2 Midstream Dividend Stocks With Growing Payouts -- One Yielding Over 6%

Energy Transfer and Enterprise Products Partners are highlighted as high-yield midstream beneficiaries of rising natural-gas demand tied to AI data-center power needs. Energy Transfer plans up to $5.9B of 2026 growth capex, targets 5-6x EBITDA build multiples, and offers a 6.4% distribution yield with 2.2x coverage; Enterprise plans up to $4B of spending, has $6.5B under construction, and yields 5.8%. Enterprise expects double-digit EBITDA and distributable-cash-flow growth in 2027 as new projects ramp, although 2026 is characterized as a transition year.

Analysis

The investable issue is not headline AI power demand but whether incremental Permian and Haynesville volumes clear existing bottlenecks at contracted returns. ET has greater upside torque because its gathering, processing and takeaway footprint can monetize both volume growth and episodic basis dislocations; that same optionality makes cash flow less predictable than its fee-revenue mix suggests. EPD's NGL/export system is a cleaner duration asset, but its nearer-term earnings bridge depends more on petrochemical utilization, LPG export economics and the timing of project ramp-ups than on data-center demand.

Consensus is likely over-attributing the next 12 months of midstream EBITDA to AI. Utility interconnection queues, gas-fired generation permitting, turbine availability and transmission buildout mean data-center load translates into physical gas demand with a multi-year lag; existing industrial, LNG-export and associated-gas trends remain the more immediate drivers. The better second-order beneficiary could be WMB, whose interstate gas exposure is more directly tied to power-demand regions, while KMI offers a lower-beta alternative if investors want gas-pipeline exposure without NGL and export-cycle sensitivity.

Over 1-3 months, the catalyst path is contract announcements, in-service execution and forward capex guidance rather than distribution yields. Over 6-18 months, sustained pipeline utilization and no material cost inflation would justify multiple expansion; conversely, weak producer activity or a collapse in Permian gas differentials would expose projects that are economically sound only at high throughput. For both names, distribution coverage should be treated as a downside cushion, not the valuation thesis: rising long-end yields can compress MLP multiples even if operating results meet plan.

Key Decisions for Investors

• Initiate a 6-12 month long ET / short EPD pair at roughly equal dollar beta only if ET underperforms EPD by 5% or more from the current ratio. Thesis: ET has greater incremental EBITDA sensitivity to Permian takeaway utilization and basis volatility; target 8-12% relative return. Exit if ET's project cost guidance rises materially or if contracted volumes fail to support 2027 EBITDA expectations.
• For lower-volatility exposure, accumulate EPD on a 5-7% pullback over the next 1-3 months rather than chase yield. Underwrite total return from distribution growth plus project ramp, not a rapid AI re-rating; reassess if 2027 EBITDA/DCF guidance is not reaffirmed as construction milestones progress.
• Add WMB to the watchlist as the more direct gas-to-power proxy. Upgrade to a long only after management identifies contracted, creditworthy power-demand or LNG-linked expansions; absent disclosed contract terms and capital intensity, treat AI-related claims as narrative rather than earnings.
• Hedge broad rate sensitivity with a modest short in XLU or long-duration Treasury puts against a basket of ET/EPD/WMB if the 10-year Treasury yield rises 40-50 bp. MLP yield buyers are vulnerable to duration-driven multiple compression even where distributions remain covered.