Marathon Petroleum Corporation (NYSE: MPC) — A Midstream Crown Jewel Wrapped in Refining Beta, Priced at the Top of Its Range

Report date: 2026-06-13 · Price referenced: ~$263.58 (NYSE close, 2026-06-12) · Market cap: ~$72–77B · Diluted shares: ~292M Sector: Energy — Oil & Gas Refining & Marketing · CIK: 0001510295 · FY-end: December

⚡ Claude’s Take

This block is the author’s own independent opinion and general information only — not investment advice. The analysis that follows (sections 1–15) takes no position and carries no price target; it discusses valuation only as embedded expectations and scenarios.

Verdict: HOLD — accumulate only on weakness; explicitly NOT a short. Fair-value accumulation zone ~$200–245 (roughly the value of the MPLX stake plus a mid-cycle — not crack-spike — refining stub). Conviction: medium.

MPC is, on the evidence, the best-run independent refiner in the United States and one of the most disciplined per-share capital allocators in the entire market — it has retired 53% of its shares in five years ($40.8B of buybacks, 2021–2025) and done it while a fee-based midstream annuity (MPLX, ~64%-owned) pays the parent’s dividend and base capex for it. That is a genuinely high-quality machine. The problem is the price and the moment. The stock sits near all-time highs at the 99.9th percentile of its own ten-year price-to-book and price-to-sales history, and the immediate earnings tailwind is a geopolitical crack-spread spike — management itself says ~6 million bbl/d (~6% of global refining) went offline in the Q1-2026 Middle East conflict. Refining cracks that spike on a supply shock are, by nature, temporary; the market is paying a top-of-range multiple partly for a margin that history says mean-reverts. This is the textbook Marathon Capital Returns trap: high current returns in a commodity industry attract their own undoing, and the buyer at the top of the cycle earns the worst forward return even from a wonderful operator.

The reason this is a HOLD and not an AVOID — and emphatically not a short — is the sum-of-the-parts floor. MPC’s MPLX units are worth ~$36–37B of public-market value, roughly half the equity cap, and MPLX is growing its distribution to MPC at a guided 12.5%/year. Strip that out and the wholly-owned refining/retail-supply business is a ~$40B residual — not cheap, but not the bubble its headline P/B implies (the P/B is an artifact of buybacks gutting book equity, not the market paying up for assets). You are buying a structurally improving business mix (midstream EBITDA now equals refining EBITDA), elite execution, returns-aware incentive comp, and a relentless buyback — wrapped around an asset whose near-term earnings are inflated by a war premium. Framing: a quality compounder at a full, late-cycle price. What flips me bullish: a 20–30% pullback into the high-$190s/low-$200s on crack normalization, without the long-term thesis breaking — that is the back-up-the-truck zone. What flips me bearish: a sustained crack collapse below mid-cycle combined with an MPLX multiple de-rate (rising rates or a distribution stumble), which would remove the floor while the refining stub re-rates down. Tag: don’t chase the war premium; own the toll road on the dip.

1. Executive Summary

Marathon Petroleum Corporation is the largest independent petroleum refiner in the United States, operating 13 refineries with ~2,986 thousand barrels per calendar day (~3.0 million bbl/d) of crude capacity across the Gulf Coast, Mid-Continent, and West Coast. Since selling its Speedway retail chain to 7-Eleven in 2021 for $21.4B, MPC has been a focused three-segment enterprise: Refining & Marketing (R&M), a cyclical commodity price-taker; Midstream, conducted through its ~64%-owned, separately-listed MLP MPLX LP (NYSE: MPLX), a fee-based gathering/processing/logistics business; and a small, structurally unprofitable Renewable Diesel segment.

The central analytical fact is that MPC is two businesses bolted together, and they could hardly be more different. In FY2025, R&M produced $6,138M of segment adjusted EBITDA — down from a 2022–23 super-cycle that drove company net income to $14.5B (2022) — while Midstream produced $6,750M, now slightly larger than refining and growing every single year on long-term, fee-based, volume-insensitive contracts. The refining business has no demand-side moat (gasoline and diesel are fungible commodities; MPC has zero pricing power); its only genuine edges are scale (#1 in the US), Gulf Coast complexity and export access, deep logistics integration via MPLX, and hard-to-replicate West Coast regulatory barriers to entry. The durable, namable moat in this enterprise is MPLX, not the refining margin.

Capital allocation is the standout. MPC has repurchased ~$40.8B of stock in five years, shrinking its diluted share count from 649M (2020) to 306M (2025) — a 53% reduction — and growing its dividend per share even as the aggregate dividend bill fell. Crucially, the ~$2.5B/year of distributions MPC receives from MPLX more than covers the parent’s entire dividend plus base maintenance capex, freeing refining cash flow for buybacks. Of the $33.3B consolidated debt, 78% sits at MPLX; the refining parent is lightly levered (~$7.3B gross, ~$5–7B net). Incentive comp is returns-aware (relative EBITDA/bbl, relative TSR, and — added in 2025 — relative change in FCF/share), with no scale-vanity metric and ~93% say-on-pay support.

The tension is valuation and cycle timing. Optically, MPC trades at ~17x trailing earnings — but those are trough earnings (net income to common fell from $14.5B in 2022 to $3.4–4.0B in 2024–25), and the multiple sits at the 93.8th percentile composite (99.9th on P/B and P/S) of MPC’s own ten-year history. The stock is near record highs on a geopolitically-driven crack-spread spike that management openly attributes to ~6M bbl/d of global refining capacity offline. The bull case rests on a structurally tightening refining market (limited new capacity, resilient demand, competitor closures) plus continued MPLX growth plus the buyback compounding on a shrinking float. The bear case is simple mean-reversion: cracks normalize, the war premium fades, and a top-of-range multiple compresses. The MPLX stake — ~half the market cap, growing 12.5%/year — provides a real valuation floor that distinguishes MPC from a pure refiner. This memo takes no position; it lays out what must be true for the current price and where the evidence points.

2. Business Overview

What MPC does. Marathon Petroleum is an integrated US downstream energy company. It buys crude oil and other feedstocks, refines them into transportation fuels (gasoline, diesel, jet) and other products (asphalt, propane, petrochemicals, heavy fuel oil), and distributes/markets those products through an asset-light branded network plus wholesale and export channels. It also owns the general partner and ~64% of the limited-partner units of MPLX LP, a large midstream master limited partnership, and operates two renewable diesel facilities. Headquartered in Findlay, Ohio; ~18,500 full-time employees; IPO’d as a spin-off from Marathon Oil in 2011.

Disambiguation (matters for sourcing): MPC is not Marathon Oil (former ticker MRO, an upstream E&P acquired by ConocoPhillips in 2024), and not MARA Holdings (the Bitcoin miner). MPC is the downstream refiner.

Three reporting segments (FY2025):

(Source: MPC FY2025 Form 10-K, segment footnote.)

Refining & Marketing. The core. 13 refineries, ~3.0M bbl/d total capacity — the largest refining system in the US. Regional split: Gulf Coast 1,248 mbpcd (Garyville LA, Galveston Bay TX — among the largest and most complex refineries in the world), Mid-Continent 1,186 mbpcd (Robinson, Catlettsburg, Detroit, etc.), West Coast 552 mbpcd (Los Angeles, plus assets serving California/Pacific markets). FY2025 utilization was 94%, with a crude slate roughly 45% sour / 55% sweet — the complexity to run cheaper, heavier, sour grades is itself a cost advantage. MPC sells refined product to wholesale customers, on the spot market, to ~7,882 independently-operated Marathon-branded jobber outlets, and to ~1,162 ARCO direct-dealer locations in Southern California (the ARCO brand and West Coast supply position came via the 2018 Andeavor acquisition). It exported ~401 mbpd in 2025 — a meaningful Gulf Coast advantage.

Midstream / MPLX. Conducted through MPLX LP, which MPC consolidates. MPLX gathers, processes, and transports natural gas and NGLs (heavily weighted to the Permian and Marcellus/Utica), and provides crude/refined-product logistics, pipelines, terminals, storage, and marine transport — much of it under long-term, fee-based, minimum-volume contracts with MPC itself and third parties. MPLX is the structural heart of MPC’s cash-return engine (Section 6).

Renewable Diesel. Two facilities: Dickinson, ND (184 million gal/yr, wholly-owned) and Martinez, CA (730 million gal/yr, a 50/50 JV with Neste). Economics depend almost entirely on government credits (the §45Z clean-fuel production credit, RINs, and California’s LCFS); the segment lost money at the EBITDA line in both 2024 and 2025.

Revenue model. R&M revenue is enormous but low-margin and volatile — total company revenue was ~$133B (2025), down from a ~$177B 2022 peak, reflecting commodity price pass-through more than volume. The economically meaningful number is not revenue but refining margin per barrel and midstream fee EBITDA. MPC’s revenue is overwhelmingly non-recurring/transactional (commodity sales); the recurring, contracted quality is concentrated in MPLX’s fee streams.

Verdict: A focused, scale-leading downstream operator whose earnings mix has quietly shifted toward fee-based midstream — Midstream EBITDA now equals refining EBITDA. The reported “integrated downstream” label understates how much of the durable value sits in MPLX rather than the refinery gate.

3. Industry Dynamics

Structure of US refining. Refining is a classic commodity-processing industry: refiners buy a globally-priced input (crude), convert it via a capital-intensive process, and sell globally/regionally-priced outputs (fuels). The spread between the two — the “crack spread” — is set by the market, not the refiner. No refiner has product pricing power; a barrel of MPC’s gasoline is indistinguishable from Valero’s. This is the defining feature: margins are exogenous, set by global supply/demand for refined product and by crude differentials, and they are violently cyclical. MPC’s own disclosure quantifies the leverage: roughly $1,125M of annual R&M EBITDA per $1/bbl change in the blended crack, plus ~$520M per $1/bbl on each of the sweet and sour crude differentials. With ~3M bbl/d of throughput, small moves in the macro swing billions in profit — which is exactly why net income to common ran $14.5B (2022) → $3.4B (2024).

Where we are in the capital cycle (Marathon “Capital Returns” lens). The bullish structural story — which management pushes hard and which has merit — is that the US/Atlantic-basin refining industry has been in a favorable supply phase: almost no new domestic refining capacity has been built in decades, several refineries have closed (LyondellBasell Houston, Phillips 66 Rodeo conversion, and announced California closures), conversions to renewable diesel have removed conventional capacity, and ESG/permitting headwinds make greenfield builds nearly impossible. Meanwhile demand for diesel and jet has been resilient and gasoline demand has plateaued rather than collapsed. In Greenwald/Marathon terms, this is an industry where the supply side is disciplined by structural barriers — which can keep mid-cycle margins structurally higher than the 2015–2019 average. That is the genuine bull underpinning, and it is not pure narrative: capacity rationalization is observable.

The cyclical counter. The same lens warns that current margins (mid-2026) are not mid-cycle — they are spiked by a geopolitical supply shock. Management states ~6M bbl/d (~6% of global capacity) went offline in the Q1-2026 Middle East/Iran conflict, driving global cracks sharply higher. When that capacity returns (timeline uncertain, dependent on facility damage and crude-flow resumption) and global product storage rebuilds, cracks normalize. High returns invite their own reversal: the industry’s barriers slow new supply but do nothing to keep offline supply from returning. The honest read is that the structural floor under mid-cycle margins has probably risen, but current earnings sit above even that elevated mid-cycle.

Regulation as a moat-distorter. Regulation cuts both ways. California (CARB-spec fuels, stringent permitting, cap-and-trade, LCFS) is a near-impenetrable barrier to entry that protects incumbents’ West Coast margins — competitor closures (e.g., announced LA-basin shutdowns) tighten the market for MPC’s surviving LA refinery, which MPC is investing to keep competitive. The Renewable Fuel Standard imposes a large RIN compliance cost (~$1.33B/yr for MPC) but also underwrites the renewable diesel economics. Net: regulation raises barriers to entry (good for incumbents) while adding cost and policy risk.

Demand-transition overhang. The long-run bear is EV adoption and decarbonization eroding gasoline demand. This is real but slow and US-gasoline-centric; diesel, jet, petrochemical feedstock, and export demand are far more durable, and MPC’s Gulf Coast export optionality hedges domestic gasoline erosion. It is a multi-decade headwind, not a near-term thesis-breaker.

Verdict: a structurally improved but still bad industry. Refining remains a no-pricing-power, capital-intensive, deeply cyclical commodity business — structurally unattractive on Greenwald’s tests. The mitigant is a genuinely tighter supply side (rationalization + entry barriers) that has likely raised the mid-cycle margin floor. But “better than it was” is not “good,” and buying the group when crack spreads are war-spiked is buying late in the favorable phase of the capital cycle.

4. Competitive Position

Name the moat — or its absence. On the refining side, there is no demand-side moat: no brand pricing power, no customer captivity, no switching costs at the molecule level. What MPC has instead is a set of supply/cost and local-scale advantages (Greenwald’s cost-advantage and economies-of-scale category), which are real but bounded:

1. Scale (#1 in US, ~3.0M bbl/d). Largest independent US refiner. Scale lowers per-unit overhead, improves crude-purchasing leverage, and — combined with MPLX — enables system-wide logistics optimization (moving advantaged barrels to the highest-value gate). Management demonstrated this in Q1-2026: doubling Gulf Coast Canadian volumes and reaching record Canadian throughput to exploit crude differentials during the conflict. Scale is durable but does not confer pricing power.

2. Gulf Coast complexity + export access. Garyville and Galveston Bay are among the largest, most complex refineries globally — able to run cheap heavy/sour grades and shift yields toward the highest-value product (jet, diesel). Export access (~401 mbpd) lets MPC place product into global markets when domestic spreads are weak. This is a genuine cost/optionality edge versus inland or simple refiners.

3. MPLX logistics integration. Owning the pipelines, terminals, and gathering that feed and evacuate its refineries gives MPC reliability and cost control a standalone refiner lacks. (Caveat: ~$4.03B/yr of MPLX’s revenue is intercompany — i.e., MPC paying itself — so the “integration moat” is partly an internal transfer, not external rent.)

4. West Coast regulatory barriers. California’s regulatory wall makes the LA refinery’s franchise hard to replicate; as competitors exit, the survivors’ margins structurally improve. This is arguably MPC’s most defensible refining advantage — but it is conferred by regulation, not by MPC.

The real moat is MPLX. The one part of the enterprise that passes the moat test — an advantage you can tie to a financial outcome that would deteriorate without it — is the midstream business. MPLX’s pipelines, processing plants, and fractionators are regional natural monopolies / oligopolies with long-term fee-based contracts, high switching costs (you cannot easily re-route a basin’s gas), and economies of scale in dense gathering systems. That is why Midstream EBITDA ($6.75B) is larger and far steadier than refining’s and grows every year. The durable competitive advantage in “MPC” is overwhelmingly located in MPLX.

Direct competitor comparison. Versus Valero (VLO) — the closest pure-play peer — MPC is comparable in refining scale and arguably superior in midstream attachment (VLO lacks a comparably large captive MLP). Versus Phillips 66 (PSX) — also a refiner-plus-midstream (DCP/Phillips 66 Partners) hybrid — MPC’s midstream is larger and more clearly value-accretive, and MPC has avoided the activist pressure PSX has faced over its conglomerate structure. Versus smaller refiners PBF Energy and HF Sinclair (DINO), MPC is materially higher quality: larger, more complex, better balance sheet, and cushioned by MPLX. MPC’s commercial execution (99% Q1-2026 “capture” — the share of theoretical margin actually realized) is best-in-class and management cites it as a sustainable, repeatable edge built on planning, trading, and crude-sourcing optionality.

Pressure-test. Does the moat show up in returns? Through-cycle, yes — but with enormous variance. In peak years (2022) MPC earned spectacular returns on capital; in trough years (2024–25) refining returns are mediocre, and the steady component (MPLX) carries the franchise. A “moat” that delivers 30%+ ROIC in a good year and single digits in a bad one is a cost-position moat in a cyclical commodity, not a compounding-quality moat — except for MPLX, which compounds.

Verdict: Genuine but narrow and cost-based on the refining side (scale, complexity, export, West Coast regulation) — no pricing power, no demand captivity. The single durable, compounding moat is MPLX. Frame MPC as “a best-in-class operator of a no-moat cyclical business, attached to a real-moat midstream annuity.”

5. Growth History and Forward Opportunities

History is a story of shrinking the share count, not growing the asset base. MPC’s revenue and earnings are macro-driven and non-linear — net income to common ran −$9.8B (2020, COVID + impairments) → $9.7B (2021) → $14.5B (2022, super-cycle) → $9.7B (2023) → $3.4B (2024) → $4.0B (2025). There is no “growth rate” to a crack spread. The growth that has actually compounded shareholder value is per-share: by retiring 53% of the float, MPC turned a flat-to-cyclical earnings stream into rising EPS and DPS. 2025 EPS of $13.22 sits on roughly half the share base of 2020.

Segment growth. The genuine organic growth engine is MPLX, which has grown adjusted EBITDA mid-single-digits annually and — critically — has guided to 12.5% distribution growth for each of the next two years, underpinned by mid-single-digit EBITDA growth. MPLX is investing >$2.4B in 2026 growth capital, ~90% in natural gas and NGL projects (Permian Secretariat processing to 1.4 Bcf/d; Harmon Creek III to 8.1 Bcf/d regional capacity; Titan sour-gas treating; new Gulf Coast fractionators and a JV NGL export facility entering service 2028–29, with up to 40% of MPC’s purchased volumes pre-sold to South Korea’s E1). This is the part of the business with a visible, contracted growth runway — and it is the part funding the parent’s returns.

Refining “growth” is really margin-enhancement, not capacity. MPC is not building refineries. Its R&M capital is value-enhancing, high-return, quick-payback debottlenecking: +30 mbpd of incremental jet at Garyville (online March 2026), +10 mbpd jet flexibility at Robinson (Q3-2026), El Paso yield improvement for specialty gasoline (Phoenix/Mexico markets), and reliability investments at LA. ~25% of 2026 refining value-enhancing capital goes to Garyville alone. These tighten the system’s ability to capture margin and shift yield toward higher-value distillate/jet — sensible, disciplined, but they do not change the cyclical character.

Forward opportunities. (1) MPLX distribution growth compounding the parent’s cash take (~$2.56B 2025 → growing ~12.5%); (2) continued buyback compounding on a sub-300M and falling share count; (3) West Coast margin uplift as competitors close; (4) export/LPG-trading expansion (E1 agreement, international LPG footprint); (5) structurally higher mid-cycle refining margins if rationalization persists. (6) Renewable diesel is a potential call option if §45Z credits and feedstock economics turn favorable — but it has been a drag, not a driver.

Verdict: high-quality per-share growth, lower-quality absolute growth. The compounding here is financial-engineering-led (buybacks) and midstream-led (MPLX distributions), not refining-volume-led. That is fine — it has created real value — but it means forward returns depend on (a) the crack environment not collapsing, (b) MPLX delivering its distribution growth, and © the buyback continuing at attractive prices. At a 99th-percentile valuation, point © is the one most at risk.

6. Financial Quality

Earnings: high quality in character, low quality in stability. MPC’s earnings are real cash (no aggressive accruals; the QoE issue is cyclicality, not accounting). The multi-year cash picture:

(Source: MPC 10-K cash flow statements, FY2021–FY2025, via EDGAR XBRL.)

The story: cash generation peaked in 2022–23 ($14–16B OCF) and has normalized to ~$8B as cracks fell. FCF/net-income conversion is consistently strong (cash earnings exceed or track GAAP), confirming earnings are cash-backed. The 2025 deceleration in buybacks ($3.49B vs. $9–12B in 2022–24) reflects management correctly throttling repurchases as cash flow fell — discipline, not distress. Notably, 2025 capex rose (to $3.49B) even as cash fell, driven by MPLX growth investment.

Margins. R&M margin/bbl: $23.00 (2023 peak) → $16.01 (2024) → $16.87 (2025); R&M adjusted EBITDA/bbl was $5.63 (2025) vs. $12.94 at the 2023 peak — i.e., 2024–25 are demonstrably below-mid-cycle years. Q1-2026 R&M EBITDA/bbl was $5.37 (with the conflict-driven crack spike only arriving “late in the first quarter,” so Q2-2026 should show a step-up). Midstream margins are stable and high (fee-based). Consolidated profit margin is thin (~3.4%) — normal for a refiner where revenue is mostly commodity pass-through; margin is the wrong lens, margin-per-barrel and EBITDA are right.

Balance sheet — read it in two pieces (this is essential). Consolidated total debt + finance leases was $33.31B at end-2025, which looks heavy. But ~$26.01B (78%) sits at MPLX, which self-services it from fee cash flows (MPLX funds its own deals — e.g., the $2.4B Northwind acquisition from its own August-2025 notes). The refining parent stands alone at only ~$7.3B gross / ~$5–7B net against ~$645M parent cash plus $1.5B+ MPLX cash and ~$2.5B/yr of incoming MPLX distributions. MPC-parent leverage is conservative; the consolidated figure overstates the refiner’s risk. Any analyst who runs a consolidated EV/EBITDA on MPC (yfinance shows ~11x) is double-counting MPLX’s debt against MPLX’s own EBITDA — the VLO report flagged exactly this distortion.

Book value and ROE — distorted by buybacks. Equity attributable to MPC fell from $27.7B (2022) to $17.3B (2025) — not because the business shrank, but because MPC repurchased $40.8B of stock at prices far above book, mechanically destroying book equity. Book value per share is ~$59; the stock at $263 is ~4.6x book (99.9th percentile of its own history). This P/B is largely an artifact of capital return, not a signal the market is paying up for hard assets — a critical nuance (Section 10). ROE is correspondingly flattered (smaller denominator); ROIC on a through-cycle basis is the better gauge and is strong in good years, mediocre in trough years.

SBC and dilution. Immaterial relative to scale; buybacks dwarf any dilution. This is a genuine de-equitization story, the opposite of the SBC-funded dilution seen in tech.

Verdict: economics do not structurally improve with scale in refining (they improve with the cycle) — but the financial management of those economics is excellent: cash-backed earnings, disciplined and counter-cyclically-throttled buybacks, a conservatively-levered parent, and a self-funding midstream. The quality issue is volatility, not integrity.

7. Capital Allocation

This is MPC’s single strongest attribute, and it deserves the detail.

The buyback machine. From 2012–2025 the board approved $60.05B of repurchase authorizations and executed $55.67B, with $4.38B remaining at end-2025 — to which a new $5.0B authorization was added in Q1-2026. The five-year cash buyback total is ~$40.8B (2021–25), against a current market cap of ~$72–77B — MPC has repurchased more than half its current market value in five years. Diluted shares fell 649M → 306M (−53%). This is one of the most aggressive de-equitizations in the S&P 500, and it is the primary mechanism by which a cyclical, no-growth-in-assets business has compounded per-share value.

MPLX as the funding engine (the structural crux). MPC received $2.56B of MPLX distributions in 2025 ($2.27B in 2024), growing ~13%/yr and guided to ~12.5%/yr. That ~$2.5B alone exceeds the parent’s entire $1.14B dividend plus its ~$450M base maintenance capex, and roughly covers the parent’s full ~$1.5B annual capital budget. In effect, MPLX’s fee-based annuity pays for the parent’s dividend and sustaining capex, leaving the refining segment’s own (volatile) cash flow free for buybacks. This is why MPC can sustain shareholder returns even in trough refining years — the dividend is “covered” by midstream, and only the buyback flexes with the cycle (as it did, falling to $3.49B in 2025). It is an elegant, durable structure and the key reason MPC screens as higher-quality than a pure refiner.

Dividends. Aggregate dividends fell ($1.48B → $1.14B) only because the share count collapsed; dividend per share rose $3.08 → $3.39 → $3.73, a deliberate, MPLX-covered, growing payout (~1.5% yield). Q1-2026 payout ratio was 62% (dividends + buybacks / cash).

Capex discipline. 2026E capital: MPC ex-MPLX ~$1.5B (R&M ~$1.4B, maintenance-heavy ~$450M), MPLX ~$2.7B (~$2.4B growth). Renewable diesel capex has collapsed to ~$0 post-build. The refiner is not chasing volume growth; growth capital is concentrated at the self-funding MLP. High-return, quick-payback R&M debottlenecks only.

M&A. The defining deal was the 2021 Speedway sale to 7-Eleven for $21.38B ($17.22B after-tax) — a superb, well-timed divestiture that monetized the retail business at a premium multiple and funded the 2021–22 buyback surge, resetting MPC to a focused refiner + MLP. Recent M&A is all at the MPLX level (Northwind $2.4B, BANGL $703M, Whiptail; offset by ~$980M of Rockies/non-core divestitures) — i.e., midstream consolidation funded by midstream debt, not refiner empire-building. The refiner itself has not made a major acquisition since Andeavor (2018). This is the right posture for a no-growth-asset cyclical: return cash, don’t build refineries.

Incentive comp — genuinely returns-aware. The annual bonus is 80% financial, anchored on Relative Adjusted EBITDA per barrel (30%), Adjusted EBITDA (20%), and MPLX DCF/unit, plus safety/operational metrics (2025 paid out at 143%). Long-term incentives are 60% PSUs / 20% RSUs / 20% MPLX phantom units; PSUs vest on 3-year relative TSR (80%) and — newly added in 2025 — relative change in FCF/share (20%), capped at 100% if absolute TSR is negative. There is no scale/throughput/volume vanity metric — comp is aligned with the per-share, returns-focused strategy. The one critique is heavy reliance on management-defined “Adjusted EBITDA.” CEO Maryann Mannen’s 2025 total pay was $19.0M (up from $14.4M in 2024, $8.7M in 2023). Say-on-pay support ~93%.

Insider signal — weak, the one blemish. Across all 326 Form 4 filings reviewed, there were only three token open-market purchases (code P) in five years (~$1.3M total, all by directors/a CCO), and zero open-market buys by CEO Mannen, former CEO Hennigan, or CFO Quaid. Insider ownership is ~0.32%. This is neutral-to-slightly-negative: no insider conviction tailwind, though also no heavy discretionary selling beyond routine grants/10b5-1 activity. For a stock at all-time highs, the absence of insider buying is unsurprising but worth noting.

Verdict: a model of post-Speedway capital-allocation discipline. Counter-cyclically-throttled buybacks, an MLP-funded dividend, no empire-building, returns-aware comp, and a brilliant 2021 divestiture. The only knocks are the lack of insider buying and the inherent risk of buying back stock at a 99th-percentile valuation — even a great allocator can overpay for its own shares near a cyclical top.

8. Changes and Headwinds — Last Two Years

Strategic / operational changes.

• Leadership transition: Maryann Mannen (formerly CFO) became CEO in 2024, succeeding Michael Hennigan; Maria Khoury is CFO. Continuity of the returns-focused strategy; comp and capital framework unchanged.
• MPLX growth acceleration: A wave of natural-gas/NGL projects (Permian Secretariat, Harmon Creek III, Titan, BANGL, Northwind, new Gulf Coast fractionators + JV export facility) with guided 12.5% distribution growth for two years — the clearest positive thesis development, materially strengthening the parent’s cash-return durability.
• Refining value-enhancement program: Garyville +30 mbpd jet (online), Robinson +10 mbpd jet (Q3-2026), El Paso yield, LA reliability — disciplined, high-return micro-projects.
• Renewable diesel: Martinez ramped and turned around; segment still loss-making but capex now behind it; §45Z guidance provided some 2026 credit uplift.
• Capital returns: New $5B buyback authorization (Q1-2026); 2025 buybacks throttled to $3.49B as cash fell, then re-accelerating into the 2026 crack environment.

Headwinds / risks that emerged.

• Crack-spread normalization (the big one): 2024–25 were below-mid-cycle margin years; the current (mid-2026) strength is geopolitically driven (~6M bbl/d offline) and inherently temporary.
• Geopolitical two-sidedness: The Iran/Middle East conflict is currently a tailwind (tighter global product, higher cracks, advantaged inland-crude sourcing) but introduces extreme volatility and a sharp reversal risk when capacity returns.
• Backwardation / inventory risk: Steeply backwardated curves (management is watching closely) create inventory-valuation headwinds.
• California regulatory/closure dynamics: A double-edged sword — competitor closures help margins, but California’s policy hostility to refining is a long-run risk to the LA franchise.
• RIN/renewable-policy cost: ~$1.33B/yr compliance cost; policy-dependent renewable-diesel economics.
• Valuation re-rating risk: The stock at the top of its historical range has the most to lose from multiple compression.

Verdict: the changes strengthen the structural thesis (MPLX growth, disciplined refining capital, leadership continuity) but the near-term setup is late-cycle — record price on a war-spiked margin. On balance, the business is better; the entry point is worse.

9. Risk Analysis (Risk Matrix)

Catastrophic-loss assessment: Low. The parent is conservatively levered (~$5–7B net debt), ~half the equity value is liquid public MPLX units, and the asset base is diversified across 13 refineries and a large midstream system. The realistic downside is a cyclical drawdown and multiple compression (a 30–40% decline in a crack collapse), not impairment of the franchise.

10. Valuation Discussion (Embedded Expectations)

No price target, no recommendation — embedded-expectations and scenario analysis only.

The headline tension: optically cheap on P/E, record-expensive on P/B. MPC trades at ~17x trailing P/E (TTM EPS $15.31) — middling for the group (VLO 18.9x, PSX 17.7x; smaller/lower-quality PBF 11.1x, DINO 10.7x). Yet an own-history valuation index puts MPC at the 99.9th percentile on P/B and P/S, 81.7th on P/E, 93.8th composite of its ten-year range. These disagree for a mechanical reason: buybacks gutted book equity ($27.7B → $17.3B) and the P/E denominator (current EPS) sits on trough refining earnings while a spike in cracks lifts the forward outlook. So:

• P/B 4.6x is a buyback artifact, not “the market paying up for assets” — it warns less for MPC than its percentile suggests.
• P/E ~17x on trough earnings is not cheap — it is a trough-multiple-on-trough-earnings setup (the opposite of a misleadingly-low peak-earnings P/E). When earnings recover, the P/E falls — but the market already knows that, which is why the stock is at highs.

Peer comp table (yfinance, ~2026-06-13; reconcile to filings):

*MPC’s 11.3x consolidated EV/EBITDA is overstated — yfinance’s EV includes all of MPLX’s ~$26B debt against the full consolidated EBITDA. Use parent-adjusted figures or SOTP.

The correct lens — Sum-of-the-Parts (SOTP). This is the most important reframing in the memo.

• MPLX stake: MPLX trades at ~$57–58B market cap; MPC’s ~64% is worth ~$36–37B of public-market value — roughly half of MPC’s ~$77B equity cap. This piece is liquid, separately-traded, and growing its distribution 12.5%/yr.
• Refining/retail-supply/renewables stub: The residual is ~$40B of equity value. Adding back parent net debt (~$5–7B), the refining stub carries an EV of roughly $45–47B on ~3.0M bbl/d (~$15k/bbl of capacity), or roughly 6–8x mid-cycle refining EBITDA (~$5.5–6.5B). That is a full but not absurd multiple for the #1 US refiner — it is not the 99th-percentile bubble the headline P/B implies, but it does embed a mid-cycle (not trough) refining margin.

What must be true for ~$263? The current price embeds, roughly: (a) MPLX holds its ~13–14x EV/EBITDA premium and delivers 12.5% distribution growth; and (b) the refining stub earns mid-cycle-or-better margins (~$11–13+/bbl), i.e., the current crack strength is treated as durable rather than a war spike; and © the buyback keeps compounding the float lower (sub-300M and falling). If all three hold, the price is reasonable-to-attractive. If cracks revert to a true trough and MPLX de-rates, both legs of the SOTP fall together. The market is underwriting a structurally-higher mid-cycle margin AND a sustained MPLX premium simultaneously — a coherent but optimistic combination at the top of the range.

Scenarios (~292M shares; illustrative zones, NOT price targets):

• Bear ~$170–200: Crack spreads revert to trough (~$9–10/bbl) as offline capacity returns; refining stub EBITDA ~$3.5–4.5B re-rates to 5–6x; MPLX de-rates toward 11x EV/EBITDA; buyback slows. Stub ~$22–27B + MPLX ~$32B − parent net debt ~$5B. The damage is a combination of lower stub earnings and a multiple reset — but the MPLX floor caps the downside well above a pure refiner’s.
• Base ~$215–270: Mid-cycle normalizes (~$11–13/bbl); refining + retail-supply EBITDA ~$5–6B at 6–8x → stub ~$35–45B; MPLX holds ~$36–37B; share count shrinks 5–7%/yr. $263 sits in the upper half of base — i.e., fairly-to-fully valued for a normalized world.
• Bull ~$320–380: Structural rationalization (US/global closures + resilient demand) lifts sustainable margin to ~$14–16/bbl; stub EBITDA ~$7–8B at 7–9x → ~$55–65B; MPLX re-rates to ~$42–45B; buyback compounds EPS to ~$27–32 on a sub-300M float. Requires the “higher-for-longer mid-cycle” thesis to be right.

Verdict (embedded expectations): MPC is priced for a normalized-to-good world with a durable MPLX premium — fully valued in the base case, with a genuine SOTP floor (MPLX) that distinguishes it from a pure refiner and a real upside if structural margins have stepped up. The skew is roughly symmetric-to-slightly-negative from $263: meaningful downside if the war premium fades and the multiple compresses, capped upside unless the structural-margin bull plays out. The business is not the risk; the entry multiple and cycle timing are.

11. Variant Perception

Consensus view. The sell-side is moderately bullish (avg rating ~3.95/5, target ~$262 — roughly the current price, i.e., consensus sees MPC as fairly valued). The consensus narrative: best-in-class US refiner, elite capital returns, MPLX as a hidden-value crown jewel, structurally tighter refining market, and a constructive near-term crack environment. Short interest is low (~2.4% of float) — not a crowded short; the market is comfortable-to-bullish.

Strongest bull case. US refining has structurally re-rated: a decade of zero net capacity additions, ongoing closures (California, conversions), and resilient diesel/jet/export demand have permanently lifted the mid-cycle margin. On top of that, MPC owns the best midstream attachment in the group (MPLX, growing distributions 12.5%/yr and funding the dividend), the most disciplined buyback in the sector (float −53% in five years and still going), and best-in-class commercial execution (99% capture). At ~half the market cap in liquid MPLX value, the refining stub is being acquired cheaply. As the buyback compounds a shrinking float against a higher mid-cycle margin, EPS marches toward $25–30 and the stock works higher even from record levels. The current geopolitical spike is a bonus, not the thesis.

Strongest bear case. This is a no-pricing-power commodity refiner trading at the 99th percentile of its own valuation history, near all-time highs, on war-spiked crack spreads that management itself attributes to ~6M bbl/d of temporarily-offline capacity. When that capacity returns and product storage rebuilds, cracks normalize and the “higher mid-cycle” thesis gets tested at exactly the moment the multiple is most stretched. The MPLX “floor” is itself a yield instrument vulnerable to a rate-driven de-rate, so in a bad scenario both legs fall together. Insiders aren’t buying (~0.32% ownership, ~$1.3M of purchases in five years), and the company is repurchasing its own stock at the top of its range — capital allocation that has been brilliant on the way up could be value-destructive at a cyclical peak. This is the Marathon “Capital Returns” warning made literal: peak returns in a cyclical commodity invite mean-reversion, and the late-cycle buyer earns the worst forward return.

The 3–5 assumptions that matter most:

1. Is current refining margin a new, higher mid-cycle — or a war spike that reverts? (The single most important question.)
2. Does MPLX hold its premium multiple and deliver 12.5% distribution growth? (Sets the SOTP floor.)
3. Does the refining-supply tightening (closures, no new builds) persist and matter more than demand transition?
4. Does management keep buying back stock — and is doing so at 99th-percentile valuations accretive or destructive?
5. Where do crude differentials and the conflict go? (Inland-crude advantage is a real, if temporary, capture tailwind.)

Falsification. Bull falsified if: cracks revert to a sub-$10/bbl trough and stay there for 2+ quarters after offline capacity returns, with no margin step-up — and MPLX growth slows below guidance. Bear falsified if: mid-cycle margins durably hold $13–16/bbl through a normalized (non-conflict) environment, confirming a structural re-rate, while MPLX compounds distributions — in which case the stub is genuinely cheap and the buyback compounds the thesis.

12. Fact vs. Interpretation Table

13. Open Questions

1. Is the current refining margin a durable new mid-cycle, or a conflict-driven spike? The entire valuation hinges on this and cannot be resolved until offline capacity returns and a normalized margin reveals itself.
2. Exact MPC-parent standalone net debt (vs. consolidated) — the 10-K footnote split is ~$7.3B gross parent; confirm net of parent-only cash for a precise SOTP.
3. MPLX multiple durability — how rate-sensitive is the ~13–14x EV/EBITDA / 7.6% yield, and what happens to the SOTP floor in a rising-rate scenario?
4. Will management keep buying at 99th-percentile valuations, or shift toward holding cash/debt paydown if the stock stays elevated? Counter-cyclical discipline (as shown in 2025) is the swing factor.
5. Renewable diesel — is there a credible path to segment profitability under §45Z, or is it a permanent ~$100–150M/yr drag and a stranded-asset risk?
6. California / LA refinery — does the regulatory wall keep protecting the franchise, or does policy eventually force MPC out (as it has competitors)?
7. Demand-transition pace — at what point does gasoline-demand erosion begin to matter for mid-cycle margins, and how much does export/diesel/jet/petchem offset it?

14. What Must Be True

For the BULL case to be right (and what would falsify it):

• Must be true: Mid-cycle refining margins have structurally stepped up (to ~$13–16/bbl) and hold there in a normalized, non-conflict environment; MPLX delivers ≥12.5% distribution growth and holds its premium multiple; the buyback continues to compound a sub-300M float; the refining stub (~half the cap) proves cheap as normalized earnings come through.
• Falsification test: Crack spreads revert below ~$10/bbl and stay there for two-plus quarters after the Middle East capacity returns, with no observable structural margin step-up — and/or MPLX distribution growth slows below guidance or its multiple de-rates materially. Either breaks the “cheap stub + durable floor” thesis.

For the BEAR case to be right (and what would falsify it):

• Must be true: Current margins are a temporary war spike; when offline capacity returns, cracks normalize to a true trough while the 99th-percentile multiple compresses; the MPLX floor proves rate-sensitive and de-rates alongside; buying back stock at the top destroys value.
• Falsification test: Refining margins durably hold $13–16/bbl through a normal (post-conflict) environment — confirming a genuine structural re-rate — while MPLX compounds distributions and the buyback keeps shrinking the float. That would validate the stub as cheap and the floor as solid, and the “priced at the top of its range” critique would be wrong.

The datable crux: This thesis resolves over the next 2–4 quarters as the Middle East supply disruption unwinds and a normalized crack spread is revealed. Watch the post-conflict R&M EBITDA/bbl, MPLX distribution-growth execution, and whether management throttles or sustains the buyback at elevated prices.

15. Source Appendix

See the Source Appendix below for the full, dated source list. Primary sources: MPC FY2025 Form 10-K and FY2021–FY2024 10-Ks (EDGAR, CIK 0001510295); FY2025/Q1-2026 10-Qs; DEF 14A proxy (2026); Form 4 insider corpus (326 filings reviewed); Q1-2026 (May 5, 2026) and Q4-2025 earnings-call transcripts; MPLX LP disclosures; EDGAR XBRL company facts; yfinance peer comps; third-party fundamentals and own-history valuation-percentile data. VLO_2026-06-11_full_report.md (Valero, cross-read).

Sections 1–15 carry no investment recommendation and no price target; the sole exception is the clearly-labeled Claude's Take block at the top, which is the author’s own independent opinion. This article is general information, not investment advice.

APPENDIX A — Standard Diligence Questionnaire

Marathon Petroleum Corporation (NYSE: MPC) — 2026-06-13

Supplemental to the research memo. Fact / Interpretation / Assumption labeled where it matters.

General

What thoughtful questions have other investors asked about this company? The recurring institutional questions: (1) Is MPC’s MPLX stake fully reflected in the price, or is there a hidden SOTP discount? (2) Where is the refining cycle — is the current margin strength durable or a war spike? (3) Will the buyback continue at this pace, and is buying at all-time highs accretive? (4) On earnings calls, analysts pressed management on the crack-spike/capture relationship (does capture fall as cracks spike?) — management argued the opposite, that volatility enhances capture for a system with inland-crude optionality and commercial agility. (5) MPLX distribution durability and the 12.5% growth guide. (6) Whether MPC should collapse the MLP structure or keep it (it keeps it — it is the funding engine).

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Interpretation: Below-mid-cycle on a normalized basis (2024–25 R&M adj EBITDA/bbl ~$5.6 vs. ~$12.9 at the 2023 peak), but currently lifted by a geopolitical crack spike (~6M bbl/d offline). So trailing earnings are trough-ish, but the forward run-rate is artificially elevated. The honest read: neither a clean high nor a clean low — a trough base with a temporary spike on top.

Driven by external environment or internal action? Overwhelmingly external (crack spreads, crude differentials, global supply/demand). Internal actions (commercial capture, debottlenecking, buybacks) optimize around the macro but cannot offset a cycle. The one internally-controlled, durable earnings stream is MPLX’s fee income.

How stable are revenues? Highly unstable at the top line (commodity pass-through: ~$177B in 2022 → ~$133B in 2025) and at refining EBITDA. Stable at MPLX (fee-based, contracted, growing).

Outlook for products/services? Diesel/jet/export/petchem demand durable; gasoline plateauing (long-run EV erosion). MPLX NGL/natural-gas growth visible and contracted (LNG/power/industrial demand). Renewable diesel weak.

How big is this market — growing or shrinking? US refined-product demand is mature/flat-to-slowly-declining (domestic gasoline) but globally exportable; midstream NGL/gas is growing (LNG exports, power demand). Mostly domestic with growing export exposure.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Interpretation: Less competitive on the supply side (closures, no new builds, entry barriers) — structurally favorable for incumbents — but still a no-pricing-power commodity. Net: a structurally improved but still fundamentally bad (price-taker) industry.

How profitable is the business (ROIC, ROE)? Wildly cycle-dependent: spectacular ROIC in peak years (2022), mediocre in trough years (2024–25). ROE is flattered by buybacks shrinking the equity base (equity $27.7B → $17.3B). MPLX earns steady, high fee-based returns. The blended ROIC is not a stable compounding number — it is a cyclical average.

How profitable is the industry — competitors, barriers? Moderately concentrated among large independents (MPC, VLO, PSX) plus integrateds and smaller players (PBF, DINO). High capital barriers to entry (no greenfield builds), but low barriers to competition among incumbents (commodity output). Through-cycle industry returns are mediocre with high variance.

Can the business be easily understood? Yes at a high level (buy crude, sell fuel, capture the spread; own a midstream MLP), but the valuation requires SOTP sophistication (separating MPLX value/debt from the refining stub) and cycle judgment.

Undermined by foreign low-cost labor? No — capital/asset-intensive, location-bound (refineries serve regional markets; logistics moats are physical). The competitive threat is foreign refining capacity (Middle East/Asia mega-refineries) on the export margin, not labor.

Do brands matter? Minimally. Marathon and ARCO brands have some retail-channel value (especially ARCO’s West Coast position), but the product is a commodity; MPC sold its company-operated retail (Speedway) in 2021. Brand is not a meaningful moat.

Nature of competition? Competition on cost position (crude sourcing, complexity, logistics, scale, operational reliability/capture) — not on price or product. Best-cost-and-best-located wins.

Customers’ switching costs? Essentially zero for refined product (fungible commodity). High for MPLX’s midstream customers (you cannot re-route a basin’s gathering/processing) — the switching-cost moat lives in MPLX.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? Interpretation: The MPLX stake’s market value (~$36–37B) vastly exceeds its carrying basis — the largest hidden/under-recognized value. West Coast refining franchises protected by regulation also carry option value. Conversely, refining assets carry long-run stranded-asset risk.

Off-balance-sheet liabilities? Standard for the sector: environmental remediation/ARO, pension/OPEB, operating leases, and RIN obligations (~$1.33B/yr compliance cost). Nothing unusual or hidden flagged.

How conservative is the accounting? Reasonable; earnings are cash-backed (FCF tracks/exceeds GAAP NI). The main caveat is heavy reliance on management-defined “Adjusted EBITDA” in comp and guidance. LIFO inventory accounting (common in refining) can create timing distortions in volatile-price periods (e.g., backwardation).

How CapEx-hungry? Moderate at the refiner (~$1.5B/yr, ~$450M true maintenance — disciplined, no capacity builds), heavier at MPLX (~$2.4–2.7B growth) — but MPLX self-funds from its own cash flow and debt. The refiner itself is not capital-hungry relative to its cash generation.

Capital Allocation & Management

How much FCF, and how is it used? ~$4.8B FCF (2025), ~$6–14B in better years. Philosophy (explicit, margin-contingent): MPLX distributions cover the parent dividend + base capex; refining FCF funds buybacks as the primary return vehicle. ~$47B total returned to shareholders 2021–2025.

Significant acquisitions recently? Not at the refiner (last major: Andeavor 2018). All recent M&A is at MPLX (Northwind $2.4B, BANGL $703M, Whiptail — debt-funded by MPLX; offset by ~$980M divestitures). The defining divestiture was Speedway → 7-Eleven (2021, $21.38B / $17.22B after-tax).

Buying back shares? Aggressively — ~$40.8B in five years, shares −53%. $4.38B + new $5B authorization remaining. Counter-cyclically throttled (2025 fell to $3.49B as cash fell). The one risk: repurchasing at the 99th percentile of historical valuation.

Issuing shares to insiders? Minimal SBC; dilution is immaterial against the buyback. This is a de-equitization story.

Compensation policy? Returns-aware: annual bonus on relative Adj EBITDA/bbl (30%), Adj EBITDA, MPLX DCF/unit; LTI PSUs on 3-yr relative TSR (80%) + relative change in FCF/share (20%, added 2025); no scale-vanity metric. CEO Mannen 2025 pay $19.0M; say-on-pay ~93%.

Motivations of management? Interpretation: Aligned with per-share value creation by design (comp metrics), but low personal ownership (~0.32%) and near-zero open-market buying dampen the “owner-operator” signal. Professional managers executing a disciplined, well-aligned plan — not founder-owners.

Valuation & Market Data

ADR, MLP, or K-1 issuer? MPC itself is a standard US C-corp common stock (1099, no K-1). Note: its subsidiary MPLX is a publicly-traded MLP that issues K-1s to its unitholders — but MPC shareholders own MPC common, not MPLX units, so no K-1 for MPC holders.

Dividend policy? Growing per-share dividend (~$3.73 DPS, ~1.5% yield), covered by MPLX distributions; ~62% Q1-2026 total payout (dividend + buyback). Buyback is the larger and more variable return lever.

How profitable? Cyclically — see ROIC/ROE above. Thin reported net margin (~3.4%) is normal for commodity pass-through; margin-per-barrel and segment EBITDA are the right gauges.

Net income diverging from cash from operations? No persistent divergence — OCF tracks or exceeds NI through the cycle (cash-backed earnings). Working-capital swings create quarterly noise (e.g., a $573M Q1-2026 inventory build).

Risks & Downside

What would cause the stock to decline? (1) Crack-spread normalization/collapse as offline capacity returns; (2) multiple compression from the 99th-percentile level; (3) MPLX de-rate (rates/distribution stumble) removing the SOTP floor; (4) a major operational/safety incident; (5) demand-transition acceleration; (6) buyback slowdown.

Risk of catastrophic loss? Interpretation: Low. Conservatively-levered parent (~$5–7B net debt), ~half the cap in liquid MPLX value, 13-refinery diversification. Realistic downside is a 30–40% cyclical drawdown, not impairment.

Chance of total loss? Very low / negligible. No realistic path to zero given the asset base, the MPLX floor, and parent balance-sheet conservatism.

Recent News & Events

Has the business environment changed recently? Yes, sharply and temporarily: the Q1-2026 Middle East/Iran conflict took ~6M bbl/d (~6% of global refining) offline, spiking cracks and driving the stock to record highs. Management is “extremely constructive.” Domestic gasoline/diesel/jet demand strong; exports adding upside; steep backwardation a watch-item. (A curated news-sentiment scan returned no material scored articles; the event timeline was built from the 10-K, 8-Ks, and transcripts.)

Significant acquisitions / accounting changes / new facilities? New $5B buyback authorization (Q1-2026); MPLX growth projects entering service H2-2026 (Secretariat, Harmon Creek III, Titan); Garyville +30 mbpd jet (online), Robinson +10 mbpd jet (Q3-2026); CEO transition to Maryann Mannen (2024). No material accounting-policy changes flagged.

APPENDIX B — Source Appendix

Marathon Petroleum Corporation (NYSE: MPC) — Research as of 2026-06-13

Primary sources before secondary; all figures reconciled to filings where possible. Accessed 2026-06-13 unless noted.

Primary — SEC filings (EDGAR, CIK 0001510295)

1. MPC Form 10-K, FY2025 — segment adjusted EBITDA (R&M $6,138M, Midstream $6,750M, Renewable Diesel −$110M); refining footprint (13 refineries, 2,986 mbpcd; Gulf Coast 1,248 / Mid-Continent 1,186 / West Coast 552); R&M margin/bbl ($16.87 2025); crack sensitivity (~$1,125M EBITDA per $1/bbl); debt footnote (consolidated $33.31B, MPLX ~$26.01B / ~78%); buyback authorization history ($60.05B approved / $55.67B executed / $4.38B remaining); MPLX distributions to MPC ($2.56B 2025). https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001510295
2. MPC Form 10-Ks, FY2021–FY2024 — multi-year net income, OCF, capex, buyback, dividend, share-count series.
3. MPC Form 10-Q, Q1-2026 (filed ~May 2026) — Q1-2026 results (adj EPS $1.65, adj EBITDA $2.8B, R&M EBITDA/bbl $5.37, 89% utilization, 99% capture), new $5B buyback authorization, parent/MPLX cash split ($645M / $1.5B+).
4. MPC DEF 14A / PRE 14A proxy (2026) — incentive-comp metrics (relative Adj EBITDA/bbl 30%, Adj EBITDA 20%, MPLX DCF/unit; PSU on 3-yr relative TSR 80% + relative change in FCF/share 20%); CEO Mannen 2025 total pay $19.0M; say-on-pay ~93%.
5. MPC Form 4 insider corpus (326 filings, ~2021–2026) — three code-P open-market purchases (~$1.3M total); zero buys by CEO/CFO; insider ownership ~0.32%.
6. EDGAR XBRL company facts — authoritative multi-year series: NetIncomeLossAvailableToCommonStockholdersBasic, NetCashProvidedByUsedInOperatingActivities, PaymentsForRepurchaseOfCommonStock, DividendsCommonStockCash, PaymentsToAcquirePropertyPlantAndEquipment, WeightedAverageNumberOfDilutedSharesOutstanding, StockholdersEquity, LongTermDebtAndCapitalLeaseObligations.

Primary — Transcripts

7. MPC Q1-2026 earnings call (May 5, 2026) — CEO Maryann Mannen, CFO Maria Khoury. Geopolitical supply disruption (~6M bbl/d / ~6% global capacity offline); constructive macro; Q2-2026 utilization guide 94%; MPLX 12.5% distribution growth guide (two years); MPLX covers parent $1.5B capex + dividend; Garyville/Robinson/El Paso projects; inland-crude/Canadian-volume capture advantage; backwardation watch-item; new $5B buyback. (company earnings-call transcript.)
8. MPC Q4-2025 earnings call (Feb 3, 2026) — full-year 2025 results and capital-return framing. (company earnings-call transcript.)
9. MPC earnings-call archive (59 earnings calls, 16 conference presentations, 1 analyst/investor day) — historical segment and capital-allocation commentary.

Primary — MPLX LP

10. MPLX LP disclosures — distribution history/coverage; ~64% MPC ownership; MPLX standalone debt (~$26B); growth-capex program (Secretariat, Harmon Creek III, Titan, BANGL, Northwind, Gulf Coast fractionators/export JV); ~$57–58B market capitalization (basis for SOTP).

Secondary / market data

11. yfinance peer comps (via the yfinance library, ~2026-06-13) — MPC, VLO, PSX, PBF, DINO price/market cap/P-E/EV-EBITDA/P-S/dividend yield. Unofficial; reconciled to filings where material. Note: consolidated EV/EBITDA overstated by MPLX-debt inclusion.
12. Third-party fundamentals & own-history valuation-percentile data — snapshot (sector/GICS, employees, short interest 2.35% float, insiders 0.32%, institutions 78%, beta 0.53); valuation_index own-history percentiles (P/B 99.9th, P/S 99.9th, P/E 81.7th, composite 93.8th). Third-party signal; valuation percentiles compared only to MPC’s own history. A curated news-sentiment scan returned no material scored articles.

Notes on data quality

• Some third-party three-statement data arrays returned stale pre-2014 figures for MPC; all financial series in this report are sourced from EDGAR XBRL.
• “Adjusted EBITDA” and segment figures are management-defined non-GAAP; treated as reported but flagged as such.
• Management forward commentary (crack outlook, MPLX growth, mid-cycle margin) is treated as hypothesis, validated against filings/financials where possible and labeled Interpretation/Assumption in the memo.