KMI: Bigger Projects — Who Funds the Growth?

The 60-second read

Five things to know

1. Earnings improved. The cash question comes after construction and distributions.
2. Quarterly funding tightened. Capex growth exceeded operating-cash growth.
3. Project economics need translation.5.7 xprojectEBITDA is not shareholder free-cash return.
4. Gross adjustments matter.$138 Maccrual and $71 Mgain partly offset.
5. Missing evidence stays missing. Current detailed notes and new broker revisions remain pending.

Three scenarios, one probability-weighted number

Analyst assumptions, not management probabilities.

Signposts: what would change our mind

Issuer guidance distinguished from analyst monitoring thresholds; no invented release appointment.

The price and the funding question

Kinder Morgan has added several billion dollars of natural gas construction commitments while its current operations are producing more earnings. Those two facts are connected, but they are not interchangeable. A project backlog measures future investment obligations and opportunities; it is not cash already available for dividends. This quarter's useful question is whether contracted growth can earn enough after financing, construction and maintenance to justify the capital committed. Our HOLD call uses a $34 analytical central value and a $25–$43 scenario range, with moderate conviction and high uncertainty around execution.

The saved October 21 regular-session close is $33.97, versus $33.57 the prior session. That observation is dated, not a verified earnings-response calculation. The release is established as October 21, 2026, in America/Chicago, but an exact actual release minute and session classification have not been established from the reviewed issuer evidence. The SEC acceptance clock is a filing timestamp and is not substituted for the release clock. Aftermarket bid and ask indications likewise are not treated as a completed trade or as the valuation anchor.

The covered security is NYSE KMI Class P common stock, denominated in U.S. dollars, with one vote per share under the current proxy. The separately registered KMI27A notes are debt, not an alternate common-stock identity. The enterprise-value calculation uses 2,227 million quarterly diluted weighted-average shares as an explicit proxy because a current quarter-end outstanding count is not supplied in the condensed release. This produces a slightly different equity value from the provider's market capitalization. That difference is disclosed rather than hidden by calling the share proxy a spot count.

Better earnings, a larger funding test

Quarterly revenue was $4.531 billion, up 9.3% from $4.146 billion. GAAP earnings attributable to Kinder Morgan were $764 million, or $0.34 per share, versus $628 million and $0.28. Adjusted earnings were $792 million and $0.35 per share. Adjusted EBITDA increased to $2.180 billion from $1.991 billion. These are genuinely better operating results, but a midstream revenue comparison includes commodity purchases and sales and cannot be interpreted as the same percentage change in fee volumes, margins or owner cash.

The year-to-date figures are useful because a single quarter can be affected by interest dates, seasonal demand and construction milestones. Through September, operating cash was $5.041 billion, capital expenditures $2.831 billion, and dividends paid $1.984 billion. The remaining $226 million is the precise table amount behind management's rounded discussion of roughly $230 million. This is positive, but it is a narrow residual relative to the $14.4 billion project backlog. A multi-year funding claim must be evaluated across the construction schedule rather than against one quarter's available cash.

Management still frames the annual outlook relative to its original budget. Adjusted EBITDA is expected to exceed the $8.6 billion budget by more than 6%, while adjusted EPS is expected to exceed $1.36 by more than 13%. The arithmetic lower bounds are therefore greater than $9.116 billion and greater than $1.5368; they are not exact point forecasts. The saved provider's $8.826 billion EBITDA consensus is below that primary-source lower bound, indicating that the provider snapshot should not be used to overwrite current guidance. There is no verified new consensus beat figure in this packet.

Four businesses, different economics

Natural Gas Pipelines generated $1.489 billion of adjusted segment EBDA, versus $1.403 billion. It remains the largest operating contributor, with growth across El Paso Natural Gas, KinderHawk, Tennessee Gas Pipeline and Texas Intrastate. Reported transport volumes rose about 3% and gathering about 14%. These volume series are presented pro forma for acquired and divested assets and at Kinder Morgan's share of joint ventures. Earnings contributions do not always use that same ownership and timing convention, so multiplying the volume growth rate by last year's profit would be a false shortcut.

Products Pipelines reported only $172 million of segment EBDA but $313 million after adjustments. The $138 million legal, environmental and other accrual is important to that difference; it is not erased economically because the company excludes it from an adjusted metric. Refined-product volumes declined about 3% and crude and condensate volumes about 11%. Part of the latter relates to converting Double H to natural gas liquids service. That conversion changes the relevant product category and is not automatically evidence that every lost crude barrel represents lost enterprise economics.

Terminals reported $385 million of segment EBDA, including a $71 million divestiture gain, while adjusted segment EBDA was $314 million. CO2 adjusted segment EBDA rose to $184 million from $136 million with stronger commodity prices and production. Together the four adjusted segment figures sum to $2.300 billion, not the consolidated $2.180 billion adjusted EBITDA. Corporate costs and other reconciliation items separate those measures. The packet therefore uses segment results to explain operating sources, not as a substitute for the consolidated cash or leverage denominator.

What sits inside adjusted earnings

The income statement reconciles $1.160 billion of operating income to $980 million of pretax earnings after equity-investment earnings, net interest and other items. Taxes reduce that to $790 million of consolidated net income, and noncontrolling interests take $26 million, leaving $764 million attributable to KMI. Interest is a real claim on enterprise cash before the common dividend, and minority partners also have an economic claim. A valuation based on unlevered cash must recognize financing separately; a valuation based on cash already after interest must not subtract debt twice.

The net GAAP-to-adjusted earnings gap is only $28 million, but its small size conceals large offsetting items. The reconciliation adds $138 million of legal, environmental and other accruals and $3 million of risk-management items, removes a $71 million divestiture gain, and includes a negative $42 million tax adjustment. These add to $28 million. Calling this merely a small adjustment would miss the underlying litigation and asset-sale economics. The correct question is whether each component is likely to recur, require future cash, or change future earning assets.

Nine quarters show that both GAAP and adjusted EPS move with seasonal operations and nonroutine items. The original second-quarter 2026 exhibit reports GAAP EPS of $0.39 and adjusted EPS of $0.37. A saved prior episode metadata field used $0.37 without the same basis label; the current historical chart follows the primary exhibit and leaves the old publication untouched. Weighted-average diluted shares have changed only modestly, so the current earnings improvement is not mainly a shrinking-share-count story. The current $0.2975 quarterly dividend rises much more slowly than adjusted EPS.

Nine checks on earnings quality

Our quality assessment distinguishes nine questions: stock compensation, adjusted exclusions, below-the-line income, minority leakage, cash conversion, working capital, maintenance versus growth spending, taxes, and guidance delivery. None is resolved by a single high EBITDA number. Current-quarter stock compensation is not separately disclosed in the earnings release, so the absence of a line is not treated as zero. The latest audited annual note reports $68 million of expensed restricted-stock awards and $11 million capitalized in 2025. These are economic costs even when their cash treatment differs.

The annual comparative note reports expensed award costs of $68 million, $64 million and $60 million for 2025, 2024 and 2023, plus capitalized costs of $11 million, $10 million and $9 million. Earlier annual filings contain different comparative award totals. We retain those originals and use the latest audited comparative table consistently rather than combining incompatible vintages. Similarly, current detailed receivable and inventory balances are not provided in the condensed release. No invented current-quarter receivable-days or inventory-days series is used to make the quality score appear complete.

Operating cash of $1.590 billion is more than twice attributable net income, but depreciation and deferred taxes explain much of that conversion. The release also includes equity-investment earnings deducted and distributions added back, which matters for a joint-venture-heavy infrastructure business. The combined working-capital-and-other line is a small $3 million outflow this quarter, not a large cash windfall. Nevertheless, strong cash conversion does not guarantee dividend coverage after construction spending. The current quarter demonstrates exactly why those are separate questions.

The cash left after construction

The cash bridge begins with $764 million of attributable earnings and restores $26 million of minority earnings, $628 million of depreciation and $194 million of deferred taxes. It removes $231 million of equity-investment earnings, adds $212 million of distributions and subtracts $3 million of working-capital and other items. The result is $1.590 billion of operating cash. Capital spending of $1.045 billion leaves $545 million of free cash flow. Dividends paid were $665 million, leaving negative $120 million after dividends for the quarter.

That deficit does not by itself imply an imminent dividend cut. Capital projects can create uneven spending patterns, and the year-to-date residual remains positive. But the direction deserves attention: operating cash grew 12.4% year over year while GAAP capital spending rose 31.8%, leaving free cash flow down 12.2%. The dividend's annualized $1.19 is about a 3.50% yield on the dated price. A yield is a distribution rate, not evidence that growth projects are fully self-financing or that the equity cannot lose value while construction progresses.

Annual cash flows provide the longer test. In 2025, operating cash of $5.917 billion less $3.026 billion capital spending left $2.891 billion before $2.604 billion of dividends. The $287 million residual excludes $648 million of acquisitions and $178 million of investment contributions. Those additional uses would turn the diagnostic residual negative before asset-sale proceeds, excess joint-venture distributions and financing. In 2023, an $870 million deferred-revenue inflow helped operating cash. We therefore avoid treating the strongest historical operating-cash year as a permanent run rate.

Company-adjusted net debt is $32.229 billion, up $513 million from December, even as the rolling EBITDA denominator improved enough to lower leverage to 3.5 times. Lower leverage does not necessarily mean absolute debt repayment. The August issuance of $1.15 billion at 5.55% and $600 million at 6.15% provides term funding, but those coupons remain future cash obligations. The last detailed liquidity disclosure reviewed is June 30: a $3.5 billion revolver with about $3.2 billion available. That is dated capacity, not a verified September availability figure.

Three routes and a distinct reverse case

Our forward discounted-cash-flow case explicitly models unlevered cash after taxes, maintenance and expansion reinvestment, and economic stock compensation, before financing. The five assumed annual cash amounts are $5.0, $5.2, $5.6, $6.1 and $6.6 billion. At an 8.5% discount rate and 2.5% terminal growth, subtracting $32.229 billion of company-adjusted net debt and dividing by the 2,227 million-share proxy gives $29.17 per share. These are analyst assumptions, not management guidance and not a mechanical annualization of this quarter's reported free cash flow.

The cash scenario requires judgment about minority and unconsolidated-joint-venture economics. We treat the projected cash as attributable operating cash on a look-through basis and do not add a second standalone joint-venture valuation. The company-adjusted net-debt bridge excludes specified fair-value and currency effects, and is not identical to every balance-sheet debt line added together. These approximations are explicit sources of uncertainty. Terminal value is a large part of the result, so modest changes in the discount rate and terminal growth produce a wide equity-value range after fixed debt claims.

The second route assumes $10 billion of normalized adjusted EBITDA valued at twelve times, less the same net debt, producing $39.41 per share. This EBITDA and multiple are assumptions; neither is a promise about the next year. The third route assumes $1.75 normalized EPS at twenty times, giving $35.00. Our $34 central judgment sits among these methods rather than pretending they are equally precise. A $25 bear case, $34 base case and $43 bull case at 25%, 50% and 25% weights also produce $34.00. The range describes scenarios, not statistical confidence.

The reverse case is distinct from the forward model. It begins with $5.0 billion cash in year one and solves a constant growth rate across the next four intervals that reproduces the dated $33.97 price under the same discount, terminal and debt assumptions. The solved rate is about 10.29% a year. It is cash growth, not revenue growth, not dividend growth and not the growth rate of the forward nonuniform series. This makes the market's requirement visible without asserting that the reverse case is management's plan or the only path that can justify the price.

Dated Street evidence

The saved target consensus is $37.20, with a $36 median and a $35–$43 reported range. These are provider snapshots, not a verified set of post-release analyst revisions. We retain dated records from RBC, Stifel and Morgan Stanley only where the headline and attribution agree. Other records assign a Capital One or Jefferies headline to Morgan Stanley and are excluded. A precise target with a mismatched author is not reliable evidence merely because it has a publication date and a familiar bank name.

The provider's annual estimates illustrate why source precedence matters. Its 2026 EBITDA mean is below the company's newly stated lower bound, while its EPS estimate is near the corresponding EPS bound. We do not label a stale estimate as a current market forecast, and we do not manufacture a percentage earnings surprise. The forward revenue and earnings table remains useful context with analyst counts, but the primary release controls reported figures and guidance. Updates should be compared using both publication date and fiscal-period identity before changing the interpretation.

Williams, ONEOK and Targa are context peers, not interchangeable valuation templates. Their business mixes, leverage, distributions and exposure to gathering, processing, liquids and interstate pipes differ. Saved provider quotes and estimates are retained for a dated comparison, with missing observations left missing. This packet does not transplant a peer multiple directly into KMI's equity value. Our twelve-times EBITDA route is an explicit analyst scenario that must be tested against KMI's own financing, construction schedule and project economics.

Projects, incentives and execution

The enlarged project backlog creates a more demanding execution test than the prior quarter's record-results narrative. Southeast Connector carries an approximately $3.1 billion KMI-share cost and a phased fourth-quarter 2030 target. Permian Link has an approximately $4.2 billion total cost, about $1.6 billion KMI share, and a third-quarter 2030 target. Permits and final scopes matter. A long-term take-or-pay contract with an investment-grade customer can reduce utilization risk, but it does not eliminate construction inflation, delays, financing costs or counterparty concentration.

Management describes a $13.7 billion subset of the backlog at 5.7 times first-full-year project EBITDA. The simple implied EBITDA is roughly $2.404 billion once that subset is fully productive, and the reciprocal is about 17.54% EBITDA on stated cost. Neither number is a shareholder free-cash yield or an internal rate of return. Taxes, interest, sustaining capital, minority interests, timing and cost overruns stand between project EBITDA and common cash. The subset excludes specified CO2 enhanced-recovery and gathering-and-processing projects, so it is also not the entire $14.4 billion backlog.

The original Q2 release said $9.6 billion of backlog. The current release describes $5.4 billion of additions, $0.7 billion placed in service and a $14.4 billion closing amount, a rounded roll-forward that leaves a $0.1 billion residual from the original figure. We disclose the residual rather than rewriting history or inventing an undisclosed project. The prior episode's $36 central value and HOLD call are retained as historical context; our current $34 value reflects a tougher funding and return test, not denial that operating earnings improved.

The proxy uses distributable cash flow per share, leverage and adjusted segment EBDA among executive performance measures. Those can encourage financial discipline, but adjusted measures can also differ from cash left after all expansion spending. Here the acronym DCF in compensation discussions means distributable cash flow, while our valuation DCF means discounted cash flow. One vote per common share, the current audited governance record and the retained ownership evidence provide context; none substitutes for monitoring actual project spending and common-shareholder cash returns.

What can break the thesis

Construction and permitting are the largest execution uncertainties in this thesis. Multi-year pipelines can consume cash well before they earn full-year returns. A delay can lower value even if the eventual EBITDA forecast is unchanged, because investors carry financing costs and wait longer for distributions. Cost escalation has the same asymmetric effect when revenues are contracted but construction expenses remain variable. Watch updated KMI-share budgets and commercial-service dates, not simply a rising backlog balance that can reflect both attractive commitments and larger obligations.

The $138 million legal, environmental and other accrual needs detailed follow-up. The condensed release does not establish an exact case-by-case allocation, so we do not attribute it to a specific dispute without the detailed note. Historical environmental and pipeline litigation exposures remain part of ordinary business analysis. They are not an automatic security exclusion under the current policy, and a non-GAAP add-back is not evidence that any eventual payment is harmless. The next detailed filing should clarify timing, estimation ranges and insurance or recovery assumptions where disclosed.

Commodity and volume risk still exists in a fee-heavy network. CO2 results and some marketing activities are exposed to commodity prices, and disclosed hedges cover specified quantities and periods rather than every future dollar. Products volumes are declining in the current pro forma series while natural gas expands. Financing risk interacts with those operations: refinancing is easier when EBITDA grows and markets are open, but current low leverage ratios can reverse when the denominator weakens. The company remains a continuing listed common issuer; historical note delistings and the completed Canadian sale are not new common-stock termination events.

What would change the call

The next detailed September-quarter filing is the first checkpoint, with its exact filing date unconfirmed. It should expand the condensed cash, debt and legal disclosures without allowing investors to retroactively call missing current data a pass. Full-year results will then test the greater-than guidance bounds and the year-end leverage outlook near 3.6 times. The thresholds in this packet are analytical monitoring rules unless explicitly labeled as issuer guidance. They are not promised release appointments or automated trading instructions.

For the funding thesis, track rolling cash after total GAAP capex and dividends, and then separately track acquisitions and investment contributions. A positive rolling residual with stable leverage would support management's internal-funding argument. Persistent negative residuals combined with rising debt would weaken it, even if EPS keeps rising. A second checkpoint is whether updated project budgets and service dates remain credible. The smaller storage and gathering expansions have earlier dates and can provide evidence before the large 2030 projects reach full operation.

A better entry price alone would not repair deteriorating project economics. Conversely, successful completion, sustained cash generation and clearer legal obligations could justify raising value even without a lower quote. Our HOLD call is therefore conditional: the operating network is improving, but the dated price is close to our central value and the funding margin is limited. We would revisit the call on a material change in price, cash conversion after construction, verified project returns or financing terms, using fresh primary evidence rather than a recurring dividend headline.

Primary sources and calculation record

The historical dataset contains nine original quarterly earnings exhibits from September 2024 through September 2026 and three audited annual cash observations for 2023 through 2025. Quarter numbers come from each original release, while the annual comparative table is consistently taken from the latest audited 2025 report. The raw filings, extracted tables, fetch times, URLs and hashes are retained alongside the packet. Reported values, calculated ratios and analyst assumptions have separate labels so a reader can distinguish an issuer disclosure from a model input.

Each quarterly operating-cash, capex, free-cash and dividend bridge is independently recomputed, and the four 2025 quarters reconcile to audited annual operating cash and simple free cash flow. Net income is reconciled from pretax income, taxes and minority interests. The current adjustment components sum to the stated $28 million. Forward valuation, reverse cash growth and probability-weighted scenarios are computed from the saved parameters. None of these checks proves future cash delivery; they establish arithmetic and source consistency for the current analysis.

The source limitations are deliberate. Current detailed Q3 notes, quarterly stock compensation, detailed quarter-end receivables and inventories, exact release minute and fresh post-print analyst revisions are not invented. Historical next-session price windows are descriptive comparisons and are not represented as causal earnings moves. The provider data is saved for cross-checking and market context, with mismatches retained. The original Q2 claim, prior publication files and all known IDs remain untouched. Public video and podcast links will be filled only from verified publication receipts later in the workflow.

• Q3 2024 primary earnings exhibit
• Q4 2024 primary earnings exhibit
• Q1 2025 primary earnings exhibit
• Q2 2025 primary earnings exhibit
• Q3 2025 primary earnings exhibit
• Q4 2025 primary earnings exhibit
• Q1 2026 primary earnings exhibit
• Q2 2026 primary earnings exhibit
• Q3 2026 primary earnings exhibit
• 2025 audited10-K, three annual statements and risks
• June2026 detailed10-Q, historical liquidity
• 2026 proxy: voting and compensation
• FMP saved market data, estimates and target records; October21,2026