Flowers Foods: $306 million of trailing free cash flow, and a dividend cut by 49.5%

Over the 53 weeks to July 18, 2026 Flowers Foods generated $421 million of operating cash flow and, after $115 million of capital spending, $306 million of free cash flow. That is 23% of its $1.30 billion market value. In May the same company cut its quarterly dividend from $0.2475 to $0.125, a cut of 49.5%. A screener will show you the first number. The dividend cut is a sharper clue to how the board reads the next twelve months, and that is the number I start from.

At the October 16, 2026 regular session close of $6.11 the shares trade 18.2% below my triangulated Base value of $7.47, upside of 22.2%. The Bear value of $3.71 is 39.3% below the price and the Bull value of $10.43 is 70.8% above it. Confidence is grade C (14/25), the required margin of safety (MOS) is 40%, and the maximum buy price is $4.48. The latest company report is the second quarter release of August 20, 2026. The third quarter report and the October refinancing are not in these numbers.

Valuation route: branded packaged bakery with a dividend that has just been reset. Primary method: economic EBITDA at a 7 times exit multiple in October 2028, less debt and other claims, plus the dividends paid on the way, all discounted at 10%. Secondary method: owner cash after interest, taxes, capital spending and working capital, capitalized at 10% with no growth. Classification: standard company, 2 of 12 on the generic scorecard. Values are present fair value estimates, not 12 month price targets.

What Flowers Foods is

Flowers Foods bakes and sells Nature's Own, Dave's Killer Bread, Canyon Bakehouse, Simple Mills, Wonder and Tastykake. It guides to fiscal 2026 net sales of $5.07 billion to $5.14 billion, 2.2% to 3.5% below fiscal 2025. Much of the bread reaches stores through independent distributors who buy their territories, which is why the cash flow statement carries route-system lines that most food companies do not have. These are repeat-purchase brands. Repeat purchase does not protect a category that is shrinking.

In 2025 the company bought Simple Mills. The announced price was $795 million, and the annual report records consideration above that. The leverage that followed and the dividend reset in May are linked in time. I cannot show that the one caused the other, and I do not need to: either way the company has moved from a defensive dividend grower to a leveraged repair job, and I value it as one.

The latest quarter

Price and mix added 1.8% (3.8% on branded retail) and did not come close to covering the volume loss. Sales fell 4.0% and adjusted EBITDA fell 19.2%, a gap that points to fixed bakery and delivery costs spread over fewer loaves (my reading, the release does not split it). The guidance cut on August 20 follows the same line.

The revised guidance also gives depreciation and amortization of $165 to $170 million, net interest of $65 to $70 million, a 26% tax rate, 213.5 million diluted shares and capital spending of $115 to $125 million. I use those as calibration, not as a forecast of the two anniversary periods I value.

Adjusted EBITDA is the number management guides to, and adjustments have run through the annual reports since at least 2023. I do not assume they are finished. My Base takes $15 million a year off the $467 million midpoint for legal, restructuring and transition costs (my estimate), which gives economic EBITDA of $452 million.

Which cash to value

Owner cash is what is left for shareholders after the spending the business needs: cash from operations less capital spending, less stock compensation, less the net cost of the route system (territory purchases plus notes issued, less principal collected). The 2023 to 2025 figures come from the annual report and a market data aggregator. Flowers does not add stock compensation back to its adjusted EBITDA, so in the forward model I deduct nothing a second time. The historical bridge deducts it once, because cash from operations adds it back.

Three-year average headline free cash flow is $273.3 million and average owner cash is $220.1 million. I accept neither as the earning power of a business with falling volumes. The trailing $306.1 million spans 53 weeks, because fiscal 2025 had an extra week, and it carries tax, legal and working-capital timing that I cannot strip out. A clean trailing owner cash yield would need a full rolling adjustment for stock pay, route investing and tax, so I do not quote one. The first draft of this analysis had the trailing figure at $306.12 million. The filings give $306.06 million.

Debt, claims and shares

I use face value, not the $1,686.2 million carrying value on the balance sheet, because a bond discount is not debt forgiveness. Current maturities are $399.9 million of that carrying value. I verified the post-employment line on the balance sheet. The deferred pay and seller payable are my count from the 10-Q, about $0.14 a share before discounting, and I did not re-read them against the filing text. The $317.1 million of operating lease liabilities is not subtracted, because EBITDA and cash flow are already after rent. Distributor receivables are operating assets, and deferred taxes of $270.4 million are not treated as payable now.

The model uses 213.5 million diluted shares, the company's guided figure, against about 212.1 million shares outstanding. Diluted weighted averages were 213.4, 212.1 and 212.1 million in 2023 to 2025, so dilution has been small, and the August shelf filing shows no preferred stock outstanding. At the $6.11 close the diluted market value is $1,304.5 million and conventional enterprise value is $2,951.7 million, 6.32 times the $467 million EBITDA midpoint. On my own claims and economic EBITDA the multiple is 6.6 times (my arithmetic).

The balance sheet is dated July 18. In April the company signed a $400 million delayed-draw term loan to repay the 3.5% senior notes due in October. It was a single draw, available through October 1, 2026, with a three-year term from funding, priced at SOFR plus 0.875% to 2.000%. The agreement caps leverage at 3.75 times, loosened to 4.00 times through the quarter ending October 9, 2027, and requires interest coverage of at least 4.5 times. I have not found a filing that confirms the draw, the cash used or the rate, so I treat the new loan as replacing the old notes and add nothing to net claims. My leverage proxy, face debt less cash over economic EBITDA, is 3.6 times, close enough to the cap to matter. The contract defines EBITDA its own way and I have not read that definition. A replacement cost of 6% against 3.5% would add about $10 million of pretax interest a year, which is why my normalized interest is $75 million against guidance of $65 to $70 million.

Primary valuation: economic EBITDA with a two year transition

The business is meant to be worth more in October 2028, once volumes and margins have either settled or not. Valuing 2028 earnings with a multiple and calling it today's value would be a trick, so the model runs two forward years. I add the dividends shareholders receive, subtract the claims that remain after retained cash has paid down debt, and discount everything at 10% (tested at 8% and 12%). Cash is counted once.

Owner cash is 41.5% of economic EBITDA, or 40.2% of the guided midpoint. Depreciation of $167.5 million appears only in the tax line, since capital spending is already deducted. Base sales are roughly flat at $5,106 million with an 8.85% economic EBITDA margin. Nothing here assumes a return to the premium multiple of earlier years.

Net claims at exit are 1,685.2 less 154.0 of debt reduction, or 1,531.2. Exit enterprise value is 452 x 7 = 3,164. That leaves $1,632.8 million of equity, $7.65 a share. Add the two $0.50 dividends and discount: $7.19 today.

Why 7 times? It is a judgment inside the range for a stabilizing, low-growth food company, not a measured peer median, because I have no synchronized peer table. The Bear uses 6 times for permanent contraction and the Bull 8 times for stabilization. The price is 6.3 to 6.6 times depending on whose EBITDA and whose claims you use, so even my Base asks for a small re-rating.

Sensitivity: primary value per share by economic EBITDA and exit multiple

Year 1 cash is held at $180 million and year 2 cash and debt are recomputed for each EBITDA. At 7 times the $6.11 price sits between the $410 million and $452 million rows, which is the reverse calculation further down.

Secondary valuation: capitalized owner cash

The second method ignores multiples. It takes year 2 owner cash of $187.53 million, assumes all of it is distributable with zero growth and the debt stock refinanced as it is, and capitalizes it at 10%: $1,875.3 million, less $38.0 million of other claims, or $1,837.3 million of equity. That is $8.61 a share at exit. Add the two dividends and discount: $7.98 today. The 0% growth is deliberate. Renewal capital spending is already fully deducted, and I do not capitalize any saving from growth investment that growth would need.

• Shared by both methods: the sales and margin normalization, the financing, the share count, the interim dividends and the time value.
• Different in each: the first rests on an exit multiple and the debt bridge, the second on renewal capital spending, cash taxes, working capital and the capitalization rate.
• Treated the same: both are after operating rent, neither subtracts lease debt again, and neither credits any minority or joint venture cash.
• What agreement proves: that the multiple-based value can be supported by cash after reinvestment. It does not prove the recovery happens, because more than half of both values depends on the same operating stabilization, so agreement earns no extra confidence.

History and growth diagnostics

A data aggregator shows year-end EV/EBITDA of about 12.4 times in 2023, 10.9 times in 2024 and 8.5 times in 2025. The definitions and lease treatment are not reconciled to mine, so these numbers get zero weight in the valuation. They show how far the multiple has already fallen, not where it should stop.

The growth-adjusted check does not work here. Third-party consensus adjusted EPS is about $0.77 for 2026 and $0.68 for 2027, a fall of roughly 11.7%, and the 2026 adjusted P/E is 7.9 times. A PEG ratio on negative growth is not meaningful, so I give no credit for growth. Consensus does not enter my Base.

For free cash flow growth I compare 2023 owner cash of $192.96 million ($0.904 a share) with the 2028 run rate in each scenario, five fiscal intervals later. The endpoints mix a historical proxy and a forecast, so this is not a verified five-year history.

Owner cash grew 16.1% a year from 2023 to 2025, but the acquisition and changing route-system outflows make that short comparison unreliable, and the larger rise in headline free cash flow is not evidence of structural compounding. The Base ends slightly below where 2023 started. Trailing headline free cash flow gives a yield of 23.5% at the current price, against the Base owner cash yield of 14.4%, a forward model figure and not a reported one. Consensus free cash flow for 2026 is about $220.8 million, 16.9% of the market value. It is before my stock pay and route adjustments, and subtracting the 2025 stock pay of $32.3 million gives $188.5 million, near my Base but from a different year and definition, so I do not count it as confirmation.

Cash flow growth quality is uncertain and normalizing, and the reliability of the estimates is low: I could not find the dispersion of cash flow estimates or their revision history. If year 2 owner cash is 20% lower, the triangulated value falls to $6.86, and if it is 40% lower, to $6.26, with forward yields of 11.50% and 8.63%. Those stress tests hold EBITDA and the multiple constant to isolate conversion, and they do not replace the Bear case.

Valuation triangulation

I weight the first method 65% because the debt bridge is what binds a leveraged company, and the second 35% because renewal cash limits what shareholders can really receive. The weights are judgment, not probabilities. Equal weights give $7.58. In Base the methods are 10.4% apart, which is comfortable, but the Bear methods are 35.6% apart: $3.22 if the operating business has little residual value, $4.62 if the cash survives. Read the downside as roughly $3.22 to $4.62, not as a floor at $3.71.

What is the market already pricing in?

Solving the primary model backward at $6.11, with Base dividends and a 7 times multiple, the price needs about $416 million of economic EBITDA, 8.0% below my Base of $452 million. At Base sales that is a margin of 8.15% against 8.85%. Holding my Base cash, debt and EBITDA instead, the price implies an exit multiple of 6.38 times against my 7. In the cash method the price is justified by about $139 million of owner cash a year in perpetuity, against $187.5 million in Base.

These are three separate answers, not one set of market expectations. The point is that the price does not require a collapse: a modest EBITDA disappointment plus a restrained multiple is enough to explain it. The price is not asking for bankruptcy. It is asking you to believe volumes and cash conversion stabilize above these thresholds. Buying at $6.11 and receiving the Base outcome would earn about 22.0% a year over two years before tax and fees, a modeled figure for information only and not a forecast.

Bear, base and bull scenarios

Bear: volumes keep falling and price and mix stop covering them, sales settle at $4.8 billion, economic EBITDA at $380 million, the dividend is cut again to $0.25 and the market pays 6 times. Cutting both earnings and the multiple is justified only because this is structural decline, not a trough. Base: guidance holds, sales are flat and margins stabilize at 8.85%. Bull: stabilization and execution lift sales to $5.3 billion and EBITDA to $505 million at 8 times, which is growth in sales and not a return to the old premium rating.

No equity raise is assumed in any scenario. An actual solvency shock could take the equity below the Bear value, including to a near-total loss, and I would rather say that plainly than draw a prettier Bear.

How the buy price is set

A Confidence score of 14 is grade C, which carries a base margin of safety of 40%. I add no modifier, because the operating, leverage and accounting risks are already priced in other places, listed below. The required MOS is 40% and the maximum buy price is $7.47 times 0.60, or $4.48. That is 26.7% below the current price and 20.8% above the Bear value. The price is above the top of my entry range of about $3.70 to $5.20, which comes from changing EBITDA or the multiple one at a time.

Four quantities are easy to confuse, so I keep them apart. Fair value today is $7.47. The modeled return at $6.11 is 22.0% a year. An entry price for an illustrative 18% return, using the same Base dividends and exit, would be $6.52. That 18% is an example, not a universal hurdle. And the margin of safety against model error gives $4.48. For a conservative decision the MOS threshold controls. There is no unconditional buy price if debt, the dividend or operating facts get worse. Where each risk is priced:

• Category decline and operating leverage: in the EBITDA and margin of each scenario and in the restrained exit multiples. No separate MOS add-on.
• Financial leverage: in the face-value debt bridge and the allocation of retained cash. Debt already reduces equity directly.
• Refinancing: in the $75 million interest allowance and the $65 to $90 million sensitivity. No second haircut.
• Legal, restructuring and product quality costs: in the $15 million reserve and the Bear cash. The tail goes into Confidence.
• Maintenance capital spending: in the $115 to $150 million sensitivity. I do not add back an unsupported growth share.
• Terminal dependence and the shared recovery driver: in Confidence. Agreement between methods earns no extra credit.
• Balance sheet after July: in Confidence and sensitivity. I do not invent a current balance sheet.

Dividend

The May reset took the quarterly payment from $0.2475 to $0.125 and the latest declaration kept $0.125. Annualize the new rate, not the old one: a screener that still shows the old dividend will tell you about a 16% yield, and that is a figure from before the cut. The new payment is more plausibly funded than the old one, but it is not protected by contract, and Flowers is now an income stock that has just cut, not a dividend grower. A simple dividend model that pays $0.50 forever at 10% gives $5.00, which I show as a diagnostic and do not average into fair value, since it ignores the cash retained to repay debt. Dividends paid to non-U.S. holders face U.S. withholding tax, and the rate depends on your tax treaty and your broker.

Reasons to own Flowers Foods

• The brands are real and bought repeatedly. Stabilization does not need a new business model, and distribution reach is still an asset.
• Cash can fund the repair. Base owner cash covers the reset dividend and leaves about $154 million of debt reduction over two years.
• A modest valuation is enough. The Base assumes 7 times EBITDA, not a return to a premium staples rating.
• Capital allocation has a repair route. The dividend reset and the refinancing facility reduce near-term financing pressure, if management follows through.

These justify watching the stock and holding a conditional thesis. They do not show that the price already gives the margin of safety I require.

What could go wrong?

• Thesis breaker: volumes keep falling. Unit volume was down 5.8% and branded retail volume down 7.6% in the second quarter. Persistent lost units can swamp price and mix and cost savings and produce the Bear cash outcome.
• Balance sheet and financing. Debt makes the equity nonlinear. A covenant problem or an unfavorable refinancing could impair the dividend and force asset sales or new shares. The leverage cap is 3.75 times, loosened to 4.00 times until October 2027.
• Product quality. On October 2 the company recalled Nature's Own Hawaiian Bread over undeclared milk. I have no verified cost and I will not invent one.
• Spare parts. The 10-Q carries $102.2 million of spare parts and supplies under a recoverability review that the company expects to finish by the end of fiscal 2026 and says it cannot yet size. That is a balance, not a known loss.
• Capital allocation and accounting. Recurring legal and restructuring costs, the execution of the Simple Mills deal and the spare parts review can all undermine adjusted earnings.
• Stale balance sheet. My net claims are from July 18. The October draw, the cash used and any change in debt are not verified.

Management execution

The evidence is mixed. In 2023 the annual report carried substantial legal and operating adjustments. The 2025 acquisition and impairments weaken the case that more investment automatically adds value per share. In 2026 the guidance cuts and the dividend reset cannot be squared with a simple defensive-growth story. Against that, arranging the refinancing early and keeping cash for debt reduction are the right corrective steps, and the credit is conditional on execution. I could not build a complete series of initial guidance against actual results, so this is a limited track record. Assessment: 8/20, recent underdelivery, a credible repair mechanism that is not yet proven.

Stock Analyza scorecard

The 2 of 12 puts Flowers firmly in the standard class. Return on invested capital is not shown as a verified five-year average, and I give no point where return quality is unreliable. The character change, from defensive dividend grower to leveraged stabilization, is negative: it lowers the multiple I am willing to pay and blocks a mechanical Buy based on yield alone.

Business quality is 45/100. The brands and distribution are the strongest part (12/20). The balance sheet earns only 4/15 because of the leverage and refinancing exposure, and fixed manufacturing and delivery costs would magnify any further volume loss.

The valuation score is 50/100. A stock can look undervalued and still have little investable cushion: the Bear value is 39% below the price, so protection earns 0 of 20.

Confidence score

Critical forward inputs are my estimates: economic EBITDA, capital spending, interest and the working capital line. That is why Confidence is C and not higher, however well the two methods agree.

What I would watch from here

Today's readings: unit volume of -5.8% and branded volume of -7.6% are red today on the levels, though my rule for total volume wants two more reports before it triggers a re-underwrite. Net debt to economic EBITDA of about 3.6 times is yellow. Base owner cash and coverage are model values, not reports. The next review point is the third quarter release in the normal November window. I have not confirmed the date with the company, and I would re-underwrite after any guidance change, another dividend reset, major M&A, a large issuance, a credit event or a leadership change.

The three most important thesis breakers

1. Volumes keep falling for several more quarters, so that economic EBITDA and owner cash cannot stabilize.
2. Owner cash settles below about $139 million, the level the price already assumes in the cash method.
3. Debt or covenant pressure prevents ordinary shareholder distributions.

Stress test and independent check

I recomputed the model in a separate script from the raw inputs: the owner cash and claims bridges, both valuation methods, the triangulation, the three scenarios and their returns, the reverse solutions, the sensitivity tables, the history ratios, the scores and the margin of safety. They matched to the cent, apart from the trailing free cash flow slip noted above. I checked the guidance, the second quarter results, cash, debt, the lease and spare parts balances, the credit agreement and the price against the company's release, the 10-Q, the 8-K and a market data aggregator. The price is the October 16 close, with the later after-hours quote excluded.

Changing one input at a time moves the value between $6.23 and $8.60, always above the $6.11 price, and the maximum buy price between $3.74 and $5.16, always below it. Even the generous pair gives a maximum buy price of $5.89, still under the price. That is why the answer is Hold with a conditional watchlist, and why I give a range of entry prices and not a precise one. A $100 million error in unobserved net claims is worth about $0.39 a share in the first method before weighting.

The strongest case against my Hold: the trailing cash is high and a halved dividend could speed up the repair, so waiting can mean missing a re-rating, and the Bull value is $10.43. I take that seriously. My answer is that repeating trailing cash means explaining away the tax timing, capital spending timing, stock pay, legal payments and permanent volume loss, and a low price to free cash flow on a screen does not do that. The strongest case against my Base: consensus EPS falls in 2027, which suggests margins will not stabilize where I put them. Then the market is roughly fair at $416 million of EBITDA or a 6.38 times multiple, and neither of those is catastrophic. The Bear case shows what a miss costs.

Main data limits: the balance sheet is from July, the October loan draw is not verified, I have no maintenance versus growth split for capital spending, no synchronized peer multiples and no dispersion of cash flow estimates. The economic EBITDA, the $15 million reserve, capital spending, working capital, interest of $75 million, the 7 times multiple, the two year cash path and the deferred pay count are my estimates. The 6.6 times current multiple and the leverage proxy of 3.6 times are my arithmetic.

Final verdict: hold at $6.11

The cut dividend and the cash-generating brands make a repair case, and the price does not demand a miracle. It also does not give the cushion that weak volumes and a leveraged equity require. The negative character change keeps the Undervalued label from turning into a Buy.

Verdict: HOLD at $6.11, watchlist for new money. Triangulated Base value $7.47, upside 22.2%, Confidence C (14/25), required margin of safety 40%, maximum buy price $4.48, entry range about $3.70 to $5.20.

The $306 million is a number about a year that has ended. In my cash method the price only needs about $139 million a year, and my Base says $187.5 million. The next quarter's volumes will say which of those the following year looks like.

Sources

Primary sources include Flowers Foods' second quarter 2026 results release of August 20, 2026, the 10-Q for the period ended July 18, 2026 and its balance sheet, the 2025 annual report and 2023 annual report, the dividend declaration of August 28, 2026, the February 2026 dividend declaration, the filing that reset the dividend in May, the April 2026 term loan agreement and the August 2026 shelf registration. The product recall is from the company's press release of October 2. Price history, consensus EPS and cash flow estimates, year-end multiples and the multi-year cash flow lines come from a third-party market data aggregator (price history, forecast, ratios, cash flow). Broker price targets and ratings are not used.

The reference market price is $6.11, the regular session close on October 16, 2026. Values per share use 213.5 million diluted shares. Net claims are $1,700.0 million of senior notes at face value, less cash of $52.8 million, plus $38.0 million of other claims. The earnings path, exit multiples, scenario values and robustness tests are analytical estimates, not company guidance.