MMM Q3 2026: Better Earnings. Which Cash Reaches Shareholders?

The 60-second read

Five things to know

1. Underlying performance improved. Adjusted organic growth was 3.3%; Safety and Industrial led with 7.0%. Transportation and Consumer were less strong.
2. Definitions changed. Adjusted results now exclude acquired-intangible amortization, with prior periods recast. Compare the new $2.50 with the recast $2.23.
3. The cash bridge matters. Quarterly cash adjustments total $967M, mainly after-tax litigation and PFAS payments. Those dollars did leave the company.
4. The balance sheet has more claimants. The new Madison venture brings redeemable minority interests and a put derivative alongside ordinary borrowing debt.
5. AI opportunity is conditional. Amazon warrant milestones are not booked revenue or guaranteed orders. New operating evidence must earn any higher valuation.

Three scenarios, one probability-weighted number

Twelve-month analytical value cases, not forecasts of where the market must trade.

Signposts: what would change our mind

Thresholds are our prospective tests, not company guidance. The prior Q2 public episode’s HOLD call is preserved; a prior research-packet model/signpost set was not verified, so no invented historical grade is shown.

The tape and the price of recovery

▶

The market welcomed the report: the saved quote was $172.72 at 11:51:37CDT,6.61% above the previous $162.01 close. That observation is useful evidence of expectations moving, but it is not a completed-session return. We hold the price and clock fixed throughout this packet so the valuation can be reproduced even if trading moves later. A rising share price does not prove the reported cash and adjusted cash definitions are equivalent.

The provider’s $89.08B market capitalization uses a different share convention from our $90.26B diluted-share model. We use 522.6M weighted diluted shares consistently to avoid switching denominators between valuation methods. The quarterly 10-Q also reports 516.73M actual common shares at September 30. Neither figure should silently replace the other: one describes actual shares at a date; the other reflects the earnings-period diluted denominator. All per-share models here retain the latter.

The print: growth, comparability and cash

The headline operating story is constructive. Sales rose 1.0% on a GAAP basis to $6.584B. Removing the exited manufactured-PFAS business from the prior comparison produces 4.2% adjusted growth, including 3.3% adjusted organic growth. These are different denominators, not competing answers. The adjusted figure helps evaluate the continuing commercial portfolio, while the GAAP total records the dollars actually reported. Both belong in the analysis.

Adjusted diluted EPS increased from a recast $2.23 to $2.50, or 12.1%. The company now excludes acquired-intangible amortization from adjusted results and recast prior periods. That change affects comparisons with older forecasts and previously reported adjusted EPS. GAAP EPS was $2.63 versus $1.55, but the large difference includes investment-value movements. We therefore separate improved industrial profitability from the accounting effects that made the GAAP percentage increase much larger.

The saved vendor calendar reports actual EPS $2.50 and revenue $6.6B, but the primary statement provides the exact $6.584B. We do not calculate a clean EPS beat against $2.38 without verifying that estimate’s accounting basis. The issue matters because a new exclusion can improve the apparent surprise without changing cash. For the same reason, the FY2026 saved consensusEPS $8.93701 is retained as a dated snapshot rather than treated as a fully updated response to today’s guidance.

Where the operating improvement comes from

Safety and Industrial drove the improvement. Revenue reached $3.182B, up 9.1% in total and 7.0% organically; operating profit increased from $792M to $960M. Acquisitions added 3.1points and divestitures subtracted 1.2points from its sales change. The distinction prevents the new Madison business from being counted as internally generated growth. Its roughly $90M contribution is commercially relevant, but the 10-Q says its operating income was not significant this quarter.

Transportation and Electronics is a more restrained part of the story: sales were $1.999B against $1.992B, organic growth 0.5%, and operating profit fell to $468M from $484M. Consumer sales slipped to $1.307B from $1.312B, with organic growth negative 0.2%, even as profit improved from $292M to $300M. A broad recovery would be more persuasive if these businesses began contributing stronger sales alongside the industrial division.

Segment comparisons require the current recast definitions. ManufacturedPFAS and respirator/asbestos matters moved into Corporate, and acquisition-related costs and acquired-intangible amortization are also handled there under the described presentation. The current quarter’s segment profits sum to $1.728B, then Corporate reduces consolidated operating profit by $170M to $1.558B. Segment margins cannot be substituted for the whole-company margin without that reconciliation. Corporate costs still affect shareholders even when they are not allocated to a selling division.

The earnings bridge: operating gains and investment marks

The $281M pretax increase in the value of retained Solventum stock is below operating income. It helped turn the other-income line positive and contributed a $0.60 after-tax EPS benefit under the company’s reconciliation. It is not revenue from selling more industrial products and is not an operating cash receipt. Its separate $2.254B carrying value belongs in the asset analysis, with an explicit valuation haircut if eventual realization and taxes differ.

Ordinary borrowing interest was $127M for the quarter, while the broader reported interest-expense line was $217M. The difference includes imputed interest on settlement obligations. Treating the entire reported amount as interest on bank and bond debt would misstate the financing burden. Conversely, excluding settlement accretion from adjusted earnings does not erase the obligation. Our valuation recognizes recorded legacy claims separately rather than assuming the adjusted EPS is free of future demands.

The new venture complicates the final step from consolidated profit to common shareholders. The $37M line includes both noncontrolling income and changes in redeemable noncontrolling interest; the 10-Q records a $41M redemption-value adjustment in the equity rollforward. This is why a simple minority profit percentage is not a reliable ownership shortcut. We retain the full primary bridge and the actual diluted denominator instead of forcing a cleaner-looking story.

Cash quality: the exclusions have consequences

The quarter generated $486M of operating cash and spent $256M on capital assets, leaving $230M of reported free cash flow. The company adds back $967M of specified cash impacts to show $1.197B adjusted free cash flow. Most of the difference is $894M of litigation andPFAS payments after tax. The remaining $73M combines acquisition and transformation cash. These exclusions can help isolate ongoing industrial performance, but they are not interchangeable with cash available to common shareholders.

Timing cuts both ways. Last year’s third quarter included net legal recoveries in the adjusted bridge; this year included payments. Reported quarterly FCF fell 85.0%, while adjusted FCF fell 8.6%. Across nine months, reported FCF improved to $1.342B from $61M, whereas adjusted FCF was almost flat at $3.086B versus $3.083B. Neither one quarterly collapse nor one year-to-date rebound fully describes the operating trend. The more useful question is whether operating improvement eventually increases cash after the obligations are paid.

The detailed 9M cash statement also records $557M consumed by receivables and $344M by inventory, partially offset by $315M from payables. These are real working-capital demands despite improved adjusted margins. The ending-balance day measures are deliberately labeled proxies because acquisitions, currency and sales mix affect them. They should prompt questions about collection and inventory conversion, not allegations of accounting misconduct. Stock compensation is modest at $46M this quarter, but the vendor’s zero field would have hidden it if used without the filing.

Balance sheet: debt, settlements and the new venture

Cash and current marketable securities totaled $5.145B against $14.997B of borrowings, leaving $9.852B net borrowing debt. Litigation liabilities are separate. The 10-Q records $7.5B ofPFAS-related environmental liabilities and $1.1B for Combat Arms earplugs. Those totals already reflect payments and discounting where applicable; the original nominal settlement headlines cannot be subtracted again as though nothing has been paid. Equally, the recorded balance is not a cap on possible losses: the company cannot estimate some excess exposures.

The public-water-supplier settlement originally contemplated $10.5–12.5B of nominal payments through 2036; the earplug settlement contemplated up to $6B through 2029. Our model uses the current recorded balances, not those original totals. The current portions,$2.8B environmental and $0.6B earplug, show that near-term liquidity claims remain meaningful. A simple net-debt-to-EBITDA ratio would miss them. Tax deductibility, payment timing and new claims can all change the economic burden.

Madison was purchased for $1.910B net of acquired cash through a venture with Bain. 3M owns 50.1% of voting interests and consolidates it under the variable-interest model. Bain contributed about $1.3B, and the venture also borrowed. Its interest can become puttable after five years and includes floor/cap provisions; the $971M redeemable interest and $340M derivative are therefore substantive economic claims. Consolidation gives 3M all the reported venture revenue, not all its economic value. Distributions require approval from both partners.

Three valuation routes, with obligations counted once

All three routes are analytical scenarios, not company guidance. Their equal-weight average is $152.43; we round the central value to $150. Our probability-weighted bear/base/bull value is $153. Our $150 base is 13.2% below the $172.72 quote. HOLD recognizes the operating recovery while declining to treat the recent rally as evidence of a margin of safety. High uncertainty requires a 25% discount to base value, or $112.50, before a fresh-buy valuation threshold is met with the thesis intact.

The legal adjustment uses $8.6B of recorded claims with an assumed 20% tax shield, producing $6.88B. This is a modeling convention, not a legal estimate or a promise that every payment receives that tax benefit. Solventum stock receives a 25% realization/tax haircut. Unknown excess legal exposure is not assigned a fabricated liability. These choices explain why the scenario range remains broad. If the tax shield vanished, value would be lower by approximately $3.29 per diluted share before any other change.

The DCF starts with normalized cash after interest, so it is an equity-cash model. Borrowing debt is not subtracted a second time, and cash is not added again when its income is already embedded in the normalized stream. Known legacy claims are deducted once because the starting adjusted-cash measure excludes their payments. Minority and put claims remove cash economics attributable to other owners. The sales method instead begins with enterprise value and therefore subtracts net debt. Mixing these conventions would materially distort the answer.

At the current price, the same cash model requires about 10.77% annual growth after the $4.8B2027 starting year, versus our 5% base assumption. That is a demanding requirement for a company whose current adjusted organic sales growth is 3.3%, although margin improvement and working-capital efficiency can let cash grow faster than sales. Roughly 74.5% of the gross baseDCF comes from terminal value. The sensitivity grid makes that dependence visible; it is a reason for moderate conviction, not a reason to present the model as precise.

Street targets: useful context with stale clocks

The newest saved target news dates to September: Bernstein $171 and Wells Fargo $195. July observations include Mizuho $180, Deutsche Bank $192 and UBS $218; Goldman’s June target was $190. These marks reflect different dates, assumptions and recommendations. We do not average them into a newly refreshed post-earnings consensus, and no fresh October 20 target revision has been verified in the saved feed. The provider’s ratings-history endpoint is a quantitative model grade, not the sell-side buy/hold/sell distribution.

The saved annual path is $25.457B sales and $8.937EPS for 2026,$26.604B and $9.862 for 2027, and $27.573B and $10.659 for 2028. The 2027 earnings and revenue estimates have 10 and 11 contributors respectively. These numbers help anchor plausible operating expectations, but they do not establish that every contributor has incorporated today’s accounting-definition change. We use 2027 as a disclosed cross-check and retain the uncertainty rather than converting the gap to current guidance into a manufactured upgrade trend.

Our disagreement with more optimistic targets is mainly about cash available after claims and the valuation paid for the recovery. A higher target can be reasonable if industrial growth broadens, margins remain durable and known obligations are paid without repeated new provisions. The question is testable: cash after capital investment and claims must improve, not merely adjusted EPS. This is why our signposts emphasize cash, net debt and accruals alongside organic growth.

Management: separate execution from presentation

Nine-month cash returns to shareholders totaled $4.441B, while reported FCF was $1.342B and adjusted FCF $3.086B. Those cover 30.2% and 69.5% of the distributions respectively. The company also spent $1.910B on Madison. Net borrowing proceeds of $2.525B, partner contributions of $1.277B, employee share proceeds and other investing cash helped fund the broader allocation. It would be misleading to describe every buyback as funded solely by cash generated after all obligations.

The October 19 Amazon agreement creates a warrant for up to 3.2M shares, with vesting contingent on cumulative sales of $6B over seven years and other milestones. That threshold is not a backlog number, contracted guarantee or current revenue forecast. It aligns an important customer with potential optical-connectivity growth while creating contingent dilution. We do not include all warrant shares as if already earned, nor capitalize the full sales threshold as a secured order. The actual realization path is a future research checkpoint.

Management’s disclosure of the new adjusted definition is transparent enough to rebuild the comparisons, but investors must perform that work. Excluding acquired amortization can be useful for comparing current operating activity; it also makes acquisition cost less visible in adjusted profit. We retain the acquisition price and future capital claims in valuation. Stronger margins are valuable only if the purchases and restructuring required to produce them create adequate cash returns.

1. How much of the $4.8B adjusted cash outlook converts to cash after legal and transformation payments?
2. What portion of Safety and Industrial growth persists without acquisition contribution?
3. What cash return is required from Madison after Bain’s economic rights and venture debt?
4. Which Amazon milestones turn conditional opportunity into booked sales and collections?
5. How will buybacks be sized relative to net debt and current settlement obligations?

Risks and the strongest competing cases

The strongest bull case is an improving industrial company whose productivity work is converting modest sales growth into much stronger profit. Safety and Industrial’s 7% organic growth is meaningful; the adjusted 26.2% margin shows that operational execution can matter even without a booming economy. Portfolio changes, optical connectivity and a disciplined acquisition could create a higher-quality mix. As large legacy payments decline, operating cash that currently goes elsewhere could become available to common shareholders.

The strongest bear case is that normalization becomes an excuse to ignore recurring economic demands. Legal payments, transformation projects, acquisition costs and minority claims can remain material for years even if no single item is considered ordinary operating expense. The stock’s rally brings forward some of the hoped-for benefits. If other segments fail to accelerate, working capital continues consuming cash and buybacks keep outrunning available funds, higher adjustedEPS may not translate into equivalent per-share economic value.

Our HOLD stance sits between those cases. We recognize real improvements rather than treating every adjustment as suspect. We also require claims to be funded rather than assuming a non-GAAP exclusion makes them disappear. The $110–210 value range reflects meaningful uncertainty in both directions. It should not be read as a statistical confidence interval. New material legal obligations, a failure of the current tax assumptions or persistent weak cash conversion would justify a lower model, while broad growth and stronger collections could justify a higher one.

What to watch next

The immediate next evidence is the year-end cash statement and the updated legal footnote. We will check whether the operating improvement survives a full-year comparison and whether the gap between reported and adjusted cash narrows. The company’s quarterly release and 10-Q are available, but this packet does not claim to have reviewed a call transcript. The vendor calendar suggests January 19 for the next print; without issuer confirmation, that remains an estimate and is not presented as an announced date.

Debt maturities, venture distributions and settlement payments do not arrive on the same schedule. A useful follow-up therefore tracks the balance sheet as well as annual EPS. The current settlement balances are the starting point, not a fixed promise about all future legal spending. We will preserve these definitions and date-stamped assumptions so the next review can grade the new signposts fairly. Moving a threshold after the outcome is known would defeat the purpose of publishing it now.

Sources, complete tables and calculation methods

Primary financial figures come from nine actual SEC earnings exhibits, the current 10-Q and three audited annual reports. Every source file and derived data table is retained with its URL and SHA256 hash. The current and historical income bridges and balance sheets were reconciled in code. The balance sheet includes redeemable noncontrolling interest outside permanent equity, so assets equal liabilities plus redeemable interests plus equity. Ignoring that separate category would create a false accounting discrepancy.

The saved vendor current-quarter cash-flow record contains zero placeholders for net income, stock compensation and depreciation and combines categories differently from the filing. We use the primary $46M quarterlySBC and actual 9M cash statement rather than those zeros. The vendor profile still describes a Health Care segment despite the completed 2024 separation; the current company has three operating segments. Market prices and analyst snapshots remain useful, but stale business descriptions and incomplete statement mappings do not override primary evidence.

Reported free cash flow is operating cash less gross capital expenditure. Adjusted free cash flow follows the company’s special-item reconciliation. Gross margin, operating margin, tax rate, growth rates, cash coverage and per-share balances are computed from those sources. Receivable and inventory days use ending balances and 91days, not average balances. The new amortization exclusion is a non-GAAP presentation change with recast comparatives, not a GAAP restatement. Older adjusted figures are not silently relabeled onto the new basis.

The valuation separates observations from judgments. The price, shares, balance-sheet claims and estimates are saved observations. Growth, discount rates, multiples, tax haircuts, scenario probabilities and thresholds are analyst assumptions. The equity DCF discounts 2027–2031 annual cash flows without a short stub and applies a terminal value at 2031. Legal and minority adjustments are applied once. The enterprise-sales route instead subtracts net borrowing debt because it begins before financing. Unknown legal losses remain an explicit limitation rather than a fabricated forecast.

Our prior public Q2 episode and its HOLD title remain part of the channel’s history. We have not verified an earlier packet’s exact fair value and numeric signposts, so we do not invent a performance score or imply this $150 value was published then. The new episode focuses on the reported-versus-adjusted cash distinction and the new venture economics. Glossary: GAAP is the reported accounting basis; adjusted is the company’s non-GAAP basis; FCF is cash after capital expenditure; RNCI is redeemable noncontrolling interest; enterprise value values the operating business before net borrowing claims.