3 High-Yield Pipeline Stocks to Buy Now and Hold Forever

Enterprise Products, Enbridge, and Energy Transfer are benefiting from AI-driven natural gas demand for data centers, with all three stocks up at least 19% year to date. The article highlights strong dividend support: EPD yields about 5.58% with 28 straight years of dividend increases, ENB yields 4.87% with 31 consecutive years of increases, and ET yields about 6.6% after raising its quarterly distribution more than 3% to $0.3375. First-quarter results were solid, including EPD DCF up 34.5% to $2.7B and ET revenue up 32% to $27.7B, while ET is singled out as the best current buy.

Analysis

The key signal is not just “AI needs power,” but that the incremental load is arriving into a constrained grid, which forces hyperscalers to contract for firm gas-linked infrastructure rather than wait for transmission buildout. That shifts midstream from a slow-growth yield sector to a quasi-utility toll collector with visible multi-year volume growth, and it should tighten valuation dispersion inside the group toward the names with the best asset adjacency to gas-fired generation, storage, and NGL takeaway.

Energy Transfer looks positioned to capture the most second-order upside because its growth vector is more embedded in new capacity and terminal throughput rather than just harvesting existing cash flows. The market is likely underappreciating that AI-related demand can lengthen the duration of the current midstream capex cycle, allowing reinvestment at attractive returns while still supporting distributions. That matters because a few points of incremental DCF growth can rerate the stock materially when it already screens as the highest yield.

The contrarian risk is that the current trade is being treated too much like a one-way “AI beneficiary” basket while midstream remains indirectly exposed to upstream discipline. If crude softens for a prolonged period, the volume story will lag by a couple of quarters before showing up in throughput, and the market could de-rate the group faster than fundamentals roll over. The other overhang is execution: the AI power thesis is real, but the timeline from demand announcement to contracted pipe/terminal volumes is measured in months to years, not weeks, so near-term price action can outrun cash flow.

Consensus may also be missing that the most durable winners are not necessarily the highest-quality balance sheets, but the assets with the most optionality to serve power generation, exports, and industrial demand simultaneously. That favors names with underappreciated network effects and terminal connectivity, while leaving more conservative operators looking expensive relative to growth. The trade is therefore less about owning “energy” and more about owning bottleneck infrastructure tied to an electrification constraint.

Key Decisions for Investors

• Long ET vs ENB on a 3-6 month horizon: ET offers higher growth optionality and a better entry point if AI-linked volumes continue to compound; use ENB as the lower-beta funding leg. Risk: if the market rotates back to pure yield defensives, the spread can compress.
• Add to EPD on pullbacks only, with a 6-12 month hold: this is the cleanest “quality compounder” in the group, but the multiple already reflects safety. Expect modest upside with lower drawdown risk; best used as ballast rather than a chase trade.
• Buy ET Jan-2027 calls or call spreads to express the multi-year AI infrastructure thesis with defined downside. This captures rerating potential if volume growth and project execution sustain; avoid outright common if you want to limit balance-sheet and execution risk.
• Pair trade: long ET / short XLU for 3-6 months as a way to monetize the market’s underpricing of gas infrastructure versus regulated utility beta. The thesis works if power constraints keep shifting capex toward contracted midstream assets rather than traditional utilities.
• Set a tactical stop on all three if upstream pricing weakens sharply for more than one quarter, because throughput sensitivity will show up with a lag. If crude breaks materially lower, trim midstream exposure before consensus revises volume expectations.