The Mosaic Company: A Long-Term Business Owner's Research Report

On the moat, my judgment is fairly restrained. MOS has no network effects, no obvious switching costs, and no data moat; the brand carries some weight in bulk fertilizer, but is far from enough to dictate price. Its real barriers lie mainly in: first, the upfront capital intensity of mineral rights, mines, and processing facilities; second, its asset footprint in North America and Brazil; third, the Canpotex export channel and its owned distribution network in Brazil; and fourth, partial scale and logistics advantages. The company believes that the vertical integration of its Brazilian business, mines and chemical plants located near customers, an owned distribution network, and brand give it an edge; in potash exports, Canpotex also provides international channels and port logistics capability. The problem is that these advantages show up more in "surviving and earning a lot in good years" than in "earning stable, high returns even in bad years."

There is another point often overlooked: the true industry leader holds a stronger cost position in potash. Mosaic disclosed an MOP cash production cost of about 84 dollars per tonne in the first quarter of 2026; Nutrien, in the same quarter, said its controllable potash cash production cost remained below 60 dollars per tonne, and that 49% of its potash ore-extraction tonnage was already automated in 2025. In other words, MOS's potash business is competitive, but it does not sit at the very leading edge of the cost curve. For a commodity business, this matters a great deal.

Scoring

If the stock market closed for five years starting tomorrow, would I be willing to hold MOS? My answer is: only when it is clearly undervalued and I define it as a "cyclical asset position"; if I treat it as ownership of a business that can compound with high certainty, I am not comfortable enough. This is the most fundamental difference between MOS and truly high-quality consumer, software, or exchange-type businesses. On business simplicity I give a high mark for understandability, but on industry and moat I will not award a high-quality score.

Management and Capital Allocation

On governance, the Proxy shows that in 2025 about 72% of the CEO's target total direct compensation was at-risk pay, the executive stock ownership guideline requires the CEO to hold 5 times base salary and other executives 3 times base salary, and the 2025 say-on-pay approval vote was about 93.5%. These arrangements suggest that Mosaic's institutional design is not bad, and that it is at least formally willing to tie compensation to long-term performance.

But if you look at "whether people and shareholders are truly deeply aligned," it is not as strong. The Proxy discloses that, as of April 2, 2026, CEO Bruce Bodine directly beneficially owned about 12,582 common shares, with options exercisable within 60 days of about 17,921 shares; all directors and executives combined held about 988,000 shares, still less than 1% of shares outstanding. This is not "no alignment," but it is certainly not the kind of strong alignment where management has most of its net worth staked in the stock. In other words, Mosaic's governance looks more like "professional-manager governance" than the strongly aligned "operator-owner-as-one" model. In addition, there was only one late executive Form 4 filing in 2025, which the company itself disclosed in the Proxy; while a minor blemish, it also shows that governance is not beyond reproach.

Capital allocation is where MOS warrants the most reservation. On the positive side, in 2024 it exchanged its 25% interest in the Ma'aden phosphate JV for about 111.01 million Ma'aden shares, a transaction valued at about 1.5 billion dollars, recognizing a net gain of 522 million dollars, and at the end of 2025 and in the first quarter of 2026 it recorded the fair value of that equity security at 1.848 billion and 1.964 billion dollars respectively. This step is not bad: it turned an off-balance-sheet JV into a more transparent, measurable financial asset, and added a "visible asset" to the balance sheet.

On the negative side, MOS's buyback history is not pretty. According to the company's statement of shareholders' equity, in 2022 it repurchased about 30.8 million shares for about 1.673 billion dollars, implying an average buyback cost of about 54 dollars per share; in 2023 it repurchased about 16.9 million shares for about 754 million dollars, implying an average price of about 44.6 dollars per share; and in 2024 it repurchased 7.94 million shares at an average of 29.63 dollars per share. Yet the current price is only about 23.89 dollars. This shows that Mosaic's 2022–2024 buybacks were less an example of outstanding "counter-cyclical capital allocation" than a case of using part of its cyclical profits in high-earnings years to buy back stock at the wrong price. For a value investor, this is a genuine demerit.

On the direction of cash use, the company has in recent years balanced among dividends, buybacks, debt repayment, and maintenance and environmental capital spending. In 2022 the company disclosed that it returned 1.9 billion dollars to shareholders that year while repaying 550 million dollars of long-term debt; in 2023 it spent about 1.108 billion dollars combined on buybacks and dividends; in 2024 about 506 million dollars; and in 2025 it halted buybacks and paid only about 280 million dollars in dividends. This shows that the company is not recklessly aggressive with spending, but it also reflects that its distribution policy is still swayed by the cycle and liquidity. For a commodity company, this is normal; for a company that wants long-term owners to regard it as a "master capital allocator," it rates only as adequate, not excellent.

Capital allocation score: 3/5. I consider management broadly honest, with disclosure that does not gloss over things and a willingness to be open about trouble, for example directly acknowledging volatile market conditions and the impact of production cuts and shutdowns in the first quarter of 2026, and lowering 2026 capital spending to 1.25 billion dollars. But if buyback timing, true shareholder alignment, and a "per-share intrinsic value growth orientation" are factored into the assessment, it remains clearly short of excellent.

Financial Quality and Owner Earnings

Let us first look at the core financial trajectory over the past five years. The data in the table is compiled and calculated from the company's 2021, 2023, 2024, and 2025 annual reports and the financial statements in its first-quarter 2026 10-Q.

This table makes two very important points clear. First, MOS's "good years" are very good, and its bad years can also get materially worse. 2022 was almost like a different company: net income, ROE, cash flow, and leverage were all excellent; by 2024–2025, operating cash flow and free cash flow fell rapidly, and in 2025 free cash flow turned negative. Second, recent profits have not become entirely fictitious; the issue is that cash profit is highly affected by working capital and capital spending. In 2025 equity net income was still 541 million dollars, but operating cash flow was only 825 million dollars, while capital spending reached as high as 1.359 billion dollars, leaving free cash flow at about -535 million dollars. This is not a typical accounting-fraud pattern; it is more typical of a cyclical, asset-heavy company: the income statement shows profit, but cash is swallowed up in inventory, payables, and capital projects.

Swings in working capital deserve particular attention. In 2025 the company's "results plus non-cash adjustments" before operating activities contributed about 1.8 billion dollars to cash flow, but changes in working capital were a drag of about 1.0 billion dollars, mainly from a 762 million dollar increase in inventory and a 360 million dollar decrease in accounts payable and accrued liabilities; the company itself explains that this is related to slower demand in the fourth quarter of 2025, higher phosphate and Brazilian inventory, and higher raw-material prices. In the first quarter of 2026, operating cash flow improved slightly from the prior-year period to 104 million dollars, but free cash flow was still -253 million dollars. This means MOS is not the kind of company where "profit naturally settles into cash," but the kind whose cash flow is frequently swallowed by the inventory cycle and capital spending.

The balance sheet is not fragile, but it is certainly not relaxed either. At the end of March 2026, the company had 282 million dollars in cash, 5.523 billion dollars of interest-bearing debt, 1.964 billion dollars of equity securities and investments, and common equity of about 11.804 billion dollars. If the Ma'aden shares are all treated as a realizable asset, net debt pressure is manageable; if they are not equated with cash, leverage is not low. More importantly, the company recorded 2.602 billion dollars of asset retirement obligations (ARO), of which the present value of ARO related to closure of North American phosphogypsum stacks is about 1.5 billion dollars, and in 2026 alone land reclamation, gypsum-stack closure, and water-treatment spending is expected to be about 260 million dollars, plus about 750 million dollars of environmental capital spending. For long-term shareholders, these are not "will-they-happen" risks but real cash obligations that are already occurring and must continue to be paid.

On accounting quality, I have not seen obvious signs of financial fraud. KPMG issued unqualified opinions on the financial statements and internal controls in both 2021 and 2025. What truly warrants caution is not revenue recognition but highly judgmental estimates: for example, water-treatment-related ARO, Brazilian deferred taxes and valuation allowances, and fluctuations in the fair value of the Ma'aden stake. In 2025 the company also recognized a tax-specific-period net expense of about 212 million dollars related to the Brazilian valuation allowance; in the first quarter of 2026 profit was again boosted by an unrealized fair-value gain of about 112 million dollars on the Ma'aden shares. In other words, MOS's profits are not "unreal," but the profit figure contains no small amount of valuation and non-cash fluctuation, and a long-term analysis should place more weight on operating cash flow and owner earnings.

Now to owner earnings. Take 2025 as an example: equity net income of about 541 million dollars plus depreciation and amortization of about 1.050 billion dollars totals about 1.59 billion dollars; if you directly deduct total capital spending of 1.359 billion dollars and then account for the actual working-capital drain, the company's distributable cash under its reported framework is very poor and free cash flow is negative. The catch is that not all of total capital spending is growth investment, and Mosaic's 2026 guidance is for full-year capital spending of about 1.25 billion dollars and depreciation and amortization of about 1.1 to 1.2 billion dollars, which shows that the vast majority of capital spending is essentially maintenance, environmental-compliance, or necessary capital spending, rather than easily reducible "growth investment." Given that the company's pre-working-capital-change cash contribution was roughly 2.0 billion, 1.3 billion, and 1.8 billion dollars in 2023–2025, I would rather peg MOS's conservative "mid-cycle owner earnings" in the 500 million to 800 million dollar range, with a core midpoint of about 600 million to 700 million dollars. This is an inference, not a fact.

On this basis: MOS can generate real cash flow over the long run, but it is neither stable nor "easily distributable." It is not a business model where "the more it grows, the more it earns"; on the contrary, it frequently needs to reinvest large amounts of capital, handle environmental obligations, and absorb working-capital swings. Therefore, from an owner-earnings perspective, MOS is closer to a resource company that "can generate excess cash in good cycles and swallow cash in bad cycles" than to a high-quality business that "continuously produces free cash flow." Against the current equity market capitalization of about 7.59 billion dollars, a conservative owner-earnings midpoint of 600 million dollars implies a share price of about 12.7 times owner earnings; at 700 million dollars, about 10.8 times. This multiple is not outlandish, but it is also far from "extremely cheap."

Valuation and Margin of Safety

First, the current "surface valuation." Based on 2025 equity net income of 541 million dollars, the current share price of 23.89 dollars, and about 317.8 million shares outstanding, MOS currently corresponds roughly to 14 times trailing-twelve-month equity net income. Against common equity of 11.80 billion dollars at the end of March 2026, the current P/B is about 0.64x; after deducting 989 million dollars of goodwill, the approximate P/TBV is about 0.70x. Based on 2025 operating profit of 822 million dollars plus depreciation and amortization of 1.050 billion dollars, 2025 EBITDA is about 1.871 billion dollars; using the market capitalization, debt, and cash at the end of March 2026 to compute "gross EV," EV/2025 EBITDA is about 6.9 times; if the 1.964 billion dollars of Ma'aden shares on the books are treated as a non-operating asset and deducted from EV, the adjusted EV/EBITDA is about 5.8 times. However, 2025 reported free cash flow was negative, so P/FCF is meaningless under the reported framework.

The meaning of this set of numbers is this: MOS looks inexpensive on book value and "last year's EBITDA," but the market is not giving it the valuation of a sustainable cash cow, because the market doubts the quality of its cash flow and the durability of its earnings. I think this doubt is justified. Especially in the first quarter of 2026, the company already posted an operating loss and a net loss, and withdrew full-year phosphate production guidance because of high sulfur prices and raw-material constraints; this makes investors even less willing to assign a high multiple.

Below are three valuation methods. Please note: Method One and future returns are both inferences, not facts.

Method One: Owner-Earnings Discount Method

I set the starting point for operating owner earnings at three tiers: conservative 450 million dollars, neutral 650 million dollars, and optimistic 850 million dollars. The discount rates used are 10% / 10% / 9.5% respectively; the long-term growth rates are 0% / 1% / 2% respectively. The reasoning is that MOS's long-term demand can grow, but the industry and moat do not support a high-growth assumption; at the same time, the depreciation, amortization, and capital-spending guidance management gave for 2026 shows that maintenance capital requirements are high, so I am unwilling to use the high-cash-growth model of an asset-light company. After completing the DCF for the operating business, I then count the Ma'aden book fair value into realizable value at about 85%, to reflect the five-year lock-up release, taxes, and discounting frictions. Ma'aden's book value was about 1.964 billion dollars at the end of March 2026; discounted at 85%, that is about 1.67 billion dollars, or about 5.2 dollars per share.

Method Two: Relative Valuation Method

Compared with peers, both MOS and Nutrien currently sit in a range that is "not very expensive, but not absurdly cheap either." Roughly calculated on current price and 2025 results, MOS is about 14 times trailing earnings; Nutrien's current share price is about 68.55 dollars, 2025 net income is about 2.297 billion dollars, and shares outstanding are about 482 million, corresponding to a market capitalization of about 33 billion dollars and a static P/E of about 14.4 times. In other words, MOS is not much cheaper than Nutrien on a static P/E basis. And in terms of business quality, Nutrien has larger potash scale, a stronger retail buffer, and a lower potash cash cost; therefore MOS deserves a discount rather than parity. For MOS, what is more telling is the 0.64x P/B and the 5.8 to 6.9 times EV/EBITDA: these multiples show that it is not expensive on assets and mid-cycle EBITDA, but that the market's discount is reasonable.

Method Three: Asset and Liquidation Value Method

MOS's common equity at the end of March 2026 was about 11.80 billion dollars, corresponding to a book value per share of about 37.1 dollars; after deducting goodwill, the approximate tangible book is about 34.0 dollars per share. On a static book basis, the current price of 23.89 dollars is clearly below book. The catch is that the "book value" of mines, processing facilities, and chemical assets does not equal liquidation value, and the company also carries 2.6 billion dollars of ARO, long-term environmental spending, financial guarantee obligations, and partly restricted Ma'aden shares. For this type of company, I will not treat 1x book as a value floor. The more realistic view is: the asset method gives you "backing from visible assets," but not a reliable liquidation cushion. I would put the "pessimistic asset value range" at 17 to 24 dollars per share, roughly 0.5 to 0.7 times approximate tangible book.

Combining the three methods, I give the following ranges:

So, on the margin of safety, the direct answer is: the current price is not without value, but for a "balanced-to-conservative, hold-for-more-than-ten-years" investor, the margin of safety is insufficient. The most fragile assumption is not "long-term demand disappears," but whether mid-cycle owner earnings can stably return to 600 to 700 million dollars. If that cannot be achieved, today's price is not cheap; if it can, today's price is merely "okay," not a "cigar-butt bargain."

Risk, Comparison, and Final Verdict

First, the most important risks. Competitive risk shows up not in brand but in the cost curve. Nutrien's potash cost is lower, global potash supply is recovering, and MOS may both have its margins squeezed by lower-cost players and be forced to cut production as supply recovers. Technological-substitution risk is low, because N/P/K still cannot be easily replaced, but product upgrades cannot fundamentally change its commodity nature. Regulatory and environmental risk is very high: ARO balances, state and federal financial assurance, gypsum stacks, and water treatment are all real money. Financial-leverage risk is currently manageable but rising; if core operating cash flow stays persistently weaker than capital spending, it will renew pressure on the balance sheet. Management risk is mainly not integrity but capital allocation, especially buyback timing. Customer and channel risk is mainly in Brazil, where in 2026 the company has already acknowledged a more challenging credit environment. Supply-chain and geopolitical risk shows up directly in sulfur, ammonia, energy, and ocean freight, and in 2026 it has already caused a clear operating shock.

The strongest opposing view is actually quite compelling: MOS looks cheap only because the market correctly judges it to be a "heavy-asset cyclical stock in a bad industry." The bears would say it has no true moat; that the high earnings of 2022 were merely a cyclical windfall; that the buyback history shows management spent money at the wrong price at the peak of the boom; that environmental liabilities and maintenance capital spending will erode free cash flow over the long run; and that the book discount seen today may, in the future, simply be an asset burden that must be continually backfilled. The operating loss, production cuts, and withdrawn guidance of the first quarter of 2026 provide the latest evidence for this bearish logic.

Which facts would overturn the investment judgment? I would watch five categories. First, if operating cash flow cannot sustainably recover in 2026–2027 to at least cover capital spending and dividends, it means the owner-earnings midpoint is even lower than I estimate. Second, if potash cost cannot move closer to lower-cost peers over the long run, or if potash profits are persistently squeezed, it means its core asset is not good enough. Third, if ARO, financial assurance, and environmental capital spending are revised up significantly again, the asset discount will be further "turned into liability." Fourth, if Brazilian receivables, bad debts, and the distribution environment deteriorate, the channel value of Mosaic Fertilizantes will be discounted. Fifth, if management again buys back large amounts at a high-boom, high-valuation stage while core FCF remains weak, the capital-allocation risk needs to be repriced.

Compared with other opportunities, my judgment leans conservative. Compared with the stronger Nutrien among peers, MOS is purer and more volatile, but also more fragile; Nutrien has larger potash scale, a retail network, and a lower-cost position. Compared with the S&P 500 index, MOS's current valuation looks low, but its single-commodity, single-industry, environmental-liability, and Brazilian risks are far higher than a broad-based index; SPY currently trades at about 594.28 dollars and represents multiple industries, multiple business models, and lower single-commodity exposure. For most long-term investors, MOS is not clearly superior to buying the index. Compared with the risk-free rate and high-grade bonds, MOS must offer a significantly higher return premium to be worth bearing these risks, and the current margin of safety is not enough for me to say with strong conviction that "it is clearly worth substituting for bonds and the index."

Investment Checklist

Information boundaries and open questions: this report gives priority to Mosaic's official 10-K, 10-Q, Proxy, and IR materials, and authoritative market and industry sources. For a fully apples-to-apples cross-comparison of peer EV/EBITDA and P/B, I deliberately did not force in a pile of third-party terminal figures, to avoid framework mismatch; the relative-valuation section therefore emphasizes MOS's own multiples and the verified comparison with its strongest rival, Nutrien. Another open question is that the future pace of monetizing the Ma'aden shares and the tax impact remain uncertain; this part provides positive support for equity value but should not be treated as 100% cash equivalent.

Final Verdict

【Final Rating】 Watch

【One-Sentence Investment Thesis】 MOS is an understandable fertilizer cycle stock with decent assets and a cheap-looking balance sheet, but it lacks a wide moat and stable free cash flow, so the current price looks more like "some value" than "cheap enough."

【Core Bull Case】

• Long-term demand for global phosphate and potash is still driven by food and soil nutrient replenishment, and is not demand that will easily disappear.

• Mosaic holds genuine asset and channel positions in North American phosphate, North American potash, and the Brazilian distribution network.

• The current share price corresponds to about 0.64x P/B, and the books also carry close to 1.96 billion dollars of Ma'aden share investment, providing some asset cushion.

• The company still maintains investment-grade liquidity thinking, and in 2026 has proactively cut capital spending, acknowledged reality, and made production-cut adjustments.

【Core Bear Case】

• The industry is a textbook commodity industry; MOS has no wide moat, and its earnings come more from the cycle than from a structural advantage.

• Free cash flow weakened markedly in 2024–2025, was already negative in 2025, and remained negative in the first quarter of 2026.

• Environmental and closure obligations are enormous, with an ARO balance of 2.6 billion dollars, and future environmental capital spending and remediation spending will keep eating cash.

• The buyback history is poor; average buyback prices in 2022–2024 were clearly above the current share price, showing that capital allocation is not excellent.

• In the first quarter of 2026, shocks from sulfur and raw materials already forced the company to withdraw phosphate production guidance and cut some production.

【Key Assumptions】

• Mid-cycle owner earnings can return to around 600 to 700 million dollars, rather than staying long-term at a lower level.

• The Ma'aden shares can gradually release liquidity in the future and roughly hold their value.

• The potash business at least maintains its current competitiveness, and the Brazilian business does not see persistent large credit losses.

• Environmental obligations do not see a major upward revision far beyond current disclosure.

【Fair Buy Price】 16 to 20 dollars/share is more ideal; below 21 to 23 dollars/share one can begin research-stage position-building; the current roughly 23.89 dollars is not outlandish, but it is not a conservative buy point. The basis is a conservative intrinsic value of about 19 to 22 dollars/share and a fair intrinsic value of about 27 to 31 dollars/share, while for this type of strongly cyclical business I want at least a 25% to 30% margin of safety.

【Target Holding Period】 Better suited as a 2-to-5-year cyclical mean-reversion position than as a naturally suited core compounding position of more than 10 years. If one must hold it with a 10-year view, the prerequisite is a clearly better entry price.

【Expected Annualized Return】 The following are inferences, not facts:

• Conservative scenario: 2% to 4%.

• Neutral scenario: 5% to 7%.

• Optimistic scenario: 9% to 11%. These estimates are derived from the current price, different intrinsic-value scenarios, and the existing dividend level, and assume neither large-scale renewed buybacks nor significant dilution.

【Maximum Loss Risk】 If high sulfur prices and high raw-material costs persist, Brazilian credit deteriorates, recovering potash supply pressures prices, and environmental obligations keep rising, while operating cash flow remains unable to cover capital spending over the long run, MOS could be priced by the market at a lower asset discount or a lower trough-earnings multiple, and a 40% to 60% downside is not unimaginable. This corresponds to a rough price zone possibly in the low teens of dollars. This is an inference based on historical cyclical volatility and current balance-sheet characteristics.

【Tracking Indicators】

• Potash and phosphate sales volumes, average prices, and unit cash costs.

• Whether operating cash flow and free cash flow recover to positive.

• Quarterly changes in inventory, receivables, and payables.

• The gap between 2026–2027 capital spending and depreciation and amortization.

• Whether ARO, environmental capital spending, and financial assurance amounts continue to be revised up.

• Changes in Brazilian bad debts, credit loss provisions, and distribution gross margin.

• The fair value of the Ma'aden shares and future disposal arrangements.

• Net debt/EBITDA and interest coverage.

• Whether buybacks restart, and the buyback price when they do.

• The output, cost, and supply-demand guidance of peers such as Nutrien and K+S.

【Signals Triggering Reassessment】

• Operating cash flow weak enough for several consecutive quarters that it cannot cover maintenance capital spending.

• A further significant increase in ARO or environmental-compliance spending.

• Deterioration of the potash business's cost position and loss of relative competitiveness.

• Material credit losses or share deterioration in the Brazilian business.

• Management again buying back large amounts at a high-valuation, weak-cash-flow stage.

• Significant deterioration in the value or monetizability of the Ma'aden shares.

【Final Recommendation】 Coolly put, MOS is worth studying, but there is no rush to buy. If you are a conservative-leaning long-term value investor, I would suggest treating it as a "cyclical candidate awaiting a better price" rather than a core asset to load up on immediately. The truly attractive buy point should appear when the market continues to push the share price down out of cyclical and raw-material fear, while the company shows no evidence of a balance sheet spiraling out of control, environmental obligations blowing out again, or a permanent deterioration in the competitiveness of its core assets. At that point, MOS would look more like a capital allocation with a margin of safety; today, it looks more like a long-term investment that has logic but not high enough conviction, with a trading character.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.