MPLX Lp MPLX

Reference table

1. What they actually own — two segments, a lot of steel, and some barges

MPLX LP (NYSE: MPLX, incorporated 2012, headquartered at 200 East Hardin Street, Findlay, Ohio; CEO Maryann T. Mannen; 5,762 full-time employees; IPO'd 2012-10-26) is, in its own description, "a subsidiary of Marathon Petroleum Corporation, with MPLX GP LLC serving as its general partner."

The business splits into two reported segments:

Revenue by type (FY25, the most recent product disclosure):

Data caveat: those named lines total $6.684B against $11.817B of FY25 income-statement revenue — a $5.133B gap the dataset does not explain. Treat the split as a mix signal only: it says the business is roughly two-thirds fee-for-service and one-third product, which is the right shape for a partnership (contracted fees, not commodity length), and it should not be used for any revenue bridge.

Second data caveat: there is NO geographic segmentation in this dataset — the field is empty. MPLX does not provide geographic segment reporting here. No geographic analysis is offered, and none is implied.

Third data caveat, and worth noting: the segment reporting format changed. Pre-2018 filings show a clean two-way split (FY17: Gathering and Processing $6.324B, Logistics and Storage $3.699B) that maps directly to the described segments; from FY18 onward the disclosure switches to Product/Service/Other. The two formats are not comparable, so no multi-year segment trend is drawn.

The asset base: net property, plant and equipment $22.213B, goodwill $8.755B (20.4% of total assets — a large intangible reflecting years of acquisitions), total assets $43.005B. Intangibles are 23.5% of total assets. That goodwill number is the balance-sheet fingerprint of a partnership that has grown substantially by purchase rather than by construction — the FY25 cash-flow statement shows $4.031B of acquisitions against just $1.808B of capex.

2. The Q2'26 print — landed this morning, and the data row is partly corrupt

MPLX released Q2'26 on 2026-08-04, the same day this dive was written. The earnings-calendar record is reliable; the income-statement row is not, and this section says exactly why.

What the earnings calendar reports for 2026-08-04: EPS $1.06 actual against $1.06 estimated (exactly in line) and revenue $3.487B actual against $3.142B estimated (an 11.0% beat).

What the income-statement row for 2026-06-30 reports, with the problems flagged:

This is a data-integrity failure in the vendor feed, not a business event, and the pattern makes it obvious: four of the corrupted fields are exact sign-flipped duplicates of the Q1'26 values. Every corrupted field is excluded from this dive. The figures used are revenue $3.487B, operating income $1.568B, net income $1.097B and EPS $1.06 (from the earnings calendar), with the unit count taken from Q1'26 at 1,017M. This is flagged loudly rather than smoothed over because a reader checking the raw file would otherwise see a headline loss where none exists.

What the clean fields say: revenue +25.1%, operating income +21.3%, net income +4.7%. The gap between operating income growth (+21.3%) and net income growth (+4.7%) is the interest bill on FY25's acquisition debt showing up — exactly the dynamic that drives the consensus FY26 EPS decline discussed in §4.

The prior quarter, for context. Q1'26 (2026-03-31): revenue $2.863B versus $2.887B in Q1'25 (-0.8%), operating income $1.024B versus $1.366B (-25.0%), EPS $0.90 versus $1.10 — and the earnings calendar records a miss ($0.90 actual against $1.05 estimated). So H1'26 has been mixed: a weak Q1, an in-line Q2 with a revenue beat. H1'26 EPS totals $1.96 against a full-year consensus of $4.28792, implying $2.33 in H2 — the second half has to accelerate materially for consensus to hold. That is the near-term number to watch.

3. The financials — high margin, cash-generative, acquisition-funded

Annual series:

Revenue up 12.1% over three years; net income up 24.5%; EPS up 28.5%. That is a business with genuine, if unspectacular, operating leverage and a slowly shrinking unit count. Note that FY25 operating income fell 9.9% while EBITDA rose 10.6% and net income rose 13.8% — the operating-expense line swung from -$471M (FY24) to +$583M (FY25), a $1.054B move that this dataset does not explain. Treat the FY25 operating-income line as noisy and anchor on EBITDA and net income.

TTM profitability — and this is where MPLX separates itself:

MPLX earns roughly seven times ET's net margin on roughly one seventh the revenue. That is not a fairer business so much as a differently shaped one: ET's revenue includes enormous low-margin commodity marketing and refined-product pass-through, MPLX's does not. But the return metrics — 33.3% ROE, 13.2% ROIC — are genuinely superior on any basis, and they are the analytical justification for the valuation premium discussed in §4.

Cash flow — the number that governs a partnership:

Two things to take from this table.

First, the capex discipline is remarkable and is the core of the quality case. MPLX spends $1.808B of capex against $5.909B of operating cash flow — a FY25 capex-to-operating-cash-flow ratio of 30.6%, versus ET's 62.1% on the same FY25 basis. (On the TTM key-metrics measure the gap is wider still: MPLX 17.0% against ET 65.7%.) That is what allows a 7.11% distribution to be covered at all.

Second, growth is being bought, not built. FY25 saw $4.031B of acquisitions — more than four times FY24's $940M and sixteen times FY23's — funded almost exactly by $4.077B of net new debt. Net debt rose from $19.917B to $24.025B in a single year, a 20.6% increase. Goodwill rose from $7.645B to $8.755B. Coverage fell from 1.37x to 1.02x in the same year.

How to read that fairly. Buying $4B of assets with $4B of debt at 3.4x leverage is a defensible use of a strong balance sheet, and MPLX still covers its distribution — which its more levered peer does not. But it means the FY26 earnings picture is front-loaded with interest and back-loaded with contribution, which is precisely why consensus models FY26 EPS at $4.288 against FY25's $4.82. The acquisitions have to earn their cost of capital, and this dataset does not tell us what was bought.

Balance sheet (FY25): cash $2.137B, total current assets $3.994B, net PP&E $22.213B, goodwill $8.755B, total assets $43.005B; total current liabilities $3.249B, short-term debt $1.557B, long-term debt $24.151B, total debt $26.162B, net debt $24.025B, total liabilities $28.477B; total equity $14.528B including $227M of minority interest. Net-debt/EBITDA 3.41x, debt-to-equity 1.86x, debt-to-assets 60.9%, current ratio 1.10x, working capital +$330M (positive, unlike ET's -$19.858B).

Insider signal: none — routine awards only. Every one of the eight most recent filings is a director A-Award dated 2026-04-30 or 2026-05-15, in fractional-unit amounts (78.436 to 1,656.317 units) at a reported price of $0. These are compensation grants, not conviction. There are no open-market purchases and no open-market sales in this dataset, and no insider signal is claimed.

4. Valuation — priced in or room?

At $58.91 (market cap $59.78B, EV $84.41B, ~1.017B diluted units, net debt $24.03B):

Three observations that drive the verdict.

First, the FY26 down-year. Consensus EPS of $4.28792 for FY26E is 11.1% below FY25's actual $4.82. This is the only name in this batch where consensus models earnings going backwards in the current year. The forward P/E therefore rises from 12.7x TTM to 13.7x FY26E before falling again — the opposite of the usual rolldown. A reader looking only at the FY27E and FY28E multiples would miss this.

Second, the growth rate is genuinely better than its peer's. Consensus EBITDA compounds 4.8% a year from FY26E to FY30E ($7.478B to $9.024B) against ET's 1.9%, and consensus EPS compounds 7.7% a year ($4.288 to $5.76). MPLX is the better business, and the street models it as such.

Third, and decisively: the premium is large. On identical methodology across this batch:

MPLX trades at a 65% EV/EBITDA premium to Energy Transfer. The justification is real — lower leverage, covered distribution, far higher margins and returns, a positive working-capital position, and a beta of 0.46 against ET's 0.562. The question this dive has to answer is whether 65% is the right number, and the answer is: it is roughly right, which is exactly why MPLX is a Hold and ET is a Stage-In. You are not being offered a mispricing here. You are being offered a fair price for a good asset.

Synthos fair values — stated arithmetic on EV/EBITDA, no DCF theater. Method: target multiple × FY27E consensus EBITDA of $7.981B = enterprise value, minus net debt of $24.025B = equity value, divided by 1.017B diluted units.

• Bear ~$51 = 9.5x FY27E EBITDA → EV $75.82B → equity $51.79B → $50.92/unit. The de-rating case: coverage slips below 1.0x, the acquisition contribution disappoints, and the premium to ET compresses toward 50%. Lands above the 52-week low of $47.80. -13.6% from spot.
• Base ~$62.50 = 11.0x FY27E EBITDA → EV $87.79B → equity $63.77B → $62.70/unit. An 11.0x multiple is set as base because it is roughly half a turn above where the units trade today, consistent with a business compounding EBITDA ~4.8% a year at 33% ROE with covered distributions — and because it sits just above the street's $60.25 consensus and $63 high. +6.1% from spot, plus the 7.11% distribution.
• Bull ~$70.50 = 12.0x FY27E EBITDA → EV $95.77B → equity $71.75B → $70.55/unit. Requires the sector to re-rate as infrastructure: gas volumes inflect on the LNG and AI-power wedge, coverage returns to 1.3x+, and the market pays a genuine utility multiple. +19.8% from spot — and above every published street target.

Cross-check on earnings: the base $62.50 is 13.0x FY27E consensus EPS of $4.82311, against 12.2x today. For a partnership yielding 7.11% with 28 analysts and zero sells, 13x forward earnings is neither cheap nor demanding. The two methods agree.

6a. What today's price assumes (the inversion)

At $58.91 (10.58x FY27E consensus EV/EBITDA, 12.2x FY27E consensus EPS), today's price embeds roughly the following. Every figure is consensus-derived arithmetic, labelled as such.

• EPS falls 11% in FY26 and then recovers to a new high by FY27. Consensus: $4.82 (FY25 actual) → $4.28792 (FY26E) → $4.82311 (FY27E) → $5.16803 (FY28E). Today's price assumes the FY25 acquisition debt is a one-year drag that fully reverses. Falsified by: FY26 EPS landing below ~$4.10, or FY27 consensus drifting below FY25's actual $4.82 — either of which would mean the $4.031B of acquisitions did not earn their cost of capital. Near-term check: H1'26 EPS of $1.96 requires $2.33 in H2 to reach the $4.288 consensus.
• EBITDA compounds 4.8% a year through FY30. Consensus: $7.478B → $9.024B across FY26E-FY30E. That is roughly 2.5x the rate the same analysts model for Energy Transfer, and it is why the premium exists. Falsified by: any downward revision to the FY28-30 EBITDA line, which would collapse the growth justification for a 10.6x multiple.
• The distribution holds at or above $4.186 per unit — a 7.11% yield — on coverage of just 1.02x. There is essentially no cushion. Falsified by: coverage falling below 1.0x for a full year, or any capex step-up from the $1.808B FY25 level that eats the margin.
• The market keeps paying a ~65% EV/EBITDA premium to Energy Transfer. This is the most fragile assumption in the price. MPLX's advantages are real but they are level advantages, not widening ones: ET's leverage could fall, ET's coverage could repair, and the moment the market decides the safety premium is excessive, MPLX de-rates toward the $51 bear anchor without anything going wrong at MPLX. A partnership priced for perfection in a sector priced for scepticism is asymmetric in the wrong direction.

6b. The return bridge (why the multiple moves)

> expected return ≈ EBITDA growth + multiple drift + distribution yield

• EBITDA growth: +6.7% (consensus FY26E $7.478B → FY27E $7.981B). Consensus EPS growth over the same window is +12.5% ($4.28792 → $4.82311) — but that is a recovery from a depressed FY26 base back to roughly the FY25 actual, not genuine new earnings power. The honest four-year rate is the 4.8% EBITDA CAGR, not the 12.5% one-year bounce.
• Multiple drift: +0.4 turns, or roughly +4% on enterprise value — from the 10.58x the market pays on FY27E today to the 11.0x base. This dive assumes only marginal multiple expansion, and deliberately so. There is no risk-normalisation story left to tell: leverage is already low, coverage is already positive, and the market already pays a 65% premium to the peer. The multiple has largely done its work. Because net debt is only ~28% of enterprise value, the leverage amplification that helps ET so much is muted here — a 4% EV move produces roughly a 6% equity move, not a 17% one.
• Distribution yield: +7.11%. Plus buybacks: FY25 saw $400M of unit repurchases against a $59.78B cap, adding roughly 0.7% of shareholder yield.

Say it plainly: the base case is 7.8% of shareholder yield plus about 6% of price, and almost none of it depends on a re-rating. That is a perfectly respectable ~14% total return for a 0.46-beta asset — and it is also exactly why this is a Hold rather than a Buy. There is no leverage to a change in perception, because perception has already changed. Compare the ET bridge, where a single turn of multiple moves the units 17%. MPLX offers the better business; ET offers the better bridge.

6c. Variant perception (where we differ, what would surprise)

• We have essentially NO variant perception on valuation, and that is itself the verdict. Consensus is $60.25 with a median of $60 and a band of $58 to $63 — the tightest in this batch. This dive's base of $62.50 sits inside that band. Twenty-eight analysts, zero sells, zero strong buys: the street sees a good asset fairly priced, and so do we. Per house rules, a no-edge name defaults toward Hold, and that is where this lands. Stating this plainly is more useful than manufacturing a differentiated take.
• Where we do differ — on the relative call within the sleeve. The interesting question is not "is MPLX worth $60?" but "is a 65% EV/EBITDA premium to Energy Transfer the right price for 1.66 turns less leverage and 0.21x more distribution coverage?" Our answer: it is roughly right today, which makes ET the better risk-adjusted purchase and MPLX the better thing to already own. That relative framing is the actual edge in this pair, and it is why both dives were written on the same date.
• Where we differ from the bull case: the FY26 down-year is being under-discussed. Consensus EPS of $4.288 is 11% below FY25's $4.82, H1'26 has delivered $1.96 of it, and Q1 was an outright miss ($0.90 against $1.05). A partnership trading 0.49% from a 52-week high with RSI at 65.3 while consensus models an earnings decline is a setup that punishes complacency.
• Positive surprise that would force a reprice up: distribution coverage returning above 1.3x (from 1.02x) as the FY25 acquisitions contribute, combined with an upward revision to FY28E consensus EBITDA above ~$9B (from $8.384B). That would validate 12x and put the $70.50 bull anchor in play. Watchable numbers: post-capex free cash flow against the $4.018B distribution, and net debt against the $24.025B FY25 level.
• Negative surprise that would force a reprice down: H2'26 EPS failing to deliver the $2.33 needed to hit consensus, or a coverage print below 1.0x. In that case the safety premium is revealed as unearned and 9.5x with the $51 bear anchor becomes the operative level. Watchable number: quarterly EPS against the $1.06-1.20 run-rate required.

5. The distribution — thin cover, real cover

The distribution: $4.186 per unit trailing-twelve-month, a 7.11% yield at $58.91 — the highest yield in this batch. The GAAP dividend payout ratio is 108.9% and the dividendPaidAndCapexCoverageRatio is 0.98x.

The coverage history on post-capex free cash flow: 1.44x (FY22), 1.40x (FY23), 1.37x (FY24), 1.02x (FY25).

Why FY25 compressed: capex rose 71% from $1.056B to $1.808B while distributions rose 12.9% to $4.018B and operating cash flow was flat at $5.909B. The compression is capex, not deterioration in the underlying business.

The honest read. MPLX's distribution is covered, but only just — and 1.02x is not a comfortable number. What makes it defensible rather than alarming is the shape of the spending: FY25 capex was only 30.6% of operating cash flow (17.0% on the TTM measure) against ET's 62.1%, so there is a large discretionary buffer if the partnership chooses to slow down. What makes it worth watching is that MPLX simultaneously spent $4.031B on acquisitions and repurchased $400M of units in the same year it took coverage to 1.02x. That is three competing claims on the same cash, and the distribution is the one with no flexibility.

Compare to the batch. MPLX 1.02x versus ET 0.81x. MPLX covers, ET does not. That single fact is worth a meaningful part of the 65% valuation premium — and it is why, if you hold only one midstream partnership for income, it should probably be this one.

Partnership mechanics you must know before buying: MPLX issues a Schedule K-1, not a 1099, arriving later than standard tax forms and complicating filing. Distributions are largely return of capital, deferring tax until sale and reducing cost basis. Holding units inside an IRA or other tax-advantaged account can generate unrelated business taxable income (UBTI). The TTM effective tax rate of 0.17% is the structural flip side — the partnership pays essentially no entity-level tax, so the holder does. Diluted unit count has been stable to slightly declining (1,051M in FY20, 1,017M in FY25, 1,017M in Q1'26), helped by the buyback programme — a genuine positive against ET's 11.8% unit growth over a similar span.

6. The knowledge base — one passing mention, and seven false positives worth naming

Direct coverage on MPLX: 1 claim, and it is a list entry.

• odd_lots, 2023-10-03 (bullish, conviction 80), entities ['Energy Transfer','ET','Magellan','MMP','MPLX','NuStar','NS']: "~40% of portfolio in oil/gas pipeline MLPs like Energy Transfer yielding ~9%, tax-deferred until sale — favorite risk-adjusted, tax-advantaged holding." MPLX appears in the entity list but is not the subject of the thesis, and the claim is nearly three years old, describing a ~9% yield environment against today's 7.11%. This supports the MLP asset class, not this ticker.

No tracked thinker has written a thesis on MPLX. kb_claim_count is 1, breadth 1, and the verdict here is fundamentals-and-quant-driven — the STM and NVTS precedent applies.

The false positives, named explicitly because they would mislead a reader who ran the same search. A naive symbol sweep for "MPLX" or "MPC" returns eight matches. Seven are unrelated to this company and are excluded:

This matters beyond bookkeeping. A three-letter symbol that collides with two common finance acronyms and one cryptography term is a live misattribution hazard, and the house rule is to name it rather than quietly filter it. Anyone reproducing this search should expect the same eight hits and discard the same seven.

The adjacent gas lane — the real thematic support, cited verbatim:

• invest_like_the_best, 2026-07-21 (bullish, conviction 82), entity ['natural gas']: "After 18 months of well-level study, US faces historic structural gas deficit from LNG exports (15→35 BCF/d) plus AI compute; upside price risk is convex and unbounded from ~$3.50 today." The highest-conviction gas claim in the corpus, and MPLX's Gathering and Processing segment sits directly upstream of it.
• doomberg, 2026-05-27 (bullish, conviction 60): "Prefer midstreamers, service providers and royalty owners that profit as energy consumption rises over capital-intensive E&P price-takers, which get low multiples and make money only in bursts." A direct endorsement of the business model.
• money_of_mine, 2026-06-05 (bullish, conviction 68), entity ['gas']: "Gas is his core energy pick — a far longer runway than oil or coal, faster demand growth, and multiple ways to invest around it."
• odd_lots, 2026-04-22 (bullish, conviction 62), entities ['natural gas','LNG']: "Natural gas is coming back as more significant for electric generation to power the US AI/data-center boom; US LNG plays a bigger economic role than recognized."

The tracked caution, which applies here as it does to ET:

• jordi_visser, 2026-07-25 (neutral, conviction 55, skill 2.0), entities ['oil','natural gas']: "Oil stocks rarely get multiple expansion because revenue is tied to the underlying commodity — if oil stays $60 or nat gas $3 forever the equity can't re-rate." This dive's base case assumes only 0.4 turns of expansion, so Visser's constraint is largely already respected in the arithmetic.

Read: Low conviction, breadth 1. The panel underwrites the commodity and the business model at high conviction and has nothing to say about this ticker. That caps position size and reinforces the Hold.

7. Technicals — the most relentless uptrend in the batch, which is the problem

• Price $58.91 on 2026-08-04, +0.79% on the day, on 1.39M units against a 1.74M average — quiet volume.
• 0.49% below the 52-week high of $59.20 (quote-based high: $60.00). +23.2% above the 52-week low of $47.80.
• Maximum drawdown from peak: -0.49%. MPLX has essentially not had a pullback. That is the single most striking technical fact on this page and it is not a compliment — it means there has been no entry discount at any point, and there is none now.
• Above the 50-DMA ($56.73) by 3.8% and above the 200-DMA ($55.43) by 6.3%. Both rising.
• RSI 65.3 — the highest in this batch and approaching the overbought threshold. MACD +0.52 — positive.
• Relative performance: 12-month +12.2% against SPY +19.9% and QQQ +23.9%; 6-month +5.4% against SPY +9.5%; 3-month +4.1% against SPY +5.1%. MPLX has ground steadily higher and steadily trailed the index — a textbook low-beta income profile.

The tactical read. There is no technical case for buying today. A 0.46-beta income asset at RSI 65.3, half a percent from a 52-week high, with a maximum drawdown of half a percent and a street target band of $58-63, is as fully discovered as a security gets. The 50-DMA at $56.73 (-3.7%) and the 200-DMA at $55.43 (-5.9%) are the levels that would offer even a modest discount, and neither has been touched recently. That structure, more than anything in the financials, is why this is a Hold.

8. Moat and competitive position

The moat is physical, regulated and genuinely wide — and it is the same moat every large US midstream operator enjoys. Gathering systems, processing plants, fractionators, pipelines and storage caverns cannot be replicated at scale in the current permitting environment. MPLX's $22.213B of net property, plant and equipment would cost substantially more to rebuild than it is carried for, and the rights-of-way underneath it are functionally irreplaceable.

Two things sharpen MPLX's version of that moat and one thing weakens it.

Sharpening it: the inland marine franchise. MPLX owns and charters boats and barges on the Ohio River and Gulf Coast and runs its own marine repair facility. River logistics is a niche with high barriers, few operators, and demand that does not disappear. It is a genuinely differentiated asset within midstream, and it barely gets discussed.

Sharpening it: the sponsor relationship. Marathon Petroleum as general partner gives MPLX a stable anchor customer, a captive volume base tied to refinery throughput, and a natural acquisition pipeline (dropdowns).

Weakening it: that same sponsor relationship is a concentration. MPC is simultaneously MPLX's controlling general partner and a major customer. This dataset does not disclose what share of MPLX's revenue comes from MPC, so the exposure cannot be sized here — and the inability to size it is itself the risk. A partnership whose largest customer also controls its board has structurally weaker negotiating leverage on tariffs than an independent operator, and the governance protection an outside unitholder has is limited. This is the single most under-quantified item in the dive and is named as the biggest risk accordingly.

Peer set (from this dataset): EPD (Enterprise Products, $81.9B cap), ET ($69.8B — also in this batch), KMI (Kinder Morgan, $69.9B), OKE (ONEOK, $55.6B), TRP (TC Energy, $68.6B), LNG (Cheniere, $54.1B — also in this batch), plus integrateds E, SU, IMO and services name SLB. This dataset does not carry forward estimates for those peers, so no external peer-multiple table is offered — the only clean comparison available is the within-batch one in §4, which shows MPLX at a 65% EV/EBITDA premium to ET and a 38% premium to Cheniere.

9. Verdict, kill-criteria and flip conditions

Hold. MPLX is the best-financed midstream partnership in this batch and one of the better ones anywhere: 3.41x net-debt/EBITDA, 38.0% TTM net margin, 33.3% return on equity, FY25 capex at just 30.6% of operating cash flow against ET's 62.1%, a 7.11% distribution covered 1.02 times, a stable-to-declining unit count, 28 analysts with zero sells, and consensus EBITDA compounding 4.8% a year — roughly 2.5 times the rate modelled for its larger peer. This morning it beat on revenue and hit EPS exactly.

And it is priced for all of it. At $58.91 the units sit 0.49% from a 52-week high, at 10.58x FY27E consensus EV/EBITDA — a 65% premium to Energy Transfer on identical methodology — with RSI at 65.3, a maximum drawdown from peak of 0.49%, and a street target band spanning $58 to $63. Base fair value of $62.50 (11.0x FY27E EBITDA) is +6.1%. And consensus models FY26 EPS falling 11% to $4.288 from FY25's $4.82, with H1'26 having delivered only $1.96 of it.

A Hold here is not indecision. It is the correct answer to a specific question: you are being offered a fine business at a fair price with no margin of safety, no conviction lane, no technical discount, and a consensus down-year in progress. The distribution does the work; the price does not.

How to act on a Hold:

• Already own it: hold, and collect the 7.11%. Nothing in this analysis argues for selling a covered 7.11% distribution on a 0.46-beta asset with 3.41x leverage. This is the income anchor of an energy-infrastructure sleeve and it is doing its job.
• New money: do not initiate at $58.91. Wait for the 50-DMA (~$56.73, a 7.4% yield) or the 200-DMA (~$55.43, a 7.6% yield), or for evidence that H2'26 is tracking above the $2.33 needed to hit consensus. Or — the more interesting option within this batch — deploy that capital into ET at 6.41x instead, accepting the leverage in exchange for a 17% gap to fair value rather than a 6% one.
• Sizing discipline: 2-3% as the income anchor, built on weakness. The low conviction count caps it below what the fundamentals alone would justify.

Pre-registered KILL / avoid-adding criteria:

• Distribution coverage falls below 1.0x on post-capex free cash flow — the thing that distinguishes MPLX from ET disappearing is a thesis break, not a data point.
• H2'26 EPS tracks below ~$2.10 (against the ~$2.33 needed to reach the $4.28792 FY26 consensus), which would confirm the FY25 acquisitions are not earning their cost of capital.
• Net debt rises above ~$27B from $24.025B, or net-debt/EBITDA passes ~3.8x — the balance-sheet advantage over ET narrowing removes the entire justification for the premium.
• Another acquisition of FY25's scale ($4B+) funded entirely with debt before the last one has demonstrably contributed.
• The 200-DMA (~$55.43) is decisively lost, which for a name that has never drawn down more than 0.49% from peak would be a genuine change in character.

Pre-registered FLIP TO BUY:

• The units trade at or below ~$55.50 (the 200-DMA, a 7.6% yield, roughly 10.0x FY27E EBITDA) with FY27E consensus EPS still at or above $4.82311 — that is a 13% gap to base and a genuine margin of safety.
• Or distribution coverage returns above 1.3x while FY28E consensus EBITDA is revised above $9B — which would justify 12x and make the $70.50 bull anchor the operative target rather than the ceiling.
• Or a tracked thinker initiates genuine coverage of this ticker, which would relax the sizing constraint that currently caps the position.

Where MPLX fits in the Synthos Framework Portfolio. The energy-infrastructure / income sleeve, as the defensive income ANCHOR — 2-3%, built on weakness rather than initiated here. On overlap within this batch: MPLX and ET are complementary rather than redundant, and the pairing is the point. MPLX is the balance sheet (3.41x leverage, 1.02x coverage, 7.11% yield, 38% net margin) — you buy it to hold through a cycle. ET is the valuation (5.07x leverage, 0.81x coverage, 6.58% yield, 6.41x FY27E EV/EBITDA) — you buy it for the gap to fair value. Cheniere (LNG) is a different animal entirely: a corporate taxpayer with a 0.84% yield and a buyback-driven return, not an income holding. A sensible sleeve construction is MPLX as the anchor, ET as the torque, sized so the blended leverage sits nearer MPLX's than ET's. Logged as a tracked Synthos call (Hold) as of 2026-08-04 at $58.91.

Single biggest risk: sponsor concentration that this dataset cannot quantify. Marathon Petroleum is both MPLX's controlling general partner and a major customer. That relationship supplies the stability, the dropdown pipeline and the anchor volumes that justify the premium multiple — and it means a single counterparty's refining economics, capital allocation and governance decisions sit underneath a 7.11% distribution. The revenue share attributable to MPC is not disclosed in this dataset, so the exposure cannot be sized, and an unsizeable concentration is a worse risk than a large measured one.

Single most fragile assumption in the price: that the market keeps paying a ~65% EV/EBITDA premium to Energy Transfer. MPLX's advantages are genuine but they are static: lower leverage, covered distributions, better margins. They do not widen. If ET deleverages, or if the sector re-rates and investors reach for the cheaper asset, MPLX can de-rate toward the $51 bear anchor while every operating metric on this page stays exactly as good as it is today.

Provenance and disclosures

• Traceability: kb_claim_count 1 — a single passing mention (odd_lots, 2023-10-03, conviction 80, MPLX listed among pipeline MLPs; nearly three years stale). Seven further symbol matches are false positives and are named and excluded in §6 (multi-party computation: real_vision 2026-02-10, ml_street_talk 2025-10-04, a16z 2026-04-02, empire 2026-03-30, raoul_pal 2024-12-05; marginal propensity to consume: mike_green 2026-02-11; Bank of England Monetary Policy Committee: michael_howell 2021-07-15). Adjacent thematic support cited verbatim from invest_like_the_best (2026-07-21, conviction 82), doomberg (2026-05-27, conviction 60), money_of_mine (2026-06-05, conviction 68) and odd_lots (2026-04-22, conviction 62); tracked caution from jordi_visser (2026-07-25, neutral, conviction 55, skill 2.0). Breadth 1; net conviction minimal-positive. The verdict is fundamentals-and-quant-driven — the STM and NVTS precedent.
• Data as-of: fundamentals 2026-06-30 (Q2'26; annual figures 2025-12-31) · estimates 2026-08-04 · prices 2026-08-04 (quote $58.91, 50-DMA $56.73, 200-DMA $55.43, 52-wk range $47.80-$60.00 per quote / $47.80-$59.20 per technicals) · KB claims swept 2026-08-04. All market data from the single pre-pulled FMP dataset for this ticker. Nothing was re-fetched, recalled from memory, or estimated.
• Data gaps and inconsistencies, stated explicitly: (1) The Q2'26 (2026-06-30) income-statement row is partly CORRUPT — grossProfit (-$1.174B), interestExpense (-$310M), bottomLineNetIncome (-$911M) and eps (-$0.90) are all exact sign-flipped duplicates of the Q1'26 values, and weightedAverageShsOutDil is 0. Those six fields are excluded; revenue ($3.487B), operating income ($1.568B) and net income ($1.097B) are used, with EPS taken from the earnings calendar ($1.06) and the unit count from Q1'26 (1,017M). This is the single largest data-quality issue in the batch and is flagged rather than smoothed. (2) FY25 named revenue lines total $6.684B against $11.817B of income-statement revenue — a $5.133B unexplained gap; the split is used as a mix signal only. (3) Geographic segmentation is EMPTY — MPLX provides no geographic segment reporting in this dataset, and none is analysed. (4) The segment reporting format changed between FY17 and FY18 (Gathering-and-Processing / Logistics-and-Storage → Product / Service / Other), so no multi-year segment trend is drawn. (5) FY25 operating income fell 9.9% while EBITDA and net income rose, driven by a $1.054B swing in the reported operating-expense line (-$471M to +$583M) that the dataset does not explain; the FY25 operating-income figure is therefore not used for trend. (6) The Q1'26 earnings-calendar revenue ($2.806B) differs from the income-statement figure ($2.863B); both are shown where relevant. (7) Insider filings are routine director awards at $0 only — no open-market transactions — so no insider signal is claimed. (8) No Q3'26 earnings date is listed; the early-November inference is stated as inference.
• Partnership caveat: MPLX is a master limited partnership. Holders own units, not shares, and receive a Schedule K-1 rather than a 1099. Distributions are largely return of capital, deferring tax and reducing cost basis until sale. Holding units in a tax-advantaged account can generate UBTI. The TTM effective tax rate of 0.17% reflects the pass-through structure. Diluted unit count has been stable to slightly declining (1,051M in FY20 to 1,017M in FY25), assisted by $400M-$600M of annual repurchases.
• Fair-value caveat: the $51 / $62.50 / $70.50 anchors are scenario multiples on FY27E consensus EBITDA of $7.981B (9.5x / 11.0x / 12.0x), converted to equity by subtracting FY25 net debt of $24.025B and dividing by 1.017B diluted units, then cross-checked against FY27E consensus EPS of $4.82311. They are scenario arithmetic, not a discounted cash flow. Because net debt is only ~28% of enterprise value, these anchors are less multiple-sensitive than ET's — one turn of EV/EBITDA moves the units approximately $7.85.
• Timing caveat: this dive is written on the day the Q2'26 result was released (2026-08-04). The market's full reaction may not be reflected in the $58.91 close, and the units sit within half a percent of a 52-week high — which is the central reason for the Hold.
• Not investment advice. Independent research, educational and informational only, never personalized.
• Version: 2026-08-04-full.