PAGPNASDAQThe short version
Plains GP Holdings, L.P.
Plains GP Holdings is a Nasdaq-listed partnership whose sole asset is a roughly 31% look-through stake in Plains All American Pipeline, a large US crude-oil gathering, pipeline and storage system anchored in the Permian Basin.
The shares collapsed 86% into the 2020 crude crash and slid another 24% in 2025; both round-trips are complete, and at $26.33 the stock sits above its pre-2025 peak.
Net debt $927.0MP/E FY27E 12.0×ND/EBITDA FY27E 0.3×
$26.33
Share price
$5.2B
Class A market cap
4.7%
Adjusted FCF yield
3.9×
Net debt / EBITDA
SwipeScroll▾
IThe business
What it is
A Nasdaq ticker wrapped around a Permian crude-pipeline system
9.7M b/d
Crude oil moved in 2025
20,405
Miles of pipeline
~31%
Look-through stake in PAA
28 yrs
Operating history at PAA
- It owns no assets. PAGP holds only an indirect ~31% look-through interest in Plains All American Pipeline; every dollar it earns comes from PAA's distributions.
- How it makes money. PAA buys crude at the wellhead, moves it across 20,405 miles of pipe to hubs and export docks, and sells it there — earning tariffs and fees on the assets plus a margin on the barrels.
- Why two tickers. PAGP is taxed as a corporation and issues a 1099 instead of a K-1, so it trades at roughly a 7% premium to PAA on an otherwise identical claim.
Where it earns
Revenue is 96% pass-through crude; the real earnings are ~$2.8B of pipeline EBITDA
Crude Oil Segment Adjusted EBITDA
- Revenue is a price mirror. Of $44.3B in FY2025 sales, $42.4B is crude bought and resold at the WTI price; only ~$1.8B is tariff and fee income. The top line moves with oil, not the business.
- The earnings that matter. Adjusted EBITDA attributable to PAA was $2,833M in 2025 and is guided to $2,880M for 2026 — a steady, capital-heavy toll business under the commodity gross-up.
- One big customer. ExxonMobil is ~31% of revenue, up from 15% in 2021 — mostly low-margin merchant volume, but a concentration that has doubled in four years.
IIIThe story now
The fit
Outside the framework's universe (U2 not met); contested: P2
The pillar ledger
| Test | Where PAGP sits |
|---|---|
| Market cap | $5.2B vs the $10B line |
| Adj. FCF yield | 4.7% vs the 25% bar |
| Diagnosis | Temporary (0.71) |
| Confidence | High |
- The gate that closes it. 197.9M Class A shares at $26.33 is $5.21B — 52% of the $10B floor; even counting the exchangeable Legacy interests it is $6.13B, still $3.9B short. The framework never reaches the pillars.
- The counter-fact, in the same breath. The business underneath clears the line easily — PAA's own units are worth $17.3B and PAGP consolidates $31.3B of assets — but the listed instrument a Class A buyer owns does not.
- Contested: P2. The rolling five-year adjusted-FCF test can't be computed here (stock comp is null every year), so two jurors read consistency met and two returned cannot-determine.
The dislocation
A real 24% drawdown in 2025 — shallow, and already fully retraced
−23.9%
Peak-to-trough, 2025$22.13 → $16.85
1.59×
Peak volume vs normalno 60–70% capitulation
+56%
Rebound off the troughnow above the pre-fall peak
386 days
Round trip
- One external trigger. The April 2025 tariff-and-OPEC+ crude shock drove 83% of the fall in four sessions; company EBITDA guidance barely moved and FY2025 landed just 1.5% below the pre-fall midpoint.
- Not capitulation. Volume peaked at 1.59× normal — and six months before the low, which itself printed on below-average turnover. The framework wants peak fear at the bottom; this wasn't it.
- What would bring it back. A crude shock three times April's size, or a Permian volume break, could open the entry that isn't on offer today.
Damage math
The price fell about $1.0B; the plausible lasting damage was ~$0.66B
PAGP value: price drop vs plausible damage
- The numerator is nearly missing. Through the whole episode Adjusted EBITDA rose ($2,779M → $2,833M), 2026 guidance was later raised to $2,880M, and the distribution went up 10%.
- The one durable drag. A ~$235M/yr Permian recontracting reset caps permanent damage at ~$0.66B to PAGP — but the 2026 bridge more than replaces it with $250M of Cactus III EBITDA plus self-help.
- The gap has closed. A blind trial ruled the impairment temporary at 0.71; at $26.33 the units sit above the pre-drawdown peak, so the correction was paid to whoever bought the low.
The year-10 gate
Volumes up 46%, but earnings per barrel down 42% — the gate doesn't clear
Crude segment Adj. EBITDA per tariff barrel
- Revenue is a WTI forecast. 96% of the top line is crude resold at market, so "year-10 revenue higher" is a bet on the 2036 oil price, not on earning power.
- Per-barrel economics keep compressing. Down 42% since 2019 even as volumes rose; the 10-K has flagged Permian overbuild "downward pressure on tariffs and margins" for five straight years.
- The counter-fact. Segment EBITDA has still risen every year and the contracted backlog is 76% larger than in 2021 — volume and deals have outrun the squeeze, but "so far" can't clear a very-high-conviction gate.
The flywheel
The buyback engine the setup needs isn't running: $8M in three years
$8M
Buybacks, FY23–25vs $2,709M of distributions
+1.1%
Unit count, 3 yearsrising on ~$50M/yr stock comp
4th of 4
Repurchases in capital prioritiesbehind debt, distribution, deals
~21 yrs
To retire the float from FCFabsurdity marker is ~3
- Debt first, buybacks last. Management ranks repurchases fourth of four uses of cash and gates them on leverage falling below 3.25×; $190M of a 2020 authorization still sits unused.
- Share count is rising. Units are up 1.07% over three years on ~$50M/yr of equity comp with no offset — the framework's rising-share-count condition, tripped.
- The fair counter. The balance sheet can outlast the problem — investment-grade, $2.0B liquidity, no note due 2027–28, $3.3B of NGL-sale cash in May 2026 — and the few units bought went at $8–$10, well below today.
The clock
Every catalyst management named has already fired
The self-help sequence, and where it stands
| Mechanism | Status |
|---|---|
| NGL sale, $3.3B net | Done |
| EPIC / Cactus III, $2.9B | Done |
| 2026 guidance +$130M | Done |
| Distribution +10% | Done |
| Cost program, $100M | Half |
| Leverage to 3.25–3.75× | In motion |
- The re-rating already ran. The shares are 56% above the 2025 low, 19% above the pre-fall peak, and 6.8% above the street's mean target — the gap the clock exists to time is closed.
- History says deep is slow. Across 12 drawdowns since 2013, every round-trip inside two years came from an episode shallower than 40%; the two 60–70% capitulations took 26.5 and 82.6 months, or never recovered.
- What's left is delivery. Roughly $50M of 2027 cost savings and a leverage move three-quarters done — increments on a delivered plan, not value the market has refused to see.
IVThe price
Yield vs the bar
A 4.7% adjusted yield against a 25% levered bar
Adjusted FCF yield vs the framework's levered bar
Current
4.7%
3-yr average
6.2%
Best-case basis
7.3%
Consensus FY29
7.5%
- After the claims ahead of it. FY2025 adjusted FCF of $1,537M falls to $865M once preferred and JV-minority distributions are paid — $1.23 per look-through unit, 4.7% at $26.33.
- Levered, so the bar is high. Net debt of $10.9B is 3.9× EBITDA, which sets the framework's 25% reference line; the yield sits ~2,000 bps below it, and below the 10% and 8–9% lines too.
- Nothing to mean-revert. Consensus free cash flow rises every year (4.97% → 7.46%, FY26–29) and guidance was raised — there is no cut to buy back. Even the best-case 7.3% stays far short.
The re-rating math
What the yield bars need is $1.9B–$4.7B; only $0.9B reaches the common
Adjusted FCF to common: actual vs what the bars need
- The math is one-directional. Hitting the 10% line needs $1.86B to the common — 58% above the FY2029 consensus — and the 25% bar needs $4.65B, more than PAA's entire consolidated EBITDA.
- No depressed base. With cash flow held flat, reaching the 10% line means a price near $12.30 — a 53% fall from $26.33. The growth that would lift FCF is already inside the rising consensus.
- What you get instead. An investment-grade midstream paying a covered 6.3% distribution — a real thing on offer, but not the buyback flywheel the setup is built to exploit.
The instruments
Long-dated options exist, but the case never reaches them
Jan 2028
Longest listed expiry17.6 months out, just short of 18
19.7%
30-day implied volatilityframework line is ~50–55
$24.64
Street mean target6.8% below the $26.33 close
- Qualifying LEAPS are on the board. PAGP options run to January 2028, with open interest behind them — so this is not a watchlist-only name for lack of instruments.
- Volatility is calm. 30-day implied volatility was 19.7% on 31 July 2026, far below the framework's reference lines — but the option data is web-sourced, with no filing to cite, so it is recorded as not verifiable.
- The street has already moved. Six buys, six holds and two underperforms, with a $24.64 mean target below the market — a name the sell side understands, not one it has given up on.
What to watch
A hard-to-replace crude network — but a $5B listed stub, a 4.7% yield, and a re-rating already paid
- 01Adjusted EBITDA attributable to PAA comes in below $2.805B
- 02A new Permian long-haul recontracting drag greater than $100M
- 03Capital allocation pivots to debt paydown over repurchases
- 04Share count inflects upward
This distills a fixed fit test built tab by tab; the full report shows every number and its source.
Compiled from the full report · 2026-08-03 · For information, not investment advice.