Marathon Petroleum: A Deep-Value Investment Study

At the industry level, I would rather define it as: a leading operator in a mature industry, not a capital-light good company in a good industry. The EIA discloses that as of January 1, 2025, the United States had 132 operable refineries, the same as in 2024, with almost no major new capacity added in 2025. The three largest refiners, Marathon, Valero and ExxonMobil, each added less than 1% of capacity, mostly from small-scale process optimization rather than new large refineries. In other words, this industry has very high barriers to entry, but the reason is mainly capital, environmental and permitting constraints, not consumers willing to pay a large premium for a brand.

Long-term demand is not about to "disappear immediately," but it is far from risk-free. The EIA's current outlook holds that, in its May 2026 base case, global liquid-fuels consumption rises from 104.0 million barrels per day in 2025 to 104.2 million barrels per day in 2026, then to 105.6 million barrels per day in 2027. But that same forecast explicitly states that the current 2026 projection is heavily affected by Middle East supply disruption, a Strait of Hormuz shutdown, and high oil prices suppressing demand, making it an abnormal environment. For MPC, this means near-term high profits do not necessarily signal long-term high quality.

How to assess the moat? My judgment is "there are barriers, but it is not a classic Buffett-style moat." Brand advantage: present, but limited. Marathon and ARCO help maintain channel coverage and brand recognition, but end fuel consumers care more about location, price and convenience than irrational brand loyalty. Cost advantage: partly present. A large-scale refining system of 3.0 million barrels per day, complex refineries, integrated logistics, and crude procurement and scheduling capability do allow it to achieve better-than-peer unit margins in certain cycles. Scale advantage: present. The company owns a vast refining, terminal, shipping and midstream network, and MPLX ties the logistics network into the system through long-term agreements. Network effects: weak. This is more like physical-infrastructure synergy than internet-style network effects. Switching costs: weak for end consumers, stronger for logistics and production coordination within the system. License/regulatory barriers: strong. Not because a permit lets you earn windfall profits, but because it is very hard for others to replicate. Operating capability and capital allocation: this is the most important "soft moat." In recent years MPC has continued to improve costs, refocused on refining and midstream after exiting Speedway, and kept buying back stock in large amounts.

On balance, I assign the following scores: Business understandability: 4/5 Industry attractiveness: 2/5 Moat strength: 3/5

If the stock market closed for five years, would I be willing to hold it? [View] Only if the purchase price is low enough. If the buy price implies "the median profit of a normal year, rather than the optimistic profit of a high-oil-price year," I would be willing to hold it as a real cash-flow business. But near the current price, I would not hold it as comfortably as I would Coca-Cola or a first-class software company. My one-line summary is: MPC is more like "an excellent company in a poor industry, plus a decent midstream asset." That is praise, but not the highest grade of praise.

Management and Capital Allocation

On management, MPC has completed a smooth leadership transition over the past two years. Maryann T. Mannen became CEO in August 2024 and became Chair and CEO in 2026; Maria A. Khoury became Executive Vice President and CFO on January 19, 2026. Public records show Mannen has a long-term energy and finance background, while Khoury's resume leans toward large-industrial financial management in the GE/Danaher tradition. The governance pages also disclose a complete governance framework, including an independent Lead Director, proxy-access provisions, whistleblower and related-party transaction policies, and a code of ethics for senior financial officers. On whether "the people are professional and the organization behaves like a large company with capital discipline," my answer is yes.

What truly determines management quality is not resumes but capital allocation. On this point, MPC's record over the past three years is quite strong. The company discloses that it repurchased 89 million, 53 million and 21 million shares in 2023, 2024 and 2025, respectively, paying roughly $11.572 billion, $9.077 billion and $3.399 billion in cash, at average buyback prices of about $131.27, $171.68 and $163.64 per share. By year-end 2025, cumulative board-authorized buyback capacity since 2012 reached $60.05 billion, of which $55.67 billion had been executed. As of March 31, 2026, the company had $3.6 billion of prior buyback authorization remaining; the board then added a new $5.0 billion authorization, bringing total available buyback capacity to $8.6 billion.

Dividends have likewise risen steadily. MPC's common-stock cash dividend per share rose from $3.075 in 2023 to $3.385 in 2024, then to $3.73 in 2025. In the first quarter of 2026 it declared a quarterly dividend of $1.00 per share, an implied yield of about 1.6% at the current price. This shows management treats "buybacks + a growing dividend" as the main form of capital return, rather than pouring all cash into high-risk expansion.

What deserves a closer look is whether these buybacks were rational. With hindsight, most buyback prices were below the current stock price, so they were not obviously value-destroying. And the combined 2023-2025 buyback scale was enormous, clearly shrinking the share count. By the company's disclosed weighted-average shares outstanding, MPC went from 407 million shares in 2023 to 305 million shares in 2025, falling further to 295 million shares in the first quarter of 2026. This meaningfully lifts both earnings per share and owner earnings per share. Still, I keep one cautious caveat: [View] a good buyback history does not mean it is right to keep buying at any price today. If the company keeps buying aggressively in a high-cycle, high-valuation zone, that would slide from "excellent capital allocation" toward "pro-cyclically dressing up per-share figures."

On the alignment between management and shareholder interests, I can confirm two points and must honestly flag one shortcoming. First, the public Form 4 summary shows that after a stock grant and tax-withholding transaction in March 2026, Mannen directly held about 111,847 shares of MPC common stock, indicating real economic exposure, though far from a founder-style high-ownership structure. Second, the company's website and governance pages show a relatively complete governance system. Third, this round of research could not reliably extract the full "combined management and director ownership" table from the 2026 proxy, so I do not want to fabricate a precise ratio. The more prudent way to state "whether they hold a large stake" is this: management and shareholders are not misaligned, but that alignment comes more from compensation design, reputation in office and continued buybacks than from a very high absolute ownership ratio.

Taken together, I assign: Management and capital allocation score: 4/5. Not because I think they are perfect, but because in such a capital-intensive, highly cyclical industry, MPC's management has at least shown a clear shareholder-return mindset, fairly strong financial discipline, and a strategy of building the midstream platform into a more stable cash-flow base. That is where I am willing to award a high score.

Financial Quality and Owner Earnings

Let me start with a table that includes only confirmed data. To avoid the problem of a table that "looks complete but is actually laced with guesses," the table below mainly uses 2023-2025 audited data and disclosed first-quarter 2026 data. For any item this research round could not confirm on a consistent basis, I clearly mark it as "additional data required."

• 2026Q1 capex approximates the disclosed 10-Q figure of $913 million in PP&E additions combined with investment/acquisition and similar cash flows; the single-quarter FCF is for reference only and should not be annualized. ** 2024 total debt is the year-end 2024 sum of 3.049 + 24.432. Data is drawn mainly from MPC's 2025 10-K and first-quarter 2026 report; capex uses the company's disclosed "capital expenditures and investments" basis.

From this table, the most important takeaway is not the level of profit but several structural signals. First, accounting profit and cash flow broadly match, and cash is even better. OCF was $14.117 billion, $8.665 billion and $8.253 billion in 2023-2025, with corresponding FCF of roughly $11.618 billion, $5.645 billion and $3.654 billion. Even on the stricter basis that "includes growth capex," MPC did not show the "earnings look great but cash is poor" problem in most years. Second, share-count compression is very clear, and it was not a fake boom driven by issuing shares. Third, profit volatility is extreme, showing this is by no means a company that "becomes more stable as it grows." Net income in 2025 was only a bit over 40% of 2023's level, and in the first quarter of 2025 the company was even in a loss position, recovering only in the first quarter of 2026.

For the profit quality of the refining core, the key is "profit per barrel" and sensitivity to the market environment. MPC's disclosed Refining & Marketing margin fell from $24.372 billion in 2023 to $17.121 billion in 2024, then recovered to $18.404 billion in 2025, far below the 2023 high-cycle peak. On a per-barrel basis, the 2024 figure was $16.01 per barrel, clearly below the $23.00 per barrel of 2023. The first quarter of 2026 improved to $17.74 per barrel versus the first quarter of 2025, driven mainly by higher crack spreads. This trajectory tells us clearly: the high margins of the past were more a cyclical windfall than a structurally high gross margin created by consumer brands.

Turning to the balance sheet, the conclusion is: it can survive, but not easily. As of March 31, 2026, consolidated total assets were $88.187 billion, total liabilities $64.760 billion, total debt $33.272 billion, and cash $2.151 billion. Of this, MPC parent-company debt (excluding MPLX) was about $7.266 billion, while cash and short-term investments excluding MPLX were about $645 million, and the parent company's $5.0 billion revolving credit facility was undrawn. This shows that short-term liquidity is not fragile; the real "sense of leverage" lies more in the consolidated picture and MPLX's expansion than in any imminent financial crisis at the parent.

Swings in receivables, inventory and payables also look more like a commodity business than accounting manipulation. At year-end 2025, accounts receivable fell from $11.145 billion to $10.317 billion, inventory rose from $9.568 billion to $10.129 billion, and accounts payable fell from $13.906 billion to $12.974 billion. By the first quarter of 2026, driven by changes in crude and product prices and volumes, receivables jumped to $14.629 billion, inventory rose to $10.764 billion, and payables rose to $17.617 billion. The company explains in its 10-Q that these changes are driven mainly by period-end commodity prices and volume changes. For a refiner, this is an operating characteristic, not an automatic sign of fraud.

The Owner Earnings analysis is the most critical step here. If you look only at GAAP FCF, you conclude "it is not cheap today." If you look only at 2023's high-cycle profit, you might conclude "it makes a lot of money, so perhaps it is not expensive." Neither view is complete. My conservative approach is this:

• [Fact] 2025 OCF = $8.253 billion;

• [Assumption] for maintenance capex, use the company's 2026 guidance of $450 million in refining maintenance capex, plus MPLX's 2026 guidance of $300 million in maintenance capex, then conservatively add $50 million for corporate-level and other miscellaneous items, for a total of about $800 million;

• [Fact] for cash not freely distributable to MPC common shareholders, also subtract $1.507 billion of distributions to non-controlling interests; This produces 2025 owner earnings of roughly $5.95 billion.

Why is this number much higher than 2025's strict FCF? Because a large share of the company's actual 2025 capex went into growth projects, rather than simply what was needed to keep the existing-asset business running. MPC itself discloses that within the 2026 Refining & Marketing capital budget of $1.41 billion, about $710 million is for value-enhancing projects and $250 million for marketing investment, while the part explicitly labeled maintenance capital is $450 million. Within MPLX's 2026 capital-spending outlook of $2.7 billion, there is also a clear $2.4 billion of growth capex and $300 million of maintenance capex. This means: MPC's real distributable cash capability is usually higher than the FCF you get when "all growth capex is treated as required spending."

But as a conservative investor, I will not treat $5.95 billion as a long-term norm. Because refining profit is inherently cyclical, and 2025 was not a "complete trough year" either. So my conservative long-term Owner Earnings figure drops to $4.5-5.0 billion per year; neutral is $5.0-5.5 billion per year; optimistic is $5.5-6.0 billion per year. Against the current equity market value of about $75.1 billion, the market's valuation of MPC roughly corresponds to:

• about 12.5x approximate 2025 owner earnings;

• about 13.7-15.0x neutral normalized owner earnings;

• about 20.6x strict 2025 FCF. For a "high-volatility refining + stable midstream" combination, this looks more like fairly valued to slightly expensive than a cigar butt.

In this financial-quality section, my final conclusion is: [View] Most of MPC's profit is real cash profit; it does not, like some heavy-asset companies, prop up a story on accounting profit over the long run. There are no clear signs of aggressive accounting or profit manipulation. The real issue is not "whether profit is real," but "whether profit durability matches the valuation price." This is precisely the dividing line between value investing and financial analysis.

Valuation, Margin of Safety, and Opportunity Comparison

Valuation must first address a real-world problem: the current macro environment is abnormal. At the most recently available price, MPC trades at about $254.65. At the same time, the EIA's May 2026 outlook holds that, in its current disruption scenario, Brent could average $95 per barrel in 2026, noting that the Strait of Hormuz shutdown and Middle East production cuts add a clear risk premium to near-term oil prices. This backdrop affects both refining margins and market risk appetite. Therefore, no valuation should linearly extrapolate the near-term tense environment.

Owner Earnings Discount Method

Below is the three-scenario valuation I use. The starting point is not GAAP net income but normalized owner earnings.

• Conservative scenario: [Assumption] starting Owner Earnings of $4.5 billion, zero growth over the next 10 years, a discount rate of 10%, and terminal growth of 1%.

• Neutral scenario: [Assumption] starting Owner Earnings of $5.0-5.5 billion, 2% growth over the next 10 years, a discount rate of 10%, and terminal growth of 1%.

• Optimistic scenario: [Assumption] starting Owner Earnings of $5.5-6.0 billion, 3% growth over the next 10 years, a discount rate of 10%, and terminal growth of 1%. The starting ranges follow from the 2025 cash-flow, maintenance-capex and non-controlling-distribution treatment discussed above; the discount rate is set to the return a balanced, conservatively biased investor should require for a cyclical stock.

Under these assumptions, the current intrinsic-value ranges I derive are roughly:

• Conservative intrinsic-value range: $160-190 per share

• Fair intrinsic-value range: $200-235 per share

• Optimistic intrinsic-value range: $240-265 per share

In other words, the current stock price sits roughly near the top edge of the optimistic scenario. For a conservative investor, this is not a margin of safety; it is closer to "paying upfront for a good outcome." [Inference] If you require at least a 20%-30% margin of safety, the ideal buy price should be $170-205. This is not to say $255 will immediately fall there; it is to say that, from the perspective of a long-term business owner, this price band is closer to buying insurance against uncertainty.

Relative Valuation Method

Start with the directly verifiable market multiples today. MPC currently trades at a P/E of about 16.8x; Valero 18.0x; Phillips 66 17.6x; HF Sinclair 10.5x; PBF Energy 10.6x. On P/E alone, MPC is a touch cheaper than the highest-quality large peers, but not by much; relative to second- and third-tier high-volatility refiners, it is clearly more expensive. This shows the market is already assigning a premium for its scale, MPLX value and capital-allocation capability.

Several of MPC's own multiples are more illuminating. Using shareholders' equity attributable to parent of $16.753 billion at the end of the first quarter of 2026, the current P/B is about 4.5x; but this figure is severely distorted by years of large buybacks and has limited reference value. Using 2025 OCF minus the company's disclosed 2025 capital expenditures and investments, the current P/FCF is about 20.6x, which is not cheap for a cyclical refiner. If instead I use my conservatively estimated 2025 owner earnings of $5.95 billion above, the current P/OE is about 12.6x, closer to fair. My reading is: the strict FCF lens looks expensive, the Owner Earnings lens looks close to fair, but it is by no means clearly undervalued.

There is also a very useful breakdown: carve out MPLX. As of the end of 2025, MPC held roughly 64% of MPLX's common units. At MPLX's current market value of about $57.3 billion, MPC's stake in MPLX is worth about $36.7 billion. Strip that out of MPC's current market value of about $75.1 billion, then add back the rough net debt of MPC's parent company (excluding MPLX), and the market assigns the "non-MPLX part," namely refining, marketing, renewable diesel and the corporate level, an implied enterprise value of roughly $45 billion. Offsetting this roughly against 2025 refining EBITDA of $6.138 billion, renewable diesel of -$110 million and corporate expenses of -$927 million, the non-MPLX business carries an implied EV/EBITDA in the high single digits. That is not extremely expensive, but for a highly cyclical refining core, it is by no means cheap. Here I explicitly label it [Inference] rather than fact, because this is a SOTP approximation I built from public data.

Asset and Liquidation Value Method

On book value alone, MPC's shareholders' equity attributable to parent at the end of the first quarter of 2026 was about $16.753 billion, far below its market value. But that does not mean it is wildly expensive, because book value has been pushed down by years of large buybacks, and the replacement cost of large refineries, pipelines, terminals and brand networks differs from their book net value. On the other hand, the liquidation value of refining assets can be very poor in an industry downturn, because there are also costs such as environmental remediation, shutdowns, debottlenecking reversal and regional policy change. Therefore, for MPC, book net assets are neither a good anchor nor a strong cushion; the more reasonable approach is a SOTP using "the value of the publicly listed MPLX stake + a mid-cycle valuation of the non-midstream business."

Margin of Safety and Opportunity Comparison

Is the current price cheap enough? My answer is no. The most fragile assumption in MPC's valuation is that the market is willing to keep the current better refining environment and higher midstream valuation in place. If future growth comes in below expectations, crack spreads roll over, RIN costs rise, or the market once again treats it as a "plain refiner stock" rather than a "refiner stock + midstream platform," then today's buy-in return can easily slip to the low single digits.

Compared with alternatives, the gap is not small either. At the most recently available market quotes, the U.S. 10-year Treasury yields about 4.5%. For a balanced, conservatively biased investor, if a cyclical refining stock can offer only about 4%-6% long-term annualized expected return in the neutral scenario, then its compensation over the risk-free rate is clearly thin. Only when you can buy it at a cheaper price, or when you have greater conviction in its buyback-driven per-share growth over the next decade, will MPC clearly beat owning bonds or a broad-based index. By contrast, a broad-based index may not be cheaper, but it offers greater diversification and lower exposure to any single policy or single commodity price.

Therefore, I offer the following price framework:

• Conservative intrinsic-value range: $160-190 per share

• Fair intrinsic-value range: $200-235 per share

• Optimistic intrinsic-value range: $240-265 per share

• Ideal buy-price range: $170-205 per share

• Acceptable holding-price range: $205-240 per share

• Clearly overvalued price range: above $260

On the current $254.65, my conclusion is: expensive under the conservative/neutral framework, near the top edge of fair under the optimistic framework, and therefore still "without sufficient margin of safety" for a conservative investor.

Risks, Counterarguments, and Checklist

Let me start with the strongest counterargument, which I find quite compelling: "MPC looks like a value stock, but it is often just a refining stock dressed up by a good cycle and large buybacks. The truly high-quality, low-volatility, toll-road-like assets are in MPLX, and MPLX does not belong 100% to MPC common shareholders. What you buy today is not a consumer company with strong pricing power, but a high-quality yet still highly cyclical energy platform." If this counterargument is right, then the problem with the current price is not "the company is bad" but "a good company at a bad price."

The most important risks, listed through the lens of "permanent loss of capital," are: First, cyclical risk. The company discloses that a $1 per barrel change in the crack spread affects annual refining EBITDA by about $1.125 billion, meaning earnings are very sensitive to the external environment. Second, regulatory and compliance risk. RIN purchase expense reached $1.33 billion in 2025, and the company explicitly warns of possible stricter environmental regulations, price/profit/inventory-related constraints in places like California, and ongoing climate litigation. Third, the risk of the business model being eroded by slow-moving variables. Long-term fuel demand may not fall off a cliff, but EV penetration, efficiency gains, low-carbon policy and regional demand shifts will lower the industry's "long-term fair multiple." Fourth, the risk of pro-cyclical capital allocation. Continuing large buybacks in a high-cycle, high-valuation environment would turn an otherwise excellent capital-allocation record into a drag on future returns. Fifth, renewable-diesel capital returns falling short of expectations. The renewable-diesel segment lost money continuously in 2023-2025, with 2025 adjusted EBITDA of -$110 million.

What facts would make me admit this investment judgment was wrong? First, if over the next two to three years MPLX's "stabilizer" role for MPC weakens, for example if MPLX's capital returns deteriorate, leverage expands too quickly, or MPC's ownership stake falls noticeably, while the refining core cannot fill the volatility gap. Second, if the central level of refining profit shifts down noticeably, showing up as MPC failing to earn meaningful free cash flow over the long run even in years without major shutdowns or accidents. Third, if management starts using more debt to support high-priced buybacks rather than acting against the grain when the stock is undervalued. Fourth, if regulatory events occur that are enough to reshape the industry's economics, such as higher RIN costs, spillover of California-style constraints, or new environmental-liability pressure.

What is the biggest permanent-capital-loss scenario? [Inference] If you buy at today's price and over the next decade you get "slowly weakening product demand + crack spreads mean-reverting to low levels + declining MPLX growth-capital returns + the market discounting cyclical-stock valuations," then it is not unimaginable that the stock stays in the $160-200 range for a long time. This means that even if the company keeps paying dividends, your ten-year annualized return could be only low single digits or close to zero. In opportunity-cost terms, that is a permanent loss of capital. In a worse downside scenario, a 40%-60% interim drawdown over the short to medium term is not rare in the refining industry.

Below is the result of my investment checklist. It is not a "fact table" but an integrated judgment based on the evidence above:

Open questions and limitations: This report has prioritized MPC's latest 10-K, 10-Q, website governance materials, and EIA and Treasury data, but two points still require additional verification. One, the combined management and board ownership table in the 2026 proxy was not reliably extracted this round, so I have not given a "precise combined ownership ratio." Two, a full, consistently defined peer table of P/E, P/B, EV/EBITDA, P/FCF and ROIC could not be fully reconstructed without bringing in more third-party databases, so for the relative-valuation section I used a combined method of "currently verifiable P/E + MPC's own P/FCF/P/OE + a SOTP approximation." This does not change my judgment on the "broad direction," but it does affect the precision of "how much cheaper than peers."

Final Investment Conclusion

[Final Rating] Watch

[One-Line Investment Thesis] MPC is a platform-type refiner that operates well, allocates capital strongly and holds high-quality midstream assets, but it still belongs to a highly cyclical industry, and the current price already prices in plenty of good news. For a balanced, conservatively biased long-term investor, what is most lacking is not a story but a margin of safety.

[Core Bull Case]

• It owns one of the largest refining systems in the United States, with refining capacity of about 3 million barrels per day, supported by a broad terminal and logistics network.

• Through MPLX it owns a more stable, fee-based midstream cash-flow base, with clear internal logistics synergy.

• Cash-flow quality over 2023-2025 was broadly good; profits were not "paper profits"; share-count reduction was significant.

• Over the past three years management has bought back stock in large amounts and steadily raised the dividend, with a capital-allocation record better than most refining peers.

• New U.S. refining capacity is constrained, and replicating its 3.0 mbpd scale and network would take years and enormous capital.

[Core Bear Case]

• Refining profit is extremely sensitive to the external price environment, where a $1 per barrel change in the crack spread is roughly a $1.125 billion change in EBITDA.

• At the current stock price, P/E, P/FCF and DCF all show no clear undervaluation.

• The truly most stable midstream asset does not belong 100% to MPC common shareholders; minority holders share part of MPLX's value.

• The renewable-diesel segment has been a drag for the past three years and has not yet proven itself a high-return reinvestment channel.

• It faces long-term, multi-front pressure from RIN costs, environmental rules, climate litigation and the slow-moving downward variable of fuel demand.

[Key Assumptions]

• MPLX's cash-flow quality and central valuation stay stable.

• The central level of refining profit does not shift down sharply from 2024-2025.

• Management continues to repurchase stock on a value-driven basis, rather than mechanically buying back when overvalued.

• RIN and environmental policy costs do not deteriorate enough to reshape the industry's economics.

• The market does not capitalize 2026's abnormal geopolitical-disruption profits over the long run.

[Fair Buy Price] $170-205 per share. Basis: this range roughly corresponds to a buy framework that applies a 15%-25% discount to my fair intrinsic value of $200-235 per share, better matching the room for error a balanced, conservatively biased investor needs for a cyclical stock.

[Target Holding Period] If bought at a price with a margin of safety, it suits 5-10 years or more; if bought near the current price, it looks more like "waiting for the cycle to deliver + waiting for the price to offer room for error" than "comfortably winning by lying low for the long run."

[Expected Annualized Return] This is an estimate based on the current price, not a price forecast, but a scenario-return framework:

• Conservative scenario: 0%-2% per year

• Neutral scenario: 4%-6% per year

• Optimistic scenario: 7%-10% per year In my view, such a neutral return is not enough to clearly beat the current U.S. 10-year Treasury yield of about 4.5%, so new buying requires a lower price.

[Maximum Loss Risk] If a multi-year low-cycle environment, a downward shift in central profit and a contraction in valuation multiples occur, the stock could fall to $160-200 and stay there for a long time; an extreme interim downturn of 40%-60% cannot be ruled out. The reason is not that the company will go bankrupt immediately, but that today you may be overpaying for a cyclical asset that should be bought at low levels.

[Tracking Metrics]

• Refining & Marketing margin per barrel

• Crack spreads, crude-grade differentials, natural gas costs

• RIN purchase costs

• The 2026-2027 ratio of maintenance to growth capex at MPC and MPLX

• MPLX leverage and distribution coverage

• The gap between attributable OCF, strict FCF and Owner Earnings

• Changes in the share count and the average buyback price

• Improvement in the Renewable Diesel segment's EBITDA

• Progress of regulatory and climate litigation

• Changes in parent-company (excluding MPLX) net debt and credit-facility usage

[Signals That Trigger Reassessment]

• Refining profit clearly below mid-cycle levels for several consecutive quarters

• Deterioration in MPLX capital returns or a marked rise in distribution pressure from external minority holders

• Buybacks continuing on a very large scale during a high-valuation phase

• A significant deterioration in compliance costs or the regulatory environment

• Renewable diesel continuing to lose money over the long run with no way to prove its strategic rationale

• Asset accidents, shutdowns, environmental compensation, or adverse developments in major litigation

[Final Recommendation] To put it plainly, MPC is worth studying and worth keeping near the top of a watchlist, but it is not worth rushing into at the current price. If you already hold it at a low cost, I lean toward viewing it as an asset you can hold, but must keep watching for the cycle and capital-allocation discipline; if you have not yet bought, my advice is: wait for a better price, rather than chasing a "quality cyclical stock" that is already no longer cheap. Real value investing is not about finding "stocks that look cheap," but about finding businesses that "can still do well even when the room for error is large enough." I largely accept MPC's "doing well," but I do not accept its "room for error being large enough," at least not at today's price.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.