1 High-Yield Pipeline Stock Investors Keep Underestimating

ONEOK (OKE) is up 30.6% YTD and, after its Q2 update, raised 2026 EPS and net income outlooks, signaling improving fundamentals. The company also secured a data center-related deal to deliver gas to a 1-gigawatt data center power plant, supporting a longer-term cash-flow tailwind tied to AI-driven demand. While the 4.5% dividend yield is only mid-pack within midstream, a 4% payout increase earlier this year aligns with its targeted 3%-4% annual dividend growth.

Analysis

The market is starting to treat OKE less like a static toll road and more like a leveraged call on basin growth plus power-demand adjacency. The important second-order effect is that any sustained AI/data-center buildout increases the value of gas transport and processing capacity before it shows up in headline EBITDA, because shippers and utilities compete for scarce pipe, compression, and NGL takeaway. That favors incumbents with existing footprints such as OKE and, in a broader basket, WMB and KMI; it is less helpful for higher-yield, slower-growth midstream names where the incremental capacity story is weaker.

The near-term risk is that investors are capitalizing a still-uncertain project pipeline as if it were already contracted cash flow. Data-center related deals often move from announcement to meaningful throughput over quarters, not weeks, so the stock can outrun the fundamental bridge if volumes or FID timing slip. If the next two quarters do not show clearer throughput growth or if the company’s 2026 outlook stops ratcheting higher, the AI premium can compress quickly.

Contrarianly, the move may be more about multiple re-rating than earnings durability. At a 4%+ yield with a sharp YTD rerating, upside now depends on proving that growth is structurally above the historical midstream pace, not just one-off guidance beats. That makes this a better relative-value story than a blind outright long: the thesis is strongest if fee-based volume growth and gas-for-power demand both persist through 2026, weaker if the AI narrative becomes the only justification.

Key Decisions for Investors

• Long OKE vs. short XLE for 1-3 months: expresses the view that gas infrastructure can outperform broader energy if AI/power demand stays a secular tailwind, with less commodity beta than an outright energy long.
• Within midstream, prefer OKE over higher-yield, slower-growth peers for a 6-12 month hold only if the next earnings print confirms continued 2026 guide upside; otherwise fade the rerating and rotate into names with more contracted growth visibility.
• Watchlist alert: if OKE does not show clear volume/throughput improvement on the next two reporting dates, reduce exposure; that would falsify the idea that data-center demand is already monetizing.
• For risk-managed exposure, use call spreads rather than stock if chasing the AI/power narrative; the upside is real but the stock has already re-rated, so the cleaner trade is defined-risk participation rather than chasing a further multiple expansion.