Damage Math

The units fell 23.9% from January to October 2025, but the cash the business generates never fell with them. Adjusted EBITDA attributable to PAA rose to $2,833M in 2025 and 2026 guidance was later raised to $2,880M — above 2025 — even after selling the NGL business; the distribution was increased 10% and forward estimates climbed. The only durable damage, a roughly $235M/yr Permian recontracting reset, carries an NPV well below the price drop. At $26.33 the units sit above the pre-drawdown peak.

The near-term hit — the numerator that is mostly missing

Ruchir's dislocation pattern needs a real near-term earnings cut that the market then extrapolates. Here the numerator is close to absent. Full-year adjusted EBITDA attributable to PAA rose from $2,779M in 2024 to $2,833M in 2025 [1], landing 1.1% above the $2,802.9M CapIQ consensus and only 1.5% under the $2,875M midpoint of the guidance range set in February 2025 [2]. The February 2026 outlook then stepped guidance down 2.9% to $2,750M on the NGL divestiture [3] — and by May 2026 that guide had been raised by $130M to $2,880M, above the 2025 actual [4].

Sources: 2024 and 2025 actuals and 2026 guidance, Plains All American investor updates [5] [6]; FY2027–FY2029 are CapIQ consensus mean EBITDA (data/sp/estimates.json).

The quarterly print was noisier than the full year. Normalized EPS missed consensus by 51.9% in Q1 2025 and 21.7% in Q4 2025 — real misses on commodity and timing swings — but Q3 2025 beat by 107.4% and Q1 2026 by 65.8% (CapIQ estimates). The signal that a permanent forecasting error would leave — a durable downward revision to forward cash flow — did not appear. Consensus normalized EPS for FY2027 rose from $1.77 six months ago to $2.19 today, and FY2028 from $1.885 to $2.35; consensus free cash flow climbs from $1,580M in FY2026 to $2,043M in FY2029 (CapIQ estimates).

Source: CapIQ consensus momentum, 180-day (2026-02-03) vs current (2026-08-02) (data/sp/estimates.json).

Capital return moved the way it moves for a business that is fine, not impaired: the distribution was raised 15 cents (10%) to $1.67 per unit annualized, the coverage threshold was lowered from 160% to 150% on "improved visibility," and the special distribution once earmarked for NGL-sale taxes was cancelled [7] [8].

The price move — the denominator that did all the work

The units peaked at $22.13 on January 30, 2025, troughed at $16.85 on October 10, 2025, and closed at $26.33 on July 31, 2026 — a 23.9% peak-to-trough drawdown, per the deterministic capitulation gauge, that has since reversed to sit 19.0% above the pre-drawdown peak.

Source: derived from daily prices; drawdown gauge in data/ruchir/fit_features.json (capitulation_gauge).

Market cap is not in the feature file — the run ships no share-count series, so fit_features.market_cap is not_computable. Reconstructing it from the 197.9 million Class A shares in the FY2025 10-K [9] puts PAGP's traded equity at roughly $4.38B at the peak, $3.33B at the trough, and $5.21B now. The peak-to-trough decline was about $1.04B of PAGP equity value.

One scaling note carries through the rest of this tab. PAGP is a look-through holder: it owns about 85% of AAP, which holds 233.0 million PAA common units, so PAGP's traded market cap represents an approximately 28% economic interest (198M of roughly 700M PAA common units) in the operating partnership [10]. The full enterprise carries roughly $11.3B of consolidated PAA face debt at year-end 2025 (about $7.7B net of cash and pro forma for the NGL-sale paydown) [11]. All the guidance drags and offsets below are stated "attributable to PAA," so the arithmetic is run at that enterprise level and then scaled to PAGP's ~28% slice.

The NPV arithmetic — two scenarios, workings shown

The one item that plausibly reduces long-run earning power is the Permian long-haul recontracting reset: as legacy Cactus I/II and Sunrise contracts rolled off, rates reset lower. Only part of it is separately quantified: the 2024-to-2025 walk shows a "lower contracted rates on certain Permian long-haul pipes" driver that Plains did not size [12] — that year the net crude contribution was a positive ~$135M, so the reset was one negative leg inside a segment that still grew — while the 2025-to-2026 walk quantifies a $100M "Cactus I Re-contracting" step [13]. Taking the durable reset at a deliberately generous $235M/yr — the disclosed $100M plus a comparably sized allowance for the earlier, unquantified step — sets the damage ceiling used below. The NGL divestiture's $365M of lost EBITDA is a different thing: it was a sale for CAD$5.15B (about US$3.75B, roughly US$3.0B net) at about 8.5x 2025 EBITDA, with proceeds redeployed into Cactus III and debt paydown [14] — a swap of cash flow for cash, not value destroyed.

Capitalizing the $235M reset under a conservative discount rate gives the plausible ceiling on permanent damage; running it as a level shift that reverses over five years gives the temporary reading. Ruchir's 10% yield bar is used as the discount rate, with 8% shown for sensitivity.

Source: derived — perpetuity ($235M ÷ r) and 5-year annuity of the $235M reset from the 2025 and 2026 guidance bridges [15] [16]; PAGP column scales by the ~28% look-through interest.

Two facts bound the read. First, in the very same 2026 bridge the reset was more than replaced by $380M of contracted growth and self-help — Cactus III +$250M, efficiency +$50M, optimization +$50M, and FERC recoup and tariff escalators +$30M — so run-rate EBITDA rose rather than fell [17]. Net of those offsets, the damage to intrinsic value rounds to zero. Second, even taking the reset gross and permanent, its NPV attributable to PAGP is about $660M (about $825M at 8%) against the roughly $1.04B PAGP equity lost at the trough.

The gap — and its disappearance

At the October 2025 trough there was a gap of the kind Ruchir hunts. PAGP equity had fallen about $1.04B; the plausible permanent damage attributable to PAGP was about $0.66B, and the temporary reading about $0.25B — so the price fell roughly $0.4B more than the harshest damage estimate and about four times the temporary one. Net of the contracted offsets, essentially the entire $1.04B was a re-rating gap rather than a fundamentals gap.

That gap has closed. At $26.33 the units sit $0.84B above the pre-drawdown peak in market-cap terms and $1.88B above the trough; distributable cash flow, the distribution, and forward estimates are all higher than before the shock. The price damage plainly exceeded the plausible NPV damage — but the correction has already been paid to whoever bought the dislocation, not to a buyer today. Worth adding for the framework's own gate: the drawdown was 23.9% on a volume spike of only 1.6x, not the 60–70% capitulation on emotion-driven volume the pattern is built around — the anatomy is covered in Dislocation.

The trial — temporary versus permanent

The temporary-or-permanent question was argued by two opposing corpus-cited briefs and ruled on by three blind judges. Both cases are strong on the record.

The temporary case. Crude Oil segment adjusted EBITDA rose every year through the episode and the FY2025 10-K states it in words — "Crude Oil Segment Adjusted EBITDA increased for the year ended December 31, 2025 compared to … 2024" [18]. The oil-price exposure that drove the selling is small: a $10/bbl move in WTI is worth only about $40M of EBITDA [19]. Guidance was raised, not cut [20], and management framed 2026 as roughly 13% crude-segment growth with volume growth expected to resume in 2027 as gas-takeaway egress constraints clear [21].

The permanent case. The event removed a real operating leg: Plains sold substantially all of its Canadian NGL business, a "strategic shift" that carried $383M of after-tax income from discontinued operations in 2025 [22]. The remaining crude business took a genuine price reset, not a volume dip — "all those volumes have been re-contracted" at lower rates [23] — and the 10-K roots that reset in industry structure, citing "downward pressure on tariffs and margins" [24] in Permian markets that "have become overbuilt" [25]. On this reading the future case rests on different assets — Cactus III, lower taxes, lower maintenance capital — while the old NGL earnings and above-market tariffs are gone.

The judges put the probability the impairment is temporary at 0.71, with the three seats at 0.71, 0.68, and 0.78 (range 0.68–0.78, spread 0.10, mean 0.72). The result is not contested, and it was stable to reading order (temporary-first mean 0.71 versus permanent-first 0.73, a 0.02 gap). This diagnosis probability is the trial's, not this tab's; the arithmetic above is consistent with it but does not set it.

Which line broke, and whether it self-corrects

The line that broke is Permian long-haul crude tariff pricing. As the basin's egress capacity was built out, legacy above-market contracts (Cactus I/II, Sunrise) rolled to lower market rates — the $235M cumulative reset above, disclosed as early as the Q1 2024 call, which flagged Cactus I recontracting at $1.25–$1.50/bbl effective September 2025 [26]. The self-correction mechanism is specific: Plains bought the competing long-haul capacity — the EPIC/Cactus III interests for a $2.9B combined valuation at about 10x forward EBITDA — turning the source of the rate pressure into roughly $250M of its own EBITDA plus $50M of synergies [27]. Post-close leverage is guided back toward the low end of the 3.25x–3.75x range (about 3.5x), and roughly $1,850M of adjusted free cash flow is expected in 2026 excluding the sale proceeds [28].

The structural argument against self-correction is on the record too, and is not softened: the 10-K's own language that Permian markets have become overbuilt, with downward pressure on tariffs and margins, is what would make the reset a recurring repricing cycle rather than a one-time step [29] [30]. What would decide it: whether a further long-haul recontracting drag above $100M appears in the FY2027 guidance bridge, and whether the crude segment grows off the 2026 base without new acquisitions — the tests carried in Durability and Clock.