Dislocation

What happened to the price

Plains GP Holdings fell 23.9% from a 30 January 2025 close of $22.13 to a 10 October 2025 close of $16.85, and has since recovered to $26.33 — above the pre-fall peak and the highest close since August 2018. The fall had one dated trigger: the 3–8 April 2025 tariff-and-OPEC+ crude shock, four sessions that carried 83% of the peak-to-trough decline. Traded volume peaked at 1.59x the pre-peak median. There is no live dislocation here.

The drawdown, quantified

Source: derived from the run's daily price file; figures as recorded in the deterministic feature file (fit_features.capitulation_gauge.drawdown).

The measured depth is 23.9%, against the framework's reference case of a 60–70% decline driven by forced selling. The current price sits 56.3% above the trough and 19.0% above the peak the drawdown began from — 26.33 / 16.85 and 26.33 / 22.13 respectively. Consensus mean target price on the same date is $24.64, so the shares change hands 6.9% above where the sell side marks them.

Source: month-end closes derived from the run's daily price file; the intramonth peak ($22.13, 30 Jan 2025) and trough ($16.85, 10 Oct 2025) are not month-end prints and sit outside the plotted line.

Four legs, not one

The 253 calendar days between peak and trough contain four distinguishable stretches, and only one of them is an event.

Source: derived from the run's daily price file; volume multiples measured against the 1,312,023-share median daily volume over the 180 days before the 30 January 2025 peak, the same denominator the feature file uses.

The first stretch is drift: 43 sessions, a 2.3% decline, volume 1.14x the pre-peak median. On the framework's own test — down 10–20% with no event and normal volume is not the moment — that stretch does not qualify as anything.

The 3–8 April window is the event. Four sessions took the shares from $21.61 to $17.25, a 20.2% fall that accounts for $4.36 of the $5.28 total peak-to-trough decline, or 83% of it. The single worst session, 4 April, was -9.6% on 4.1x median volume.

The 19 June to 10 October grind is the other 14.7%, spread over 79 sessions at 1.02x median volume — the shares were changing hands at an ordinary rate the whole way down.

The trigger

The dated adverse event is external and industry-wide, not company-specific. On 2 April 2025 the US administration announced its reciprocal-tariff schedule; on 3 April eight OPEC+ members announced they would advance voluntary-cut unwinds, adding 411,000 barrels a day in May — three monthly increments at once. WTI spot fell 15% from $72.12 on 2 April to $61.05 on 7 April (US Energy Information Administration weekly spot series, reported by Hart Energy and S&P Global Commodity Insights). No Plains filing or release carries an 8-K, guidance revision or press statement dated inside that window.

Management named the same two mechanisms on the next earnings call, 9 May 2025: "The ongoing uncertainty on trade tariffs is weighing on economic forecasts and creating significant volatility. Additionally, the dissension among OPEC members and the prospects of incremental supply coming to market have resulted in a lower price commodity than anticipated at the beginning of the year" [1]. The FY2025 10-K frames the same supply picture structurally, describing a market that is well supplied near-term with uncertainty around OPEC's ability to keep raising production [2].

What the company did not do is cut the number. February 2025 guidance was Adjusted EBITDA attributable to PAA of $2,800–$2,950 million with adjusted free cash flow of $1,150 million [3]. On 9 May the range was left intact, qualified only by "Lower half of guidance range in a $60 - $65/bbl WTI environment", with adjusted free cash flow trimmed to $1,095 million [4]. In August the range was again unchanged, the qualifier restated, and adjusted free cash flow cut to $870 million [5]. In November the range was narrowed to $2,840–$2,890 on a $65/bbl WTI assumption, and the free-cash-flow line swung to negative $900 million [6]. That swing is capital deployment — the $2.9 billion EPIC Crude purchase closed 31 October and 1 November 2025 — not an operating shortfall; the EBITDA line moved by $10 million of midpoint.

Sources: February 2025 deck [3]; November 2025 deck [6]; FY2025 actual per the February 2026 deck [7]. Price column is the closing price on each release date, derived from the run's daily price file.

The delivered figure was $2,833 million of FY2025 Adjusted EBITDA attributable to PAA [7] — 1.5% below the midpoint guided before the fall began, and inside the range set in February 2025. Over the same span the shares fell 23.9%.

The one company-specific event of size inside the window was a disposal, not a disappointment: on 17 June 2025 Plains agreed to sell substantially all of its Canadian NGL business to Keyera for C$5.15 billion (US$3.75 billion) [8]. The shares rose 4.2% the following session on 2.3x median volume.

Earnings dates themselves moved the price very little. The largest reaction to any of the six results releases in the window was -2.7%, on 9 May 2025 — the day a quarter that had missed consensus normalized EPS by 51.9% was reported.

The fear gauge

Source: fit_features.capitulation_gauge.volume_spike for the 1.59x figure (max 20-day average volume in the peak-to-trough leg divided by the median daily volume over the 180 days before the peak); the other two multiples derived from the same daily price file on the same denominator.

The measured spike multiple is 1.59x — the highest 20-day average volume anywhere between peak and trough was 2,090,970 shares, against a pre-peak median of 1,312,023. That window ends 24 April 2025, so the volume peak sits in the April event leg, roughly two weeks after the price low of that leg and nearly six months before the eventual trough.

Source: derived from the run's daily price file; denominator is the 180-day pre-peak median used by fit_features.capitulation_gauge.

Two features of that series bear on the capitulation question. First, the highest monthly volume readings in the whole 23-month window are not in the fall at all — February 2026 (1.63x) and January 2025 (1.57x) sit above April 2025 (1.54x), and the February 2026 reading came while the shares were rising. Second, the 20-day average volume into the 10 October 2025 trough was 0.86x the pre-peak median: the low was set on below-normal turnover.

The behaviour that reads as emotion is confined to four sessions in April, at 3.02x median volume, and it was a whole-market energy event rather than a Plains event. The rest of the decline was orderly. Against the framework's requirement of a traded-volume spike marking peak fear, 1.59x is the measured number, and it does not sit at the low.

Who was selling

Reported short interest is unavailable for this security in the run's data: FINRA returned no position rows and no short-sale volume rows, so short level, change and days-to-cover cannot be stated at all. That is a genuine hole in the seller picture and nothing below substitutes for it.

What the record does show:

No forced or distressed seller is disclosed. No index deletion, fund liquidation or block unwind appears in the SEC filing set, the news corpus or the disclosure trail across the drawdown window.

No insider sold. Across 2024–2026 the Form 4 record contains grants, phantom-share settlements and one 100,000-share gift by director Greg L. Armstrong on 11 August 2025 — and not a single open-market sale. The last open-market insider purchases on file are CEO Willie Chiang's, in May and September 2023 at $13.25 and $16.24.

The company bought, in small size. CFO Al Swanson, on the 9 May 2025 call: "We did buy a small amount in April just as we're going into blackout. It was only about $7.5 million worth. I think it was about 475,000 units" [9]. The feature file records $8 million of FY2025 repurchase cash in total, against a market value in the billions — a gesture, not a support bid.

Dedicated income holders trimmed, gradually. Massachusetts Financial Services filed 5.0% in May 2025, 5.1% in August 2025 and 4.8% in January 2026, dropping below the 5% reporting line; its position has fallen steadily from 8.6% in February 2021. Energy Income Partners went from 5.8% in February 2024 to 4.47% in February 2026. Both are multi-year reductions by long-horizon midstream and income specialists, not a repricing of a view inside the drawdown window.

Taken together, the seller composition is the opposite of the framework's pattern. The evidence shows no anchored holders cutting into a vacuum; it shows a shrinking specialist holder base, an inert insider group, and one macro session at 4.1x volume.

Estimates against price

For the drawdown itself, the run's CapIQ file carries no dated FY2025 or FY2026 consensus history — the revision series (momentum) covers only FY2027 and FY2028 at 30, 90 and 180 days. Company guidance is therefore the dated forward number available across the fall, and it moved 0.3% at the midpoint between February and November 2025 while the price moved 23.9%. Delivered FY2025 EBITDA finished 1.5% below the pre-fall midpoint. Consensus FY2025 EBITDA of $2,802.9 million against the $2,833 million actual says the sell side was, if anything, a touch too low into the print.

Where dated revision data does exist — the recovery — price and estimates moved together, not apart:

Source: consensus figures from the run's CapIQ estimates file (data/sp/estimates.json, momentum block); closing prices derived from the run's daily price file.

Between 3 February 2026 and 31 July 2026, FY2027 consensus EPS rose 23.7% ($1.770 to $2.190) and the share price rose 24.7% ($21.11 to $26.33). The move tracks the estimate change close to one-for-one. The company's own forward number did the same: the 2026 Adjusted EBITDA guide was $2,750 million ±$75 million in February 2026 [7] and was raised by $130 million to $2,880 million ±$75 million on 8 May 2026 [10], on a WTI assumption of $85/bbl for the year [11] against the $65/bbl assumed a year earlier [6].

The signature the framework looks for — a price fall that outruns the estimate cut — is not present in either direction here. On the way down, the price fell 23.9% against a forward number that barely moved; on the way up, price and estimates moved within a percentage point of each other.

Bottom line

There is no dislocation in Plains GP Holdings today. The 2025 drawdown was real, dated and externally caused, but it was 23.9% deep rather than the 60–70% the framework is built around; it carried a 1.59x volume multiple rather than a capitulation; the trough was set on below-normal turnover; and it has been fully retraced and then some, with the shares at $26.33 against a $22.13 pre-fall peak and a $24.64 consensus target. The fall also outran the damage only modestly and briefly — FY2025 EBITDA landed 1.5% below the guidance set before the fall — and the subsequent recovery has tracked rising estimates almost exactly. What the arithmetic in this tab feeds is a starting price near the top of its own multi-year range, not a moment of fear. The quantified gap between damage and price change is worked in Damage Math; what the current price implies for the yield bar is worked in Yield.

Two things would change this read. A crude-price shock of the April 2025 kind at three times the magnitude, or a Permian volume break that pushes the EBITDA guide down by more than the 1.5% the 2025 shock ultimately cost, would create the entry condition that does not exist now. Reported short-interest data, absent here, could also change the seller picture materially in either direction.