General Mills: a 7.5% yield that needs the cash to stop falling

General Mills pays a 7.5% dividend yield. In fiscal 2026 it paid out $1.82 billion to shareholders, $1.32 billion in dividends and $0.50 billion in buybacks, from $1.63 billion of free cash flow. The gap was covered by other cash while total debt fell by $1.34 billion, and that is not a pattern a company can repeat.

At $32.57, the October 21, 2026 regular session close, the shares trade 4.2% below my triangulated Base value of $34. The Bear value is $23, 29.4% below the price, and the Bull value is $44, 35.1% above it. Confidence is grade C (14/25), the required margin of safety (MOS) is 40%, and the maximum buy price is $20.40. After hours the shares traded at $32.60 and a broker ask earlier showed $32.66, so the verdict does not depend on which of the three you use. The 52-week range is $31.18 to $49.71, which puts the price near the bottom of it.

Valuation route: a branded food company, valued on owner cash: operating cash flow less capital spending and stock compensation, after interest. Primary method: that cash capitalized at 11.8 times, with 8.0 times in the Bear and 15.3 times in the Bull. Secondary method: 9.4 times operating profit plus depreciation, less net debt. A dated check follows both, discounting cash flows and an exit value at 10%. Classification: standard company, 4 of 12 on the scorecard.

What General Mills is

General Mills sells packaged food and pet food through grocery and foodservice channels. It is a business of familiar brands that is losing volume and earnings at the same time. Fiscal 2026 net sales were $18.42 billion, down 5.4% from $19.49 billion, adjusted operating profit fell 16.1% to $2.81 billion and adjusted earnings per share fell 15.7% to $3.55.

Reported results are far worse than adjusted ones. GAAP operating profit was $886 million against $3.30 billion a year earlier, and GAAP earnings per share were negative $0.16, after $1.80 billion of goodwill and brand impairments (including $1.5 billion on the North America Pet reporting unit) and a $1.03 billion valuation loss on the planned Brazil divestiture. Those charges are non-cash and I do not subtract them from cash flow. They do tell you what the company paid for some of its past acquisitions and what it now thinks they are worth.

Management guides fiscal 2027 to a further decline: adjusted operating profit down 8% to 13% in constant currency, adjusted earnings per share of $3.00 to $3.20, organic sales between down 1.5% and up 0.5%. It plans at least $750 million of cost savings in the year, toward a $3 billion cumulative target through fiscal 2030. The business is being asked to shrink profit while savings hold the cash together.

The latest results

Management said the company is off to an encouraging start in fiscal 2027 and reaffirmed the full-year outlook, which includes free cash flow conversion of about 95% of adjusted after-tax earnings (85% in fiscal 2026). The quarter's own cash flow tells a quieter story: free cash flow of $207 million did not cover $331 million of dividends. One quarter is not a year, and the first quarter is seasonally light, so I do not read it as a coverage conclusion. It does mean the buyback stopped.

Which cash to value

General Mills reports free cash flow as operating cash flow less capital spending. I also deduct stock compensation, because shares issued to employees are paid for by owners, and call the result owner cash. The amount is small here ($79 million), so the choice matters less than for a technology company.

The fiscal 2027 line is $3.10 of adjusted earnings per share times 537.9 million shares, times the 95% conversion, less stock compensation at last year's level. All of it is my arithmetic from management's outlook, not a company cash flow guide. Fiscal 2022 to 2024 free cash flow, taken from the analysis I checked and not re-verified by me, was $2.75 billion, $2.09 billion and $2.53 billion, which makes the four-year decline to fiscal 2026 12.3% a year. Fiscal 2025 and 2026 I confirmed from the filings.

The base for valuation is $1.547 billion of owner cash, or $2.876 a share on 537.9 million diluted shares. It sits 2.8% above the $1.505 billion that guidance implies for fiscal 2027, so the Base already assumes that the decline stops about where it is. Capital spending is $540 million against depreciation and amortization of $555 million, so there is no large investment gap hiding in the number, and no sign that capital spending is being starved to flatter cash.

Cash before and after financing, and the debt

The framework asks for three cash flow levels once debt is a real burden, and General Mills is past the trigger: net debt of $13.18 billion is 3.9 times operating profit plus depreciation of $3.37 billion (an estimate, not a reported EBITDA).

Level A adds back net interest of $538.6 million at the adjusted tax rate of 21.1%. Fiscal 2026 shareholder payouts exceeded owner cash by $0.27 billion and debt was repaid on top of that, so roughly $1.6 billion came from somewhere else. The most likely source is divestiture proceeds, but the amounts are not itemized in the filings I read, so I label that figure unavailable. Total debt fell from $14.88 billion to $13.54 billion over the year, and the change includes currency and non-cash effects.

The fiscal 2027 leverage figure holds Q1 net debt constant and uses the guidance midpoint for operating profit (down 10.5%) plus last year's depreciation. If cash stays near $1.5 billion and the dividend near $1.3 billion, the room left to repay debt is about $0.2 billion a year. At that rate leverage does not come down by paying it off. It only comes down if profit stops falling.

Refinancing looks manageable in the near term. The current portion of long-term debt is $1.05 billion and commercial paper is $190 million. A $2.7 billion committed credit facility runs to October 2029 and is undrawn, and the company reports covenant compliance. Two notes due in early 2027 ($750 million at 3.2% and $500 million at 4.7%) were repaid with the proceeds of junior subordinated notes issued in the fourth quarter of fiscal 2026, which carry coupons of 4.75% and 5.25%. Net interest rose to $538.6 million for the year and to $142.2 million in the first quarter, so cheaper debt is being replaced with dearer debt. I did not have the full debt maturity ladder or the lease liabilities, and both are listed under data limits.

Shares, claims and the equity bridge

I use 537.9 million diluted shares, the first quarter weighted average. Shares outstanding on September 16 were 534.7 million, and the company repurchased $500 million in fiscal 2026 and nothing in the first quarter. A flat count is consistent with a business that is not issuing much stock, since stock compensation is $79 million a year.

Debt is deducted only in the secondary method, because the primary method already pays interest inside owner cash. Cash held for sale in Brazil ($30.7 million) and operating lease liabilities are not in the bridge. The lease liabilities were not in the excerpt of the quarterly filing that I could read, and I flag that as a data limit.

The primary valuation: owner cash times a multiple

The primary method capitalizes fiscal 2026 owner cash of $2.876 a share. The value is cash times a multiple, and the multiple is where the whole valuation sits, so here is what each one assumes.

The 10% rate is my assumption and a simplification (value equals cash divided by the rate less growth). It is there to show the reader what the multiples assume, not to claim precision. In the Base, owner cash has to stop falling and grow 1.5% a year indefinitely, from a base that guidance says will be about 3% lower in fiscal 2027. That needs the savings programme to offset inflation and reinvestment, and volume to stabilize. The Bear allows a persistent 2.5% annual decline. The Bull requires volume recovery with profit, which nothing in the current guidance shows.

Reinvestment is low, with capital spending about equal to depreciation, so little of the value depends on new projects earning a return. The weak point is leverage: net debt is 3.9 times operating profit plus depreciation, and interest takes about 19% of adjusted operating profit ($538.6 million of $2,811.5 million). The multiples are a starting point, not a table default, and the Bear is not a floor. If cash settled at $1.4 billion and the market paid 6 times, the shares would be worth $15.62, 52% below the price. I show that as a stress and not as a scenario.

All value in this method sits in the capitalized figure: there is no explicit forecast period, so the continuation value is 100% of the total and the dividend of about $1.3 billion a year is not modeled separately. The dated check below puts the first 19 months on a calendar.

The secondary valuation: operating profit plus depreciation

The second method values the whole business on operating profit plus depreciation and subtracts net debt. It uses fiscal 2026 adjusted operating profit of $2.8115 billion plus depreciation and amortization of $0.5552 billion, a total of $3.3667 billion, which is a proxy and not a reported adjusted EBITDA.

The secondary Base is $34.33. The 9.4 times is also an assumption, and it rests on the same recovery as the primary one: a business whose profit is falling needs a lower multiple than a stable one, and the 9.4 times only holds if the decline stops. With debt this size the equity value moves a great deal for a small change in the multiple: 0.5 turn of operating profit plus depreciation is $1.68 billion, or $3.13 a share.

The two methods on one basis

The two methods agree within 1.2%, and the table shows why: they assume nearly the same valuation of the same business. That is not independent confirmation. Both depend on the cash stabilizing and on the same net debt, so their agreement does not raise Confidence. It also means the market is already paying 9.1 times operating profit plus depreciation, close to what I use for the Base.

What the price already assumes

At $32.57 and 537.9 million shares the market value is $17.52 billion, or 11.3 times fiscal 2026 owner cash. At a multiple of 12 times, the price needs $1.46 billion of sustainable owner cash, 5.6% below last year and 3.0% below what guidance implies for fiscal 2027. At 11.8 times it needs $1.48 billion. In plain terms, the market is not paying for a recovery. It is paying for a stop.

That makes the downside case hard to ignore as well. If owner cash falls at the rate of the last four years, 12.3% a year, then fiscal 2028 would be about $1.3 billion, and the Bear value of $23 would look more like an expectation than a stress.

Bear, Base and Bull, with a dated check

These three cases change only the multiple and hold cash constant. They do not model different operating outcomes, and I say so plainly because the framework's three cases are meant to differ in the business, not just the price. I rebuilt the cases as a dated cash path to test them.

The Base barely moves ($33.54 against $33.93), which supports the headline value. The Bear is lower by $2, and the Bull higher by $4, because cash that differs between scenarios adds a second effect to the multiple. I keep the framework values of $23 and $44 as the stated range and show the dated figures beside them so the reader can see both. In the dated Base the exit value is 85.3% of the total over a horizon of 1.6 years (to the end of fiscal 2028, May 2028), with dividends of about $1.3 billion a year paid in the meantime. That is right at the level where I check whether the exit is supported by sustainable cash flows, and the flat cash outlook with capital spending near depreciation is the evidence. I count that dependence once, in the valuation robustness part of Confidence, and not again in the margin of safety.

At $32.57 the dated Base cash path returns 12.2% a year to the end of fiscal 2028 if the shares are then worth 11.8 times owner cash. That is a modeled owner cash return, not a forecast of dividends or the share price, and it requires the stabilization the Base assumes. It is also the return at risk: the dated Bear path, with lower cash and 8 times at the end, is worth $20.96 today.

A price check against history

I could not verify a long series of past multiples, so history carries no weight. Owner cash fell from about $2.2 billion in fiscal 2025 to $1.55 billion in fiscal 2026 (down 29.7%) and the shares are 34.5% below their 52-week high of $49.71, roughly in step with the cash. They are cheaper mostly because the cash is smaller. On guidance-implied fiscal 2027 earnings of $3.10 a share the shares trade at 10.5 times, but that is adjusted earnings, and I do not mix it with the GAAP figure. A forward PEG ratio is not meaningful with earnings falling, and a forward free cash flow growth rate is unavailable.

Sensitivity

One more case matters more than the grid. If fiscal 2027 cash lands where guidance implies ($2.80 a share) and the Base multiple holds, the value is $33.01. Owner cash would need to be $3.25 a share and the multiple 12 times or more to reach $39. Even the lowest cell of the grid, 10 times $2.50, gives $25.00, above the maximum buy price of $20.40. The entry price at a 40% margin of safety is $20.40 on the Base of $34, $20.60 on the secondary value alone and $20.12 on the dated Base. No reasonable single change creates a Buy at $32.57.

How the buy price is set

Confidence C calls for a margin of safety of 35% to 50%. I use 40% with no add-ons, because leverage, opacity and the terminal dependence are already in the multiples and in Confidence, and the framework does not let me count the same risk twice. The maximum buy price is $34 x (1 - 40%) = $20.40. Even at a 30% margin of safety the ceiling would be $23.80, still 27% below the price. It is a decision reference, not an order to place.

Three things are reported separately: what the shares are worth ($34), the modeled return from buying at $32.57 on the dated Base (12.2% a year, assuming the exit multiple holds), and the price that earns a chosen return. At a required 10% the dated Base gives $33.54. The 12.2% is not an invitation. It depends on the unchanged multiple, and a price with a 40% margin of safety is $20.40.

The dividend

The quarterly dividend is $0.61, $2.44 a year, a yield of 7.49% at $32.57. In fiscal 2026 dividends of $1.315 billion were 80.9% of free cash flow of $1.626 billion and 85.0% of owner cash. On the fiscal 2027 estimate the payout is 82% of free cash flow (cover of 1.21 times). The first quarter's dividend of $330.5 million was more than its free cash flow of $207.3 million, which is a seasonal pattern as much as a warning.

A high yield is a sign of risk as much as a promise of return. The market is asking for 7.5% because it expects either a cut or a falling price. If cash stays near $1.5 billion, the dividend leaves $0.2 billion to repay $13.6 billion of debt. I do not add the dividend to the value because the owner cash method already values the cash once, and I make no claim about whether the board will cut.

Reasons to own

• Brands that sell. Organic net sales were flat in the first quarter, and guidance for the year spans down 1.5% to up 0.5%, so volume is not collapsing.
• Cash generation survives the earnings decline. Free cash flow was $1.63 billion in a year that included $1.8 billion of impairments, and capital needs are modest at about $540 million.
• Savings are large against profit. At least $750 million in fiscal 2027 is 27% of fiscal 2026 adjusted operating profit. If they hold while volume stabilizes, owner cash could be flat.
• The price is already low. At 11.3 times owner cash and near the bottom of the 52-week range, the market is not paying for a recovery. A stop in the decline is worth something.
• Financing is in place. A $2.7 billion undrawn facility to 2029, covenant compliance and the early 2027 maturities already replaced.

Main risks

• Volume erosion. Adjusted operating profit fell 16% in fiscal 2026, 11% in the first quarter, and is guided down a further 8% to 13%. Guidance has organic sales roughly flat, so the profit decline comes mostly from costs and reinvestment, not from lost sales.
• Debt of $13.6 billion. Net debt is 3.9 times operating profit plus depreciation and, on guidance, rises to about 4.3 times. Refinancing at higher coupons raises interest of $538.6 million further.
• Acquisition economics. $1.8 billion of goodwill and brand impairments in one year, including $1.5 billion on Pet, say something about what the company paid. The annual report flags a risk of decreasing coverage on the Blue Buffalo and Progresso brands, and the carrying values were not available to me.
• Dividend coverage. The dividend takes 81% to 85% of cash flow, so there is little room for a bad year. A cut would hurt the shares even if it is the right decision for the balance sheet.
• Savings that do not reach profit. At least $750 million of savings has to offset inflation, price investment and volume. Management also flags a nine-point drag on operating profit from the 53rd week, incentive pay and divestitures.
• Valuation construction. All the value sits in a capitalized figure, and the multiple is a judgment. The Bear is not a floor.
• Currency. For an investor who does not hold dollars, the return also depends on the exchange rate.

Management execution

The record is mixed. Management has delivered the lower numbers it guided to, and the first quarter was consistent with the full-year outlook, which it reaffirmed. But adjusted earnings per share fell 16% in fiscal 2026 and 13% in the first quarter, and the acquisition record is weakened by the Pet impairment. I rate management credibility 10 of 20 (provisional, my estimate) and management and allocation 6 of 10 inside business quality. The buyback was cut from $1.2 billion in fiscal 2025 to $0.5 billion in fiscal 2026 and to none in the first quarter, which is the right call with leverage at 3.9 times.

Stock Analyza scorecard

The business quality score is judgment, not an audited figure. The company scores 4 of 12 on the standard company scorecard, which fits a food business with strong brands and a weak cash trend. The valuation score is 45 of 100, my estimate: 16 of 40 for the discount to value (the price is near the Base but far from the buy ceiling), 6 of 20 for protection against the Bear, 9 of 15 for agreement between methods (low weight because they share a driver), 5 of 10 for history, 6 of 10 for what the price implies and 3 of 5 for data quality. Business quality and valuation are separate numbers and are never blended.

Confidence score

The score is held down by what I could not close: the debt maturity ladder, lease liabilities, a verified multi-year history of owner cash per share, and a second method that is genuinely independent of the first. To reach grade B I would need those, plus two quarters showing cash stabilizing at or above $1.5 billion.

What I would watch from here

The thresholds are my monitoring triggers, not company guidance. The next review is the second quarter of fiscal 2027.

The thesis breaker

Continuing cash decline on top of the debt. If owner cash keeps falling at the pace of the last four years while the dividend stays near $1.3 billion, either the dividend is cut or deleveraging stops, and in both cases the Base multiple of 11.8 times is too high.

Stress test and independent check

The analysis I was given described itself as a research draft with an unfinished audit. I rebuilt it before using any of it. I reproduced owner cash of $1.5469 billion ($2,166.2 million less $539.9 million less $79.4 million), $2.876 a share, the three primary values ($23.01, $33.93, $44.00), the operating profit plus depreciation base of $3.3667 billion, the secondary value (EV of $31.65 billion, equity of $18.46 billion, $34.33), the 1.2% divergence, the $20.40 buy price, the nine-cell grid and the 4.0% to 4.4% upside range.

From company sources I confirmed the fiscal 2026 results (sales, operating profit, adjusted operating profit and earnings per share, cash flow, capital spending, dividends, buybacks, stock compensation, depreciation, impairments, debt and cash), the first quarter fiscal 2027 results, the balance sheet and cash flow lines from the quarterly filing, the credit facility and the fiscal 2027 outlook. The price comes from stockanalysis.com: a close of $32.57 on October 21 and $32.60 after hours.

I made six changes to the draft. The reference price is the regular session close of $32.57 instead of the broker ask of $32.66, which moves the Base upside from 4.1% to 4.4%. The share count is the reported 537.9 million rather than a rounded 538 million. I added the three cash flow levels and the debt discussion that the newest version of my method requires at this leverage. I added what each multiple assumes. I added the dated cash path as a check on the capitalized values, because the draft itself noted that a transition model was required if stabilization was more than a year away. And I added a valuation score of 45, which the draft did not contain. I left out all position sizing and trading material, since a stock research page does not carry personal sizing. The verdict did not change.

The choice between the two methods does not change the decision, since both give a buy price near $20 and the price is $32.57. A conditional verdict is therefore not needed, though the range of defensible values for the Base is wider than the point value suggests.

The strongest case against Hold is that the shares are already priced for a stop in the decline, a 7.5% yield pays you while you wait, and a stabilization would lift the shares to the Bull value of $44. That is a fair case. The strongest case against owning at all is that shrinking brands and $13.6 billion of debt make even 8 times cash generous, and the dividend funds itself only by not deleveraging. I do not dismiss either, which is why the verdict is Hold and the margin of safety is 40%.

Data limits: no debt maturity ladder, lease liabilities, pension funded status or itemized divestiture proceeds from the filings I could read; fiscal 2022 to 2024 cash flow taken from the supplied analysis and not re-verified; no peer set or synchronized analyst estimates; the full dollar impact of the 53rd week was not separated. These are input uncertainty, not construction errors, and they are why Confidence is C.

Final verdict: hold at $32.57

The yield is the story, and it is the part to be careful with. A 7.5% payout from a business whose cash fell 29.7% last year is a price, not a gift. The shares are worth about what they cost today, and the safety you would want is another 40% lower.

Verdict: HOLD at $32.57, watchlist for new money. Triangulated Base value $34, upside 4.4%, Confidence C (14/25), required margin of safety 40%, maximum buy price $20.40. Main thesis breaker: cash flow keeps falling while debt stays near $13.6 billion.

For holders this is a review, not an instruction to sell. For new money it means waiting either for a price near $20 or for two quarters that show cash stabilizing at or above $1.5 billion. The dividend isn't the thesis. The next two cash flow statements are.

Sources

Primary sources include General Mills' first quarter fiscal 2027 results and quarterly filing, the fiscal 2026 results exhibit and annual report, and the fiscal 2025 annual report for the comparison year.

The reference market price is $32.57, the October 21, 2026 regular session close, shown on the stockanalysis.com quote page, which also showed the $32.60 after-hours price, the 52-week range and the dividend. Per share values use 537.9 million diluted shares, the first quarter weighted average. The multiples, the 10% discount rate, the dated cash path, the fiscal 2027 cash estimate and the leverage estimates are analytical assumptions, not company guidance.