest rate sensitivity.

For allocator monitoring, XLU/10Y is foundational utility cycle indicator. April 2026 reading: 10Y 4.31 percent + XLU $46 (data center narrative dominant). Pair complements XLF/10Y (NIM beneficiary), XLK/10Y (AI capex), XLRE/10Y (REIT duration), KRE/10Y (regional banks) for comprehensive sector rate-sensitivity read.

XLU $46 (April 24 2026), 10Y yield 4.31 percent. NextEra 14.01 percent of XLU (+26 percent past 6 months); Southern 7.23 percent (8 percent annual demand growth through 2029); Duke 6.96 percent ($103B record capex April 25 2026); Constellation 6.38 percent (nuclear data center deals). Data centers 200 TWh (2022) projected 260+ TWh (2026).

Forward-looking through 2026: hyperscaler capex announcements remain key. Microsoft, Meta, Google, Amazon Q1 2026 earnings (April-May 2026) will signal capex trajectory. NextEra investor day announcements. Constellation Energy nuclear power purchase agreement details. Duke Energy capex execution.

Key watches: 10Y trajectory (sustained above 4.5 percent would test duration framework reassertion). FOMC May 6-7 for rate path. Hyperscaler capex guidance. Nuclear regulatory developments.

Key risks: AI capex disappointment (would compress data center demand growth narrative); nuclear safety incident; regulatory tightening on utility ROEs; 10Y collapse below 3.5 percent (would catalyze bond proxy + growth combination); 10Y surge above 5 percent (would test duration override).

Expected XLU range $44-$48 absent major catalyst. 10Y range 4.0-4.5 percent. Configuration suggests continued utility outperformance with both bond-proxy and growth narrative benefits.