Capstone: Hershey's 2025 accounts, from performance to credit

All from the consolidated statements (Item 8) unless marked; millions of dollars.

From the MD&A:

• Net sales rose 4.4%: price realisation of about 6% and about 1% from acquisitions (Sour Strips in 2024, LesserEvil in 2025), offset by a volume fall of about 1% "primarily driven by price elasticity impacts".
• Cost of sales rose by 736.6 million of higher commodity costs, 287.2 million of higher supply chain costs including tariffs, and 491.0 million of unfavourable mark-to-market activity on commodity derivatives "intended to economically hedge future years' commodity purchases". These are changes from 2024, not 2025 levels.
• The segment note shows the levels: unallocated mark-to-market losses on commodity derivatives of 423.2 million in 2025, against gains of 460.4 million in 2024, which the company excludes from segment income.
• The effective tax rate rose from 10.2% to 27.3%: 2025 "primarily impacted by state taxes and tax reserves"; 2024 "benefited from investment tax credits".
• Cocoa products "continued to experience elevated market prices compared to historical levels" (MD&A; Item 1A, Risk Factors, calls cocoa "structurally elevated").
• U.S. retail takeaway rose 5.4%; candy, mint and gum share fell about 10 basis points; salty snacks share rose about 40 basis points.

• Margins: gross 33.5% (47.3%), operating 12.3% (25.9%).
• DuPont: net margin 7.6%; asset turnover 0.88 (11,692.6 ÷ average assets 13,344.1); equity multiplier 2.85 (13,344.1 ÷ average equity 4,675.7); ROE 18.9%, against about 50% in 2024 (2,221.2 ÷ 4,406.9). Almost all the fall is margin; turnover and leverage barely moved.
• Growth: price, not volume. Sales grew on price while volume fell: pricing power is holding, but at a cost in volume.

• Operating cash flow (2,277) was 2.6 times net income (883); accruals (cash flow method) are −10.4% of average assets: cash ran well ahead of profit.
• Why: the cash flow statement adds back 530 million of unrealised losses on derivative contracts, but these mostly reverse gains booked as profit in 2024 on contracts that settled in 2025 (derivative assets fell from 524 million to 4 million). So the add-back reverses non-cash gains reported in 2024 (the 10-K says cash moves with commodity brokers daily, so it doesn't say exactly when the cash arrived); depreciation of 503.7 million adds the rest.
• Free cash flow 1,822 million (2,277 − 455) covered dividends of 1,085 million; the 756 million LesserEvil acquisition and the year's payouts together took slightly more than free cash flow, and debt rose by 0.3 billion.

• Most of the margin fall is the swing in mark-to-market, not 2025 alone. Hershey's commodity derivatives aren't designated as accounting hedges, so their changes in value go straight into cost of sales. Taking out the unallocated mark-to-market in each year (the segment note's figures), gross margin was about 43.2% in 2024 ((5,300.9 − 460.4) ÷ 11,202.3) and 37.2% in 2025 ((3,922.7 + 423.2) ÷ 11,692.6). Of the 13.8-point fall, about 7.7 points is the swing from 2024's 460 million gain to 2025's 423 million loss; the remaining 6 points is the real rise in commodity and tariff costs and lower volume.
• 2024 was flattered, 2025 depressed, by the same mechanism. In 2024 the derivatives gained value, consistent with rising cocoa prices, before the cocoa was bought; in 2025 those contracts settled and the hedged cocoa reached cost of sales at high prices. It's a shift between years, not a lasting loss and not a reserve that will come back: with derivative assets now about 4 million, little of the 2024 gains is left to unwind; but new positions (open contracts with a notional value of 973 million at the year end) can create new swings either way.
• Working capital: receivable days fell (about 23 against 26) and so did inventory days (about 67 against 78, partly because cost of sales rose), so no sign of sales pulled forward.
• Tax: the effective rate rose from 10.2% to 27.3% (state taxes and tax reserves in 2025; investment tax credits in 2024). At 2024's rate, 2025 net income would have been about 1,090 million: the tax change cost about 207 million.

EBITDA = operating profit + depreciation and amortisation; without mark-to-market, 2024 less 460.4 and 2025 plus 423.2. Interest is net interest expense (190.2 and 165.7); on gross interest expense of 224.8, 2025 cover is 6.4×. Operating leases are 336 million; the 74 million of finance leases are already inside debt. Leverage has risen on both bases: by about half a turn on the underlying measure (1.8× to 2.3×) and by about 1.3 turns as reported.

Hershey: a real cost shock, exaggerated by mark-to-market on economic hedges; profits fell, cash didn't · Source: Form 10-K for 2025

Summary. Operating profit halved (1.44 billion against 2.90 billion) and net income fell 60%, but operating cash flow fell only 10%, to 2.3 billion. More than half of the 13.8-point fall in gross margin is the swing in mark-to-market on cocoa derivatives, from a 460 million gain in 2024 to a 423 million loss in 2025; without it, gross margin fell from about 43% to about 37%, still a real cost shock from cocoa, tariffs and lower volume. Pricing lifted sales 4.4%, at the cost of about 1% of volume.

Performance. ROE fell from about 50% to 19%, almost all through margin; turnover and leverage barely moved. Salty snacks grew share while candy share slipped, and the LesserEvil purchase adds to salty snacks.

Cash and quality. Cash was 2.6 times net income, partly because the add-back for derivatives reverses gains already reported in 2024; free cash flow (1.8 billion) covered the dividend; no sign in working capital of pulled-forward sales. A higher tax rate took another 0.2 billion off net income.

Credit. Debt is 2.8× EBITDA (2.3× without mark-to-market), up from 1.5× (1.8×); net interest is covered 7.6×. Comfortable, but leverage has risen with the acquisition and the profit fall.

What would change my view: gross margin staying below about 37% once the hedges reflect current cocoa prices; volume falling faster as prices rise; or debt rising further on acquisitions while earnings stay down.

1. Profit halved in 2025, but cash fell only 10%. Exhibit: operating profit, net income and operating cash flow, 2024 against 2025.
2. Sales grew on price, not volume. Exhibit: the bridge (+6% price, +1% acquisitions, −1% volume, small currency drag).
3. More than half the margin fall is the hedge swing; the rest is a real cost shock. Exhibit: gross margin 47.3% → 33.5% reported, 43.2% → 37.2% without unallocated mark-to-market.
4. Cash ran ahead of profit, partly because last year's derivative gains reversed. Exhibit: operating cash flow against net income; derivative assets 524 → 4; receivable and inventory days.
5. Leverage rose on any basis (underlying 1.8× to 2.3×); cover is still comfortable. Exhibit: debt ÷ EBITDA and interest cover, 2024 and 2025, reported and without mark-to-market.
6. Our view turns on cocoa after the hedges, volume and debt. Exhibit: the three triggers, with the levels that would change the view.

The three likely questions:

• "Isn't the hedge loss just accounting?" It's timing between years: 2024 booked gains on derivatives for cocoa bought later; 2025 used that cocoa at high prices while the gains, already counted, turned into cash. Strip mark-to-market out of both years to see the trend; little of the 2024 gains is left to reverse, though new positions can swing either way.
• "Can they keep raising prices?" Volume fell about 1% on price elasticity and candy share slipped 10 basis points: pricing is holding, but the cost in volume is showing.
• "Is the dividend safe?" Free cash flow of 1.8 billion covers 1.1 billion of dividends; the risk is debt-funded acquisitions while profit is down.

How it's built. The three habits of this level do the work. Check the source and the base: "491 million unfavourable" is measured against 2024, so the analyst finds the levels in the segment note before adjusting anything. Follow the cash: cash ran ahead of profit, and the derivative assets show why. Know what would change your mind: the triggers come from the weak points (cocoa cost, volume, debt). The memo combines the four modules into one story: a real cost shock, made to look worse by mark-to-market swings, with cash and credit intact. The deck's titles tell that story on their own.

What a reviewer would flag:

• Treating the MD&A's 491 million as 2025's loss, when it is the change from 2024.
• Calling this "hedge accounting": these derivatives aren't designated as hedges, which is why the swings hit profit.
• Treating the mark-to-market as a reserve that will reverse in 2026.
• Missing the tax-rate jump between operating profit and net income.

Performance: net sales fell 3.5% to 24.9 billion and gross margin slipped from 34.7% to 33.3%; the MD&A attributes a 715 million fall in operating income before impairments (about 5.4 billion to 4.6 billion) to cost inflation, volume and mix and other factors. Earnings quality: the 4.7 billion operating loss comes from 9.3 billion of non-cash goodwill and intangible impairments (5.6 billion more than 2024's 3.7 billion), which mean the carrying values are no longer supported by expected cash flows. Cash: operating cash flow rose to 4.5 billion; free cash flow of about 3.7 billion covered 2.3 billion of dividends and buybacks. Credit: debt of 21.2 billion is about 3.8× EBITDA before impairments, with interest expense (gross, 947 million) covered about 4.9×; the 10-K also reports the pause of work on its planned separation into two companies. What would change my view: further impairments with falling cash flow, or underlying operating income continuing to fall.