Enterprise Products Partners L.P. EPD

Reference table

1. What they actually sell — four segments, and a disclosure quirk you must understand

Enterprise Products operates through four reportable segments: NGL Pipelines and Services (19 natural gas processing facilities across Colorado, Louisiana, Mississippi, New Mexico, Texas and Wyoming, plus NGL pipelines, fractionation plants, storage and NGL marine export/import terminals); Crude Oil Pipelines and Services (pipelines, storage, marine terminals and a fleet of 255 tractor-trailer tank trucks); Natural Gas Pipelines and Services (gathering, treating and transmission, including leased underground salt-dome storage); and Petrochemical and Refined Products Services. CEO A. James Teague. 7,300 full-time employees. Listed 1998-07-28.

The segment disclosure carries an unusual quirk that must be explained before the numbers are used. Segment revenues are reported gross of intersegment activity, and the eliminations are enormous — -$131.540B in FY25, roughly 2.5 times the $52.596B of consolidated revenue. This means segment revenue percentages of consolidated revenue are meaningless and are not calculated here. What the segment data is good for is showing direction and relative change, and on that basis it is genuinely informative:

This is the most important qualitative table in the dive. The mix is shifting decisively toward NGL (+22.1%) and natural gas (+39.7%) and away from crude (-16.9%) — which is precisely the shift the panel describes. Doomberg (2026-07-16, conviction 80): "China made the hydrocarbon suite fungible; exploding US NGL volumes ('peak not oil') plugged China's hole, replacing ~4M bbl/day of crude demand and keeping crude capped." EPD is the largest NGL midstream operator in the country and its own segment data corroborates that claim from the inside.

Geographic segments: the seg_geo block is EMPTY. Enterprise Products does not report geographic revenue segments in this dataset. For a business with NGL marine export terminals whose economics are increasingly tied to international demand, that is a real disclosure gap, and no geographic split is asserted anywhere in this dive.

2. Read EBITDA, not revenue — the single most important framing in this name

Look at what happens when you read the revenue line and the EBITDA line side by side:

Revenue has fallen in two of the last three years. EBITDA has risen in every single year for six years. That is the whole argument for calling this a toll road rather than a commodity bet: the top line is a pass-through of the value of molecules being handled, and the earnings line is a function of the volume handled and the fee earned. Any analysis of EPD that leads with revenue growth is measuring the wrong thing.

The quarterly detail makes the point even more sharply:

In Q1'26 revenue fell 6.7% and EBITDA rose 11.6%. In Q2'26 revenue rose 60.8% and EBITDA rose 17.1%. The revenue line moved 67 points between those two quarters; the EBITDA line moved 6. That is what a fee-based business looks like.

Trailing-twelve-month, computed by hand from the quarterly statements: revenue $58.471B, EBITDA $10.611B (+9.7% on the prior TTM period's $9.677B), EPS $2.88 (+7.5% on $2.68). And H1'26 EPS of $1.52 is up 16.9% on H1'25's $1.30, annualising at $3.04 — roughly 4% ahead of the FY26 consensus of $2.914.

3. The distribution, the capex cycle, and the coverage problem

This is the section that decides whether EPD is a Buy or a Hold, so it gets the most detail.

Read this honestly, because it can be read two ways and both readings are legitimate.

The bear reading: free cash flow has covered the distribution less well every single year for four years, and has been below 1.0x for two consecutive years. The FY25 shortfall of $1.713B was funded by $2.510B of net new debt. Net debt has risen from $28.891B (FY23) to $31.679B (FY24) to $33.681B (FY25). A partnership that borrows to pay its distribution is on a path that ends badly, and 0.63x is a low number.

The bull reading, and it is the one this dive adopts: the entire deterioration is explained by growth capex, which has nearly tripled from $1.964B to $5.620B. Maintenance capex and growth capex are not separately disclosed in this dataset — a genuine gap — but the direction is unambiguous: this is a company in an investment cycle. Operating cash flow still covers the distribution 1.84x and that ratio has been stable or rising for three years. Net debt at 3.17x TTM EBITDA is within the normal investment-grade band for large midstream, interest coverage is 5.05x, and the payout ratio on earnings is a comfortable 56.4%. Critically, the capex is converting: TTM EBITDA is up 9.7% and Q2'26 EBITDA up 17.1%. When a capex cycle turns down while EBITDA keeps rising, free-cash-flow cover repairs mechanically.

The test is simple and quarterly: if capex moderates from the FY25 run-rate while EBITDA keeps compounding at 9-10%, cover returns above 1.0x within roughly two years and this is a straightforward Buy. If capex keeps rising and EBITDA growth stalls, the distribution stops growing and the units de-rate. That is the single most fragile assumption in the price and it is named as such in section 6a.

Balance sheet (2025-12-31): total debt $34.926B, cash and short-term investments $1.679B, net debt $33.681B. Total equity $29.739B plus $831M of minority interest. Property, plant and equipment net of $51.796B on total assets of $77.902B — this is a genuinely asset-heavy business, which is the source of both its moat and its capital intensity. Book value per unit $13.91, so the units trade at 2.72x book.

4. Technicals — the calmest chart in the batch

• Price $37.87, above both moving averages: 50-DMA $37.62 (+0.7%) and 200-DMA $35.41 (+7.0%), with the 50 above the 200. A quiet, intact uptrend.
• RSI(14) 51.7 — perfectly neutral. MACD +0.227 — marginally positive. Nothing stretched in either direction.
• -4.85% from the $39.80 52-week high; +25.4% above the $30.19 low. The maximum drawdown from peak across the entire twelve-month window is 4.85% — the same number, meaning the units have essentially never had a meaningful decline in a year.
• Beta 0.469 — roughly a third of the market's volatility, and the lowest in this batch by a wide margin (SpaceX has no meaningful beta, Cloudflare 1.66, Snowflake 1.353, Southern Copper 1.124).
• Returns: 3-month -0.4%, 6-month +14.1%, 12-month +22.2%, versus SPY at +5.1% / +9.5% / +19.9%. EPD has beaten the S&P over twelve months with a third of the beta and a sub-5% maximum drawdown, while paying a 5.82% cash yield. That is the entire case for the income sleeve in one sentence.
• The three-month flat patch (-0.4% against SPY +5.1%) is why the entry is available at all.

5. Knowledge base — thin at the name level, coherent on the model

Enterprise Products has 1 claim tagged to the EPD ticker and 1 further claim naming "Enterprise Products Partners" directly — both from Lyn Alden (skill 1.1). That is a thin name-level lane and the front matter reports kb_claim_count: 1 accordingly. What compensates is that the surrounding model lane is unusually coherent and directionally aligned. Quarantined and misattributed entries are excluded; all quotations are verbatim.

Name-level claims:

• Lyn Alden — 2023-03-29, bullish, conviction 72 (the EPD-tagged claim): "Locked in cheap 20yr fixed-rate debt against real assets; 7%+ yield, 25yrs of distribution increases — sits opposite the banks' rate problem." Note the dates: this is a 2023 claim and the yield today is 5.82%, not 7%+ — the units have appreciated. The structural point about long fixed-rate debt against real assets remains the correct frame.
• Lyn Alden — 2025-03-20, bullish, conviction 60: "Cites EPD breaking out above its ~$30 resistance while paying high distributions as an example of how to play energy." The unit price is now $37.87, +26% above the $30 level she flagged, which makes this a claim that has worked.

The supporting model lane — this is where the real conviction sits:

• Doomberg — 2025-12-22, neutral-constructive, conviction 75 (the cleanest statement of the thesis anywhere in the base): "Don't own commodities or price-taking producers; own companies leveraged to volume — service providers, enablers and midstream — since energy demand always grows regardless of price."
• 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."
• Doomberg — 2026-07-16, bearish-on-crude but bullish-on-volume, conviction 80: "China made the hydrocarbon suite fungible; exploding US NGL volumes ('peak not oil') plugged China's hole, replacing ~4M bbl/day of crude demand and keeping crude capped — permanent changes that outlast the war." This is the single most on-thesis claim for EPD specifically, because EPD is the largest NGL midstream operator and its own FY25 segment data shows NGL gross revenue +22.1% against crude -16.9%.
• Lyn Alden — 2025-01-23, bullish, conviction 80: "US natural gas trades at an energy-density discount to oil and will fuel data centers; producers and transporters well-positioned amid LNG-export/permitting boom." Corroborated by EPD's natural gas segment at +39.7%.
• Lyn Alden — 2022-08-08, bullish, conviction 78: "Energy pipelines and oil producers are attractive on a 5-10 year view — it's the average price under the curve, not the near-term spot, that matters."

The counterweight, and it belongs prominently in the record because it comes from the panel's highest-skill voice:

• Jordi Visser — 2026-07-25, neutral, conviction 55, skill 2.0: "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."

Read. This is a Moderate, positive-thin conviction. The name-level coverage is two claims from one voice, both of which have worked. The model — own volume, not price; own the toll road, not the producer — is endorsed at conviction 60-80 across two independent voices, and EPD's own segment disclosure independently corroborates the specific volume claim. The counterweight from Visser is not a bear case on the business; it is a ceiling on the multiple, and it is exactly why the bull case in section 6 is $50 rather than $65.

6. Valuation — priced in or room?

At $37.87 (market cap $81.932B, net debt $33.681B, minority interest $0.831B, EV ~$116.4B hand-computed):

Other current metrics: EV/EBITDA ~11.0x (hand-computed on TTM EBITDA of $10.611B), P/B 2.72x, distribution yield 5.82% at a 56.4% payout, return on equity 21.2%, interest coverage 5.05x, net debt / TTM EBITDA 3.17x, price/operating cash flow 14.6x, price/free cash flow 57.0x (elevated purely by the growth-capex cycle).

Coverage note: FY27 and FY28 carry 10 EPS analysts each — reasonable depth for a name this size. FY26 has 6, FY29-30 only 4. The FY26 revenue estimate rests on just 4 analysts, so revenue estimates should be treated as indicative; EPS coverage is the more reliable series and is what the valuation is built on.

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

At $37.87 — 13.0x FY26E EPS, ~11.0x TTM EV/EBITDA, 2.72x book, a 5.82% distribution — the price makes these falsifiable claims. Sources labeled.

• That the distribution is safe and grows. Arithmetic: the payout ratio on earnings is 56.4% and operating cash flow covers the distribution 1.84x. But free cash flow covered it only 0.63x in FY25 and 0.79x in FY24, with the gap funded by $2.510B and $3.169B of net debt. This is the single most fragile assumption in the price. The market is assuming the capex cycle is temporary. If FY26 capex exceeds FY25's $5.620B while EBITDA growth slows below ~7%, that assumption fails.
• That EBITDA compounds at roughly 6-9%. Consensus: EBITDA of $10.363B (FY26E) → $10.920B (FY27E) → $11.539B (FY28E). Arithmetic: TTM EBITDA is already $10.611B, so the FY26 consensus of $10.363B is BELOW the trailing actual — the sell side is modelling a modest step back. H1'26 EPS annualising at $3.04 versus a $2.914 FY26 consensus says the same thing: consensus looks conservative, which is a favourable configuration.
• That leverage stays near 3.2x. Arithmetic: net debt of $33.681B against TTM EBITDA of $10.611B is 3.17x. The price implies no ratings pressure and no distribution cut. Interest coverage of 5.05x supports that, but the trend in net debt (from $28.891B in FY23) is one direction.
• That the multiple does NOT expand. Arithmetic: the price implies 13.0x FY26E falling to 11.0x FY28E purely on earnings growth — i.e., the market is paying for cash flow, not for re-rating. Panel, labeled as such: Jordi Visser (2026-07-25) explicitly argues commodity-linked equities rarely re-rate. The price agrees with him, which is a large part of why the risk here is low.
• That volumes keep shifting toward NGL and gas. From the company's own segment disclosure: NGL gross revenue +22.1% and natural gas +39.7% in FY25, crude -16.9%. The price assumes this continues; the panel (Doomberg, 2026-07-16, conviction 80) independently argues it will.

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

Expected return ≈ EPS growth + multiple drift + shareholder yield. This is the one name in the batch where all three terms are legible and two of them are unambiguously positive.

• Shareholder yield: +5.82% from the distribution, plus a small unit repurchase ($300M in FY25, roughly 0.4% of market cap). Call it ~6.2% of cash return.
• EPS growth: ~9% a year on consensus ($2.66 FY25 actual → $3.457 FY28E), and running ahead of that in H1'26 (+16.9%).
• Multiple drift: our base case assumes essentially NONE, and this is deliberate. The base fair value of $40 is 12.5x FY27E EPS of $3.193 against a current 13.1x trailing — i.e., we assume the multiple holds roughly flat to very slightly lower. We do not underwrite any re-rating, for the exact reason Visser gives: commodity-linked equities rarely get one. If your thesis requires the multiple to expand toward, say, 15x, that is your fragile leg and you should name it. Ours does not.

So the expected return is ~9% earnings growth plus ~6% cash yield, minus a small multiple drag — call it low-double-digit annualised, which is precisely what a well-run income holding should deliver and is why the position is sized as income rather than growth.

Note what is NOT re-rating here. The P/E falling from 13.0x (FY26E) to 11.0x (FY28E) at a constant price is mechanical rolldown from earnings growth, not the market getting more optimistic. It is stated that way throughout.

Scenario arithmetic — and usefully, two independent methods converge in every case:

• Bear ~$30 (-21%). Volumes stall, the capex cycle overruns, distribution growth stops and the units de-rate toward a distressed midstream yield. Method 1: 10.5x an EPS of $2.85 (roughly FY25's level, i.e. no growth) → $29.93. Method 2: the current $2.205 distribution at a 7.35% demanded yield → $30.00. Cross-check: $30 is 0.6% below the $30.19 52-week low.
• Base ~$40 (+6%). Method 1: 12.5x consensus FY27E EPS of $3.193 → $39.91. Method 2: the current $2.205 distribution at a 5.50% demanded yield → $40.09. Total return with the 5.82% distribution ≈ +11.4%. Cross-check: the street's own consensus target is $40.60 — a third independent method landing in the same place.
• Bull ~$50 (+32%). The NGL and gas volume thesis compounds and the capex cycle converts cleanly. Method 1: 13.0x consensus FY30E EPS of $3.84 → $49.92. Method 2: the distribution grown ~3% a year for four years to $2.48 at a 5.00% demanded yield → $49.63. Cross-check: $50 is above the highest sell-side target ($45), so the bull case requires the street to raise.

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

Where we differ from the market:

• We think consensus is too conservative on the near term, and the data says so. The FY26 EBITDA consensus of $10.363B is below the $10.611B already delivered on a trailing basis, and the FY26 EPS consensus of $2.914 is below the $3.04 that H1'26 annualises to. When trailing actuals exceed forward consensus, the estimate revisions usually go one way. This is our clearest positive divergence.
• We think the market reads the revenue line and gets the wrong answer. FY25 revenue fell 6.4% and EBITDA rose 3.5%; Q1'26 revenue fell 6.7% and EBITDA rose 11.6%. Any screen or headline that ranks EPD on revenue growth is systematically mispricing a fee business. That mispricing is part of why a 21.2%-ROE asset trades at 13.1x earnings.
• We take the free-cash-flow coverage problem MORE seriously than the market does. 34 Buy ratings and a target above spot suggests the street is comfortable. A 0.63x FCF cover ratio, down from 1.48x in four years, deserves more attention than it is getting — and it is why this dive names it as the single biggest risk despite issuing a Buy.
• We put no weight on a re-rating. Much bullish midstream commentary implicitly assumes multiple expansion as the asset class is "rediscovered." We assume none, in line with Visser's 2026-07-25 caution, and the base case still works. That is what makes it a comfortable Buy rather than a stretch.

Positive surprises that would force a repricing upward:

• Free-cash-flow distribution cover returning above 1.0x — the cleanest possible signal that the capex cycle has turned. Watch the FY26 full-year cash-flow statement.
• Capex guided below the FY25 run-rate of $5.620B while EBITDA growth holds near 9-10%.
• A Q3'26 beat on 2026-10-29 against the sequentially lower $0.72 consensus, extending the H1 momentum.
• NGL and natural gas segment growth continuing above 20%, validating the Doomberg volume thesis with the company's own numbers.
• Datacentre-driven natural gas transport volumes becoming visible in disclosure — the Lyn Alden 2025-01-23 claim converting from theme to line item.

Negative surprises that would confirm the bear case:

• FY26 capex exceeding $6B with no corresponding EBITDA acceleration — the capex cycle becoming structural rather than temporary.
• Distribution growth stopping. Cash distributions have risen every year ($4.095B → $4.301B → $4.512B → $4.678B). A flat year would be the first real crack.
• Net debt / EBITDA rising above ~3.5x, or interest coverage falling below ~4.5x from the current 5.05x.
• A quarter where EBITDA declines year-on-year. It has not happened in six years of annual data; Q3'25's +0.4% was the closest call.
• NGL segment growth decelerating below 10%, which would undercut the mix-shift argument that is the core of the volume thesis.

7. Moat and competitive position

A genuine, physical moat — arguably the most durable in this batch. Enterprise Products owns $51.796B of net property, plant and equipment: pipelines, 19 gas processing facilities, fractionators, salt-dome storage, and marine export terminals. You cannot compete with an existing pipeline by building a parallel one; the permitting, right-of-way and capital requirements make duplication uneconomic, and the incumbent can always price below a new entrant's cost of capital. This is the classic regulated-adjacent infrastructure moat, and it shows up as 21.2% return on equity and 5.05x interest coverage in a business that takes minimal commodity price risk.

The competitive set (from this data file's peer list, prices as supplied): Energy Transfer $20.28 / $69.8B, MPLX $58.91 / $59.8B, Kinder Morgan $31.40 / $69.9B, ONEOK $88.24 / $55.6B, Williams $70.43 / $86.1B, TC Energy $65.86 / $68.6B. EPD's $81.9B market cap makes it the largest pure US midstream partnership in the set. Note that the peer list also contains producers (EOG, Canadian Natural, Equinor, Petrobras) which are a different business model entirely — the price-taking producers Doomberg explicitly advises against owning relative to midstream.

Where the moat is weaker:

• Volume, not price, is the exposure — but volume is still an exposure. If US hydrocarbon production genuinely rolls over, tolls fall regardless of how good the pipeline is.
• Permitting cuts both ways. It protects incumbents from new competition and constrains EPD's own growth projects.
• Contract renewal risk. Long-term fee contracts eventually reprice, and this dataset does not disclose the contract duration profile — a genuine gap.
• The multiple is structurally capped. Visser's 2026-07-25 point stands: commodity-linked revenue rarely earns multiple expansion, however good the asset.

8. Management, capital allocation and insider activity

Capital allocation is textbook midstream, with one caveat. FY25: $8.585B of operating cash flow deployed into $5.620B of growth capex, $4.678B of distributions and $300M of unit repurchases, funded with $2.510B of net new debt. No stock-based compensation appears in the FY23-25 cash-flow data (and the TTM ratio is marginally negative), which is unusual and shareholder-friendly relative to the software names in this batch, where it runs 20-34% of revenue. The caveat is the debt-funded distribution gap set out in section 3 — the one place where the capital allocation is genuinely doing something that cannot continue indefinitely.

CEO A. James Teague (co-CEO, and also a director).

Insider activity contains the single clearest positive signal in this batch. Of the eight filings in the data file:

The CEO made an open-market PURCHASE of 2,665 units at $37.55 on 2026-03-20 — roughly $100,000, small in dollar terms but categorically different from an award or a vesting. It is the only open-market insider purchase across all five names in this batch. Every other transaction in the file is a phantom-unit exercise or the associated in-kind tax withholding, which are compensation mechanics and carry no signal. Post-transaction, Teague holds 3,110,280 units (after the February exercise) and Bachmann 2,270,826; Duncan Williams holds 1,417,390.

Read it for what it is: a modest but genuine expression of confidence at $37.55, essentially the same price as today's $37.87. Not a thesis in itself, but the right direction, and a useful contrast with the insider selling at Snowflake and Cloudflare in this same batch.

9. Structural considerations — the partnership form

EPD is a limited partnership, not a corporation, and units are not shares. Three practical consequences that a reader should understand before evaluating the yield:

1. Tax reporting is via Schedule K-1, not a standard dividend form. This is administratively more involved and typically arrives later in the tax season.

2. Distributions are not dividends. They generally include a return-of-capital component that reduces cost basis rather than being fully taxed as income in the year received — which is why holders sometimes describe the yield as tax-deferred until sale.

3. Account type matters. Partnership income held inside tax-advantaged accounts can raise unrelated-business-taxable-income considerations that do not arise with corporate shares.

None of this is a comment on the business quality, and none of it is tax advice — but a 5.82% yield with K-1 reporting is a different instrument from a 5.82% yield on a corporation, and the difference belongs in the analysis rather than in a footnote. The panel notes the same trade-off: Odd Lots (2023-10-03, conviction 80) described a large MLP allocation as "tax-deferred until sale—favorite risk-adjusted, tax-advantaged holding."

10. Verdict, kill-criteria and flip conditions

Buy — income sleeve.

Enterprise Products is the only name in this batch where the arithmetic works in the buyer's favour today. At $37.87 you own a $51.8B net property base earning a 21.2% return on equity, at 13.1x trailing earnings and 2.72x book, paying a 5.82% distribution at a 56.4% payout, with 5.05x interest coverage, a 0.469 beta, and a maximum twelve-month drawdown of 4.85% — while still beating the S&P over that period. Base fair value of $40 plus the distribution is roughly an 11% total return, and it is corroborated three independent ways (12.5x FY27E EPS, a 5.5% demanded yield, and the street's own $40.60 target). Every sell-side target, including the lowest at $38, sits above the current price — the exact opposite configuration to Southern Copper, Snowflake and Cloudflare in this same batch. Trailing EBITDA already exceeds the FY26 consensus, and H1'26 EPS annualises 4% ahead of the full-year estimate. The CEO bought units on the open market in March.

Build it in two tranches:

1. Tranche 1 — now (~two-thirds), at ~$37.87. Above both moving averages, RSI neutral, base fair value above spot, every analyst target above spot, and you begin collecting 5.82% immediately.

2. Tranche 2 — on a pullback to the 200-DMA (~$35.41) or after the 2026-10-29 print confirms the H1 momentum against a sequentially lower $0.72 consensus. At $35.41 the yield is 6.23% and the base case is +13%.

Pre-registered KILL / trim criteria:

• Free-cash-flow distribution cover falling below ~0.55x or failing to improve toward 1.0x by the FY26 full-year cash-flow statement. This is the number that matters most.
• Distribution growth stopping. Cash paid has risen every year for four years; a flat year would be the first genuine crack in the thesis.
• Net debt / EBITDA above ~3.5x or interest coverage below ~4.5x (currently 3.17x and 5.05x).
• A year-on-year EBITDA decline in any quarter — it has not happened in the six years of annual data available and would signal that volumes, not just prices, have turned.
• NGL segment gross revenue growth decelerating below 10%, undercutting the volume mix-shift that is the core of the structural case.

Pre-registered FLIP TO HIGHER CONVICTION (upsize toward the top of the 3-5% range):

• FCF distribution cover back above 1.0x with the distribution still growing — the single cleanest confirmation available.
• Capex guided below $5B for FY26 while EBITDA growth holds near 9-10%.
• Datacentre-driven natural gas volumes appearing explicitly in company disclosure, converting the Lyn Alden thesis into a reported line.

Where EPD fits in the Synthos Framework Portfolio. The income / real-assets sleeve, as the anchor holding at 3-5% core weight. Its role in the book is structural, not tactical: with a beta of 0.469 and a sub-5% maximum drawdown it is the ballast against the high-multiple, high-beta names elsewhere in the portfolio (Cloudflare at 1.66 beta and 44x sales; Snowflake at 1.353 and 21x sales). On overlap: EPD is complementary, not redundant to a copper miner like Southern Copper — SCCO is a price-taker whose earnings are the commodity price, EPD is a volume-taker whose earnings are the fee. Owning both gives real-asset exposure across price and volume, which is precisely the distinction Doomberg draws. Logged as a tracked Synthos call (Buy) as of 2026-08-04 at $37.87.

Single biggest risk: distribution coverage on free cash flow. FCF covered the distribution 1.48x in FY22 and just 0.63x in FY25, with the shortfall funded by $2.510B of net new debt while capex nearly tripled to $5.620B. Operating cash flow still covers 1.84x and the capex is visibly converting into EBITDA growth — which is why this is a Buy — but if the capex cycle extends without EBITDA following, the distribution stops growing and a 5.82% yield re-rates to 7%.

Most fragile assumption in the price: that the elevated capex is a cycle rather than a level — i.e., that $5.620B of FY25 capex moderates while EBITDA keeps compounding at 9-10%, mechanically repairing free-cash-flow cover.

Provenance and disclosures

• Traceability: 1 knowledge-base claim tagged to EPD (Lyn Alden, 2023-03-29, bullish, conviction 72, skill 1.1) plus 1 further claim naming Enterprise Products Partners directly (Lyn Alden, 2025-03-20, conviction 60). Supporting model-level lane: Doomberg (2025-12-22 conviction 75; 2026-05-27 conviction 60; 2026-07-16 conviction 80) and Lyn Alden (2025-01-23 conviction 80; 2022-08-08 conviction 78). Counterweight: Jordi Visser (2026-07-25, neutral, conviction 55, skill 2.0) on why commodity-linked equities rarely re-rate. Net conviction: positive but thin at the name level, coherent at the model level. Quarantined and misattributed entries excluded. All quotations verbatim with source and date.
• Data as-of: fundamentals 2026-06-30 (Q2'26, filed 2026-06-30; annual FY25 ended 2025-12-31, filed 2026-02-27) · estimates 2026-08-04 · prices 2026-08-04 (quote $37.87, 50-DMA $37.62, 200-DMA $35.41, RSI 51.7, MACD +0.23, 52-wk range $30.19-$39.80, beta 0.469) · knowledge-base claims to 2026-07-25. All market data comes from the pre-pulled Synthos data file for EPD; no figure has been sourced from memory or any other retrieval.
• IMPORTANT vendor-data warning. The key-metrics and ratios block for EPD is partly corrupted and several standard fields are zeroed or wrong. Specifically: enterpriseValueTTM is reported as $81.932B — identical to the market cap — which is incorrect for a company carrying $34.926B of total debt; netDebtToEBITDATTM, currentRatioTTM, debtToEquityRatioTTM, debtToAssetsRatioTTM, returnOnAssetsTTM, returnOnInvestedCapitalTTM, assetTurnoverTTM, cashRatioTTM and several turnover ratios are all reported as 0. This dive therefore hand-computes enterprise value as market cap ($81.932B) plus net debt ($33.681B) plus minority interest ($0.831B) = ~$116.4B, and hand-computes net-debt-to-EBITDA (3.17x), EV/EBITDA (~11.0x) and interest coverage (5.05x) from the raw statements. The vendor's evToEBITDATTM of 7.72x is understated because it uses the broken EV and should not be relied upon. Every affected figure is labeled where it appears.
• Named data gaps: seg_geo is EMPTY — Enterprise Products does not report geographic revenue segments in this dataset, which is a real gap for a business with international marine export exposure, and no geographic split is asserted anywhere in this dive. Segment revenues are reported gross of intersegment eliminations of -$131.540B (FY25), roughly 2.5x consolidated revenue, so segment shares of consolidated revenue are not meaningful and are not calculated. Maintenance capex and growth capex are not separately disclosed, which limits precision on the distribution-coverage analysis in section 3. Analyst coverage is moderate: 10 EPS analysts on FY27 and FY28, but only 4-6 on FY26 and FY29-30, and the FY26 revenue estimate rests on 4 analysts.
• Fair-value method: scenario multiples on consensus EPS, each cross-checked against a demanded-distribution-yield calculation. Base $40 = 12.5x FY27E EPS of $3.193, cross-checked at the $2.205 distribution divided by a 5.50% demanded yield ($40.09) and against the street target ($40.60). Bear $30 = 10.5x a $2.85 EPS, cross-checked at a 7.35% demanded yield ($30.00). Bull $50 = 13.0x FY30E EPS of $3.84, cross-checked at a distribution grown ~3%/yr to $2.48 at a 5.00% demanded yield ($49.63). No discounted cash flow is used and none is implied.
• Sell-side note: consensus target $40.60 (high $45, low $38), 7.2% above the current price, with even the lowest target above spot, on ratings of 34 Buy / 9 Hold / 2 Sell. FMP composite rating C, overall score 2 of 5. Earnings record across the last five reported quarters is genuinely mixed: 3 beats and 2 misses (beat 2026-07-30, missed 2026-04-28, beat 2026-02-03, missed 2025-10-30, beat 2025-07-28) — stated rather than smoothed.
• Structural note: EPD is a limited partnership. Units carry K-1 tax reporting, return-of-capital distribution treatment and account-type considerations that differ from corporate shares. This is disclosed as a structural fact and is not tax advice.
• Not investment advice. Independent research, educational and informational only, never personalized.
• Version: 2026-08-04-full.