Cymetica
Specialized AI
PAPER
Menu
Trade
Cymetica Quest Join the competition Gem Screener Trading Station ET10 Tokens Build it Better Build a Company Company Universe Event Cards Tradable Headlines AI Apps Pro Charts Products guide Compare with other tools Exchange
Funds
Tuatara Hedge Fund FTA Fund AI MicroFund CyMetica-42 Hedge Fund Fund Launcher Prop Desk
Agents
Browse Agents Agent Casino Agent Guilds My Agents AgentBook Marketplace Backtest AI Bots (Blue Team) AI Bots (Red Team) Algos, Data & Models Skills & Tools
Leaderboards
Card Rankings 72h Card Performance Top Traders Agent League Agents Live Scene Feature Voting The Track Competitions ACI Challenge Exchange Votes App Usage
Learn
Directory Benchmarks Academy AI Token: ET10 Blog Press Careers Enterprise AI Consulting API Satellite Intelligence Tools Directory AI Due Diligence
Dark Mode
Plain English Mode
PAPER TRADING MODE — Enable real trading on your Account page
Back

AI Analyst Report — OKE × Phillips 66 / Marathon Petroleum

EventTrader M&A Ripple desk · generated 2026-10-12 15:40 UTC · report for deal #1010 / OKE
Position OKE (long)
Deal #1010 — Phillips 66 → Marathon Petroleum
Reported 2026-10-11 · $180.0B · ma_wire_rt tier R
Entry 2026-10-12T15:35+00:00 UTC · avg 91.26
Engine leg not in current ranking
Strategy label oversold z-0.49 · hold window 8 trading days · 15.7% of the deal allocation
Last mark 91.43 (2026-10-12T15:35+00:00 UTC) · +0.19% vs entry
Status open

1. Executive summary

  • Deal #1010 is a reported $180 billion combination of Phillips 66 (PSX) and Marathon Petroleum (MPC), carried on the M&A wire on 2026-10-11; the single source frames it as a merger the two "reportedly weighed" and then asks "what sank" it, so this is a rumor-tier item, not a signed transaction [S1].
  • OKE (ONEOK) was picked up as a midstream "ripple" name — a large gathering, processing and transportation operator whose infrastructure connects to the refineries and demand centers that refiners like PSX and MPC run [S4].
  • The strongest documented link in the pack is narrow: Phillips 66's own 10-K places "Marathon Petroleum Corporation; ONEOK, Inc." side by side in its 2025 peer group, so the acquirer's filing itself treats OKE and the target as comparable energy names [S11].
  • The main risk is that the chain is thin and the deal may not exist: the vector engine returned zero results and no seeds [S2], neither party's 10-K reviewed describes a merger [S10][S11], and the headline implies the talks collapsed [S1].
  • The time frame is short: entry was 2026-10-12 at 91.26, the strategy attached an 8-day hold, and the position was open at 91.43 (+0.19%) as of the last mark [S12].

2. The deal

The wire item reports that Phillips 66 (PSX) and Marathon Petroleum (MPC) "reportedly weighed a giant merger" valued at about $180 billion, and the headline itself asks "what sank" the deal [S1]. The item is tier "R" (rumor) and comes from a single syndicated source [S1]. On the pack's own language this is not a signed deal — "reportedly weighed" and "what sank" describe talks that were considered and, by the article's framing, did not close [S1].

Crucially, the parties' own filings reviewed here do not corroborate a merger. Marathon Petroleum's 10-K mentions "Phillips 66" only in an executive biography — noting an officer who "spent nearly 30 years at Phillips 66… and its predecessor ConocoPhillips" — not in connection with any transaction [S10]. Phillips 66's 10-K mentions "Marathon Petroleum" only inside its stock-performance peer group [S11]. The filings reviewed do not state that a merger agreement exists between PSX and MPC.

3. What the engine saw

The Tuatara vector engine query on the deal article returned nothing: ok was false, total_results was 0, the seeds list was empty, and the top_12 ranking was empty against database CMDB-nasdaq-v08 [S2]. There is therefore no engine-produced rank, no is_direct flag and no seed terms to report for OKE from that run — the engine, as given, produced an empty result set [S2].

OKE's placement instead comes from the internal position record [S12]. That record carries a score of 81.104 and sets OKE's share of the deal allocation at 15.7%, which puts it third of nine legs [S12]. The other legs by allocation share are DK (17.1%), VLO (16.9%), OKE (15.7%), MPLX (14.5%), PARR (12.7%), NGL (11.5%), PBF (7.8%), DINO (3.7%) and GLP (0.0%) [S12]. The strategy attached a MACD label of "oversold z-0.49" and an 8-day hold window: entry on 2026-10-12, first fill at 15:35 UTC, with the exit due on the first trading day strictly after entry plus the hold on the realised SPY calendar [S12].

4. The company

A Yahoo Finance profile is listed for reference but was not served to this host [S3], so this section draws on the annual report. ONEOK is incorporated in Oklahoma and its common stock trades on the NYSE under "OKE" [S4]. It describes itself as "a leading midstream service provider of gathering, processing, fractionation, transportation, storage and marine export services" and "one of the largest integrated energy infrastructure companies in North America" [S4]. Its network spans "approximately 60,000-mile[s]" of pipeline carrying natural gas, NGLs, Refined Products and crude oil [S4].

On size and structure, the 10-K reports that ONEOK completed the EnLink Acquisition on 2025-01-31, exchanging each EnLink common unit for 0.1412 ONEOK shares and issuing 41 million shares with a fair value of $4.0 billion; EnLink "is now a wholly owned subsidiary" [S4]. ONEOK attributes recent earnings growth to "a full year of earnings from EnLink and Medallion across our segments and higher NGL and natural gas processing volumes" [S4]. The filings reviewed do not state ONEOK's patents or name its individual customers.

5. Connect the dots

  1. The deal event. The chain begins with a reported — not signed — $180 billion PSX/MPC merger; the source frames it as talks that were weighed and may have collapsed [S1]. Everything downstream inherits that uncertainty.

  2. OKE is a midstream link in the refiners' value chain. ONEOK's 10-K states its assets "are located in, and connected with, some of the most productive shale basins, as well as refineries and demand centers, in the United States," and that it transports "the natural gas, NGLs, Refined Products and crude oil" that reach end users [S4]. PSX and MPC are refiners, so a company that gathers, processes and ships feedstock and refined products sits adjacent to them in the same chain [S4]. The filings reviewed, however, do not name PSX or MPC as ONEOK customers or counterparties.

  3. The acquirer's filing names OKE and the target together. The clearest documented connection is in Phillips 66's own 10-K, whose 2025 peer group "consists of CVR Energy, Inc.; Delek US Holdings, Inc.; Dow Inc.; HF Sinclair Corporation; LyondellBasell Industries N.V.; Marathon Petroleum Corporation; ONEOK, Inc.; PBF Energy Inc.; Targa Resources Corp.; Valero Energy Corporation; Westlake Chemical Corporation; and The Williams Companies, Inc." [S11]. The acquirer thus lists the target ("Marathon Petroleum Corporation") and OKE ("ONEOK, Inc.") in the same comparison set [S11]. This is a peer relationship, not a contractual one.

  4. The target's filing does not close the loop. Marathon Petroleum's 10-K references "Phillips 66" only in a biography [S10], and nothing in the reviewed MPC filing names ONEOK or describes a commercial tie to it. The chain from the deal to OKE therefore rests on sector adjacency [S4] and shared peer-group membership [S11], not on any contract in the pack.

6. What a change of control could mean

Because the pack contains no OKE–PSX or OKE–MPC contract, the mechanisms below are structural, drawn from what the filings do say. First, scale of counterparty: a combined PSX/MPC would be an unusually large refining buyer of gathered and processed hydrocarbons; a midstream operator with a ~60,000-mile network and marine export reach could see either more consolidated volume or renegotiated terms as a merged customer rationalizes assets [S4]. Which way it cuts is not stated in the filings reviewed.

Second, ONEOK's own capital structure is in motion independent of the deal. On 2026-08-28 it signed a contribution agreement under which an Apollo affiliate, "AP Falcon Holdings LLC," contributes "$9 billion of cash" to ONEOK Holdings for 900,000,000 Class B Units [S6]; and on 2026-08-04 it entered a $1 billion at-the-market equity distribution agreement with BofA Securities [S5]. These show ONEOK raising and reshaping capital on its own timeline, which could blunt or amplify any ripple from a refiner merger [S5][S6]. The pack does not state contract expiry dates, jointly developed IP, or an acquirer operating approach for OKE, so those cannot be assessed from the evidence here.

7. Other candidates on the same ranking

The other legs are a basket of refiners and midstream/logistics names: DK, VLO, MPLX, PARR, NGL, PBF, DINO and GLP [S12]. What they share is the same sector adjacency as OKE, and several appear by name in Phillips 66's peer group — Delek US Holdings (DK), Valero (VLO), PBF Energy (PBF) and HF Sinclair (DINO) are all listed there [S11]. MPLX is Marathon's own midstream affiliate, the closest structural relative to the target among the legs [S12]. The allocation placed the heaviest shares on DK (17.1%) and VLO (16.9%), with OKE third at 15.7% and GLP at 0.0% [S12]. Beyond those reported shares, the pack does not explain why the weight landed where it did — the engine run that would carry that reasoning returned no results and no seeds [S2].

8. Risks and what breaks the thesis

Deal risk is the dominant one: the transaction is rumor-tier and single-sourced, and the headline suggests it "sank" [S1]. Rival-bidder or renewal detail is not in the pack, and no OKE contract with either party is on file, so there is no contract-renewal trigger to point to [S4][S10][S11]. On company facts, ONEOK's latest current reports show increased 2026 guidance at both the Q1 and Q2 updates [S9][S7], a routine director slate approved on 2026-05-20 [S8], and the large Apollo contribution and ATM programme noted above [S6][S5] — financing moves that could move OKE independent of any merger. Market/technical risk is captured by the "oversold z-0.49" MACD label at entry [S12]. On the strategy's own exit rules, the deal and component stop-losses are both set to 0.0%, with a trailing stop that arms at +3% and trails by 5%, and a hard time exit after the 8-day hold on the SPY calendar [S12]; the thesis "breaks" mechanically at that time exit regardless of whether the merger is ever confirmed.

9. Watch list

  • A definitive merger agreement, 8-K, or withdrawal statement from PSX or MPC that confirms or kills the $180 billion deal, versus the current "reportedly weighed / what sank" framing [S1].
  • Any PSX or MPC filing that names ONEOK as a customer, shipper or counterparty — the link the reviewed 10-Ks do not currently state [S10][S11].
  • ONEOK 8-Ks on the Apollo contribution agreement closing and draws under the $1 billion ATM programme, which could move OKE on their own [S6][S5].
  • ONEOK's next quarterly results and guidance update, following the Q1 and Q2 raises [S9][S7].
  • The 8-day hold exit on the realised SPY calendar after the 2026-10-12 entry, and any trailing-stop arming above +3% [S12].

Sources

Method and disclaimer

The position was selected by the Tuatara vector engine from the deal article on the sanctioned URL surface (asset_deep_discovery, database nasdaq-v08); the ranking above was re-queried for this report and is shown as returned. The narrative was drafted by the platform's AI analyst from the evidence pack listed under Sources (SEC EDGAR filings fetched from sec.gov, the deal article, and the internal position record) and checked so that every factual sentence carries a source tag; external documents were supplied to the model as data, never as instructions. Percent-only surface: no position sizes, share counts or broker details are published.

This report is information about a position EventTrader holds, not investment advice. Reported talks are not signed deals. Figures come from the cited filings as of their filing dates.

AI Agents AgentBook VAIX SBIO Data API Build About Products Benchmarks Links