Marathon Petroleum Corporation
Orin's own judgment of how firmly the evidence points one way — not the chance the price goes up.
MPC has rallied to $455.03 with RSI at 73.5 (overbought) after Q2 2026's blowout EPS of $17.68 on revenue of $52.0B (up 53.8% YoY), beating estimates by 24.2%, and the WSJ reports refiners are set to crush those results again in Q3.
However, the newly announced Trump-Putin diesel agreement introduces a potential supply-side margin compressor just as the stock sits well above its 50-day MA of $378.97, insiders remain net sellers (-$12.8M over 24 months), and total debt of $34.4B against equity of $17.3B constrains financial flexibility. With a P/E of 15.66 trading only 7.2% below the peer median, the valuation already discounts near-peak refining margins, making the risk/reward balanced but fragile into a shifting supply landscape.
What could go wrong
- Russian diesel supply shock. Trump's October 9 announcement that Russia will send millions of tons of diesel to world markets could compress the crack spreads that drove Q2 2026's exceptional margins.
- Overbought technicals. RSI at 73.49 and the stock trading well above both the 50-day MA ($378.97) and 200-day MA ($268.59) leave it vulnerable to a sharp pullback on any negative margin news.
- Balance-sheet leverage. Total debt of $34.4B against equity of $17.3B, with FCF declining from $12.2B in 2023 to $4.8B in 2025, limits flexibility if the refining cycle turns.
- Insider and smart-money selling. Insiders net sold $12.8M over 24 months with zero open-market acquisitions, and the smart-money score is slightly negative at -0.03 as of Q2 2026.
What would change my mind
Where this comes from: Q2 2026 results and earnings vs estimates · technicals, 9 Oct · peer comparison · FY23 and FY25 fundamentals. Orin's read on MPC; not advice.
How has the share price moved?
Up 122% over twelve months, from $196.23 on 27 Oct 2025 to $435.98 at the Tue 27 Oct 2026 close. Its highest close in that time was $484.92 on 14 Oct 2026; the last close is 10.1% below it.
- Green line: the closing price each trading day.
- Grey line: the average closing price over the last 50 trading days.
- Dashed amber line: the same over the last 200 trading days — a slower, smoother line.
- Bars along the bottom: shares traded each day — green when it closed higher than the day before, rose when lower (the most, 9.9M, on 18 Sep 2026).
- Flags: the company's filings, on the day they were made (a 10-Q is the quarterly report and a 10-K the annual report; an 8-K reports a major event).
On file
All MPC filingsLargest holders · 13F Q2 2026
Money flow| Filer | Position | Of fund |
|---|---|---|
| BlackRock | $6.6B | 0.1% |
| Vanguard Capital Management | $4.9B | 0.1% |
| State Street | $4.7B | 0.1% |
| Vanguard Portfolio Management | $4.0B | 0.2% |
| Geode Capital Management | $2.1B | 0.1% |
| Morgan Stanley | $1.5B | 0.1% |
| FMR | $1.4B | 0.1% |
| Bank of New York Mellon | $1.0B | 0.2% |
| Wellington Management Group | $831.5M | 0.1% |
| Northern Trust | $799.2M | 0.1% |
125 of the 815 filers we track · quarter ended Q2 2026 · positions, not flows
Insiders, last 24 months: 90 Form 4 filings; open-market trades net −$12.8M. Of the 50 on hand, 0 were open-market purchases and 14 sales — the rest are grants, option exercises and tax withholding.
Ask Orin about MPC
Orin answers questions about MPC from its filings, transcripts and numbers — with the source behind every claim. Your own desk starts with 3 free days, card required.
Start 3 free days →How is it priced, against its own past and its sector?
Priced at 15.0× its last 12 months of profit (Tue 27 Oct close); the typical Energy company on the NYSE is at 17.6×.
| Measure | Now | Its usual | Sector |
|---|---|---|---|
| P/E | 15.0× | — | 17.6× |
| P/S | 0.80×above usual | 0.34× | — |
| EV/EBITDA | 8.5×above usual | 6.9× | — |
| P/B | 6.7×above usual | 2.5× | — |
- P/E: no usual level — it ran from 4.1× to 23.6× over its last 5 fiscal years, so a middle value says little.
- "Its usual" is the middle of its own last 5 fiscal years.
P/E: the share price divided by one year of profit per share — roughly how many years of today's profit the price pays for. A company that loses money has no P/E.
P/S: the company's total market value divided by one year of revenue. Used when there is no profit to measure.
EV/EBITDA: the price of the whole company (its shares plus its debt, minus its cash) divided by one year of operating profit before interest, tax and write-downs.
P/B: the share price divided by the book value per share — what the company owns minus what it owes, as its accounts record it. When book value is tiny, P/B swings wildly and says little.
What do the analysts who cover it publish?
The analysts' average 12-month price target is $444 (from $321 to $560) — 2% above the Tue 27 Oct close price. That is their published opinion, not Orin's.
$444
$321 – $560 · +2% against the last close
- 10 buy or overweight
- 6 hold
- 0 underweight or sell
Price target: the share price an analyst publishes for about 12 months ahead. It is the opinion of the firm that wrote it; the average is across the firms that cover the company.
Overweight / underweight: analysts' rating words: overweight means the firm expects the stock to do better than the others it covers, underweight worse. Their words, not Orin's.
How is it priced next to the companies it is compared to?
Priced at 15.0× its last 12 months of profit. Only 2 of the 2 companies it is compared to made a profit (PSX 15.2×, VLO 8.7×), so there is no typical P/E.
| Company | Market value | P/E | EV/EBITDA | Gross margin | Net margin | ROE |
|---|---|---|---|---|---|---|
| MPCMarathon Petroleum Corporation | $122B | 15.0× | 8.5× | 11.6% | 5.6% | 49% |
| PSXPhillips 66 | $108B | 15.2× | 9.7× | 9.8% | 4.6% | 24% |
| VLOValero Energy Corporation | $112B | 8.7× | 6.6× | 15.5% | 8.8% | 50% |
- The list of companies is the data vendor's choice, so it can include a company a person would not have picked.
Market value: the share price times the number of shares — what the whole company is priced at.
P/E: the share price divided by one year of profit per share — roughly how many years of today's profit the price pays for. A company that loses money has no P/E.
EV/EBITDA: the price of the whole company (its shares plus its debt, minus its cash) divided by one year of operating profit before interest, tax and write-downs.
Gross margin: the share of each $1 of sales left after the direct cost of making what was sold.
Net margin: the share of each $1 of sales kept as profit after every cost. Below zero, the company lost money.
ROE: return on equity: one year of profit as a share of the shareholders' money in the business (its book value).
Is it priced high or low compared with its own past?
Priced at 15.0× its last 12 months of profit (Tue 27 Oct close). It has no full fiscal year on file since its history starts, so there is nothing of its own to compare with.
- Now15.0×
P/E: the share price divided by one year of profit per share — roughly how many years of today's profit the price pays for. A company that loses money has no P/E.
Fiscal year (FY): the company's own 12-month accounting year, which need not match the calendar year.
How has its sector been priced?
Energy companies on the NYSE were priced at a typical 17.6× their yearly profit (Tue 27 Oct close).
- Now17.6×
- Each year is the middle of that calendar year's daily sector P/E; "Now" is the Tue 27 Oct close figure, the same one in the table above.
Sector P/E: the middle P/E of the companies in the same sector on the same stock exchange, as the data vendor works it out each day.
What does a cash-flow model say a share is worth?
A standard cash-flow model puts it at $411 a share — 6% below the $439.17 price it was computed against (27 Oct).
| Model | Value a share | Against the price then |
|---|---|---|
| Standard | $411 | 6% below $439.17 |
| Levered | $366 | 17% below $439.17 |
Discounted cash flow: a model's estimate of what the company is worth today, from the cash it is forecast to make in future years. Small changes in the forecast move it a lot, and it says little about a company that loses money.
Standard / levered: two versions of the cash-flow model: the standard one counts the cash before the company pays its lenders, the levered one after.