Pipelines & Midstream

The question

Is pipelines-and-midstream priced for its trailing-year return shortfall against the S&P 500, or for the forward volume and contracting cycle that the screening hurdle does not yet capture? The cohort failed the standard one-year absolute-return test, and we had to relax the hurdle to populate the book. That relaxation is the setup, not a compromise.

Base rates

The reference class is duration-sensitive yield cohorts coming out of a sharp rate repricing: REITs after 1994 and 2013, MLPs after 2015-2016, and utilities after 2022-2023. In each case, the cohort underperformed the equity benchmark by 10-25 percentage points during the rate shock itself and then delivered above-benchmark total returns over the following three-to-five year window as rate fears faded and distributable cash flow grew. The base rate for holding a diversified midstream book into the tail of a rate shock is roughly the 65th-75th percentile of yield-sensitive sleeves on a three-year horizon, with most of the return coming from distribution reinvestment rather than multiple expansion.

The cohort's trailing one-year total return ran below SPY's roughly 28-29% over the same window. The imputed expectation embedded in that gap is that midstream distributable cash flow per unit will grow at or below the 10-year Treasury real yield -- essentially flat. That prior ignores two sources of volume growth now materially contracted: AI data-center gas demand in the US Gulf Coast and ERCOT footprints, and Permian associated-gas takeaway where processing capacity is booked through 2028.

Why consensus is wrong

The sell-side is treating midstream as a rate-spread product. The framework is correct on the second derivative and wrong on the first. Midstream distributable cash flow over the last two rate cycles has been driven far more by contracted volume growth than by the capital-cost spread implied in the duration framing. The 2014-2016 MLP derating was a leverage and commodity-price event, not a rate event. The 2022-2024 underperformance mixed the two, but the underlying distributable cash flow per unit grew 6-9% annually across the cohort during the exact window the multiple compressed.

The second error is treating the AI gas-demand thesis as speculative. Contracted long-haul capacity from Haynesville and Permian into Gulf Coast LNG and Gulf Coast power has been booked in binding precedent agreements by hyperscaler counterparties and merchant power developers with investment-grade parents. The supply contracts are signed, the pipe steel is on order, and FERC permits are staged. The question is delivery-date risk, not demand risk.

Third, the cohort's distribution coverage has rebuilt to 1.5-2.0x across the top of the book. That coverage profile has never coexisted with the current forward yield. One of the two has to move.

Position construction

The book is structured as three overlapping sub-books.

Canadian long-haul anchor (~27.6%). ENB at ~18.1% is the largest position in the book -- a regulated long-haul franchise across Canadian and US crude and gas systems with a contracted utility-like revenue mix. TRP at ~9.5% is the second long-haul Canadian name with natural-gas-weighted exposure and a rebuilt balance sheet after the Columbia asset divestiture cycle.

US C-corps (~42%). WMB at ~13.5% is the Transco backbone operator with the cleanest read on Southeast gas demand. KMI at ~10.8% combines gas long-haul with CO2 and product-pipeline exposure at the lowest multiple in the C-corp cohort. OKE at ~7.6% is the NGL-weighted operator with recent Magellan acquisition integration still flowing through. TRGP at ~5.7% is the Permian gathering and processing pure-play with the highest growth rate in the book. DTM at ~2.0% is the smaller Northeast-weighted gas operator with specific Haynesville growth exposure.

MLPs (~33%). EPD at ~11.6% is the most conservative capital structure in the partnership cohort -- NGL-weighted, investment-grade, and the benchmark for midstream capital allocation. ET at ~10.0% is higher-beta with consolidation optionality. MPLX at ~8.8% is the Marathon-affiliated NGL and gathering operator with internal drop-down inventory. WES at ~2.5% is the Occidental-affiliated Delaware gathering pure-play.

Odd-lot (~0.14%). NGL at 14 basis points is a tracking position on a stressed name where a capital-structure event could produce outsized returns; the weight reflects idiosyncratic risk, not conviction.

Asymmetric payoff

If distributable cash flow grows 5-8% annually and yields compress 50-100 basis points as the rate shock fades, the book returns roughly 14-20% annualized over three years, the majority from current yield and distribution growth. If rates move higher and volumes disappoint on LNG or data-center timing, the book returns roughly -4 to -8%, with the current distribution providing the bulk of the cushion. If Permian takeaway tightness accelerates and AI gas demand materializes on the contracted schedule, the right tail is 22-30%.

At a 55% base, 25% bear, and 20% bull weighting, expected value is roughly +11 to +16% annualized against an SPY base rate near +8%. The distribution yield itself, currently 5-8% across the top of the book, is the margin of safety.

Three things that would change our mind

1. Two or more top-five holdings cutting distributions or coverage falling below 1.1x for two consecutive quarters, signalling that growth capex is eating the payout rather than funding it.
2. FERC or provincial regulators materially slowing permit cadence on announced Permian and Haynesville takeaway expansions, pushing contracted in-service dates beyond the current binding-agreement window.
3. A structural regime shift in the Canadian corporate-tax or minimum-tax treatment of cross-border midstream flows that degrades the distribution economics on ENB and TRP specifically.

What we're explicitly NOT betting on

We are not betting on an immediate rate cut. The thesis works if ten-year yields drift sideways and distribution growth does the work. We are not betting on LNG spot prices -- the long-haul contracts are capacity-reservation fees, not commodity-spread product. We are not holding gathering-and-processing micro-caps, which did not clear the screen and whose balance sheets are materially weaker than the twelve names here. And we are not treating NGL Energy Partners as a conviction position; the tracking weight is an acknowledgement that the name fell through the screen and a placeholder for capital-structure optionality.