The Mosaic Company (NYSE: MOS) — The World’s #2 Rock, Marked at Half of Book at the Bottom of the Cycle

Independent research note. Report date: 2026-07-04. Price referenced: $21.13 (2026-07-02 close).

This article discusses valuation only as embedded expectations and scenarios. The analytical body contains no buy/sell recommendation and no price target. The single, deliberate exception is the Author’s Take block immediately below, which is fenced off as a subjective view.

⚡ Author’s Take

This block is the author’s own subjective opinion. It is general information and not investment advice. The analysis that follows takes no position and carries no price target.

Verdict: HOLD / a cost-average-ish cyclical trading below tangible book at the bottom of its cycle / accumulate-on-weakness in the high-$teens to ~$21 / not-a-short here. Medium-low conviction.

Mosaic is the mirror image of the businesses I have written up all quarter. Where the typical large-cap is a great franchise at its richest-ever multiple, Mosaic is a mediocre-quality, price-taking commodity producer at very nearly its cheapest-ever multiple — 0.55x book, 0.68x tangible book, 0.55x sales (all around the 4th–5th percentile of its own decade), against a mid-single-digit through-cycle ROIC that is presently 3.0% and troughed at 1.5% in 2024. The market is not wrong that this is a low-return business; three of Mosaic’s four earnings engines (phosphate, Brazil distribution, and the higher-cost end of potash) are structurally advantaged only at the bottom of the cost curve on a good day, and phosphate in particular is a cyclical price-taker with a large hidden environmental liability (the Florida phosphogypsum stacks) sitting on the balance sheet. This is emphatically not a compounder, and I would not pay up for it.

But at $21 you are not paying up. You are buying ~$38 of book (~$35 tangible) and a genuinely irreplaceable set of hard assets — the world’s #2 potash position (Esterhazy, Belle Plaine), an integrated Florida/Louisiana phosphate complex, a half-share of the Canpotex export cartel it co-owns with Nutrien, and a Brazil distribution footprint — for roughly half of replacement cost, in a $13.7B enterprise at ~8x trough EBITDA / ~6x mid-cycle EBITDA. The framing, grounded in the factor tape, is an abandoned, out-of-favor deep-cyclical value name (FactorsToday: Value β +0.46, OilPrice +0.43, GoldPrice +0.32, DividendYield +0.22; zero momentum, zero quality loading), sitting on a –70% round-trip from the 2022 super-cycle peak with a brutal risk-adjusted record (lifetime max drawdown –95%, negative Sharpe on every horizon). That is exactly the profile that is dangerous to short into a supply shock and unrewarding to own in size: a falling knife whose knife is a real asset. The 2026 Iran/Hormuz phosphate-and-sulfur squeeze is the swing factor — it has tightened finished-product prices but compressed stripping margins so hard that Mosaic is curtailing its own phosphate production, so the geopolitics cut both ways. One thing keeps my conviction low rather than medium: the capital-allocation and insider tape is unsupportive. Management bought ~$3.06B of stock pro-cyclically at ~$45 blended (an ASR at $64.37 near the 2022 peak) and then did zero buybacks at the 2025 trough; the CEO sold his entire direct stake at $31.56 in May 2025; and there is no cluster of conviction insider buying at these prices. For a stock that screens as a cyclical low, the people who know it best are not backing up the truck — so I won’t either. My accumulation zone is the high-$teens to ~$21, where you are paid a ~4% dividend and are buying assets below liquidation value while you wait for the cycle; I would fade strength into the low-$30s where the last two cyclical bounces died. Conviction: medium-low. Flips more bullish if potash tightness proves structural and sulfur/ammonia normalize so phosphate stripping margins re-expand and — the real “show-me” — management finally deploys the Ma’aden monetization runway into a counter-cyclical buyback at half of book. Flips bearish if BHP’s Jansen potash floods a Belarus/Russia-normalized market, sulfur stays elevated, and Brazil re-impairs — turning “cheap on book” into “book is too high.” Tag: the world’s #2 rock, marked at half of book — own the asset, don’t marry the business.

📈 Stock Price Action — Five-Year Event Map

Factual price history — not a recommendation and not a price target. Price moves are FACT (AZI adjusted-close CSV, 2020→2026-07-02); attributed drivers are INTERPRETATION, cross-referenced to earnings prints, benchmark fertilizer prices, 8-K events, and the news feed.

The arc. Mosaic has completed a violent commodity round-trip and now sits at the low end of it. On split/dividend-adjusted closes the stock bottomed near $8 (23-Mar-2020, COVID crash), reflated through the 2021 ag boom, and spiked to a five-year high of $70.81 (18-Apr-2022) on the Russia/Ukraine potash-and-phosphate super-shock. It then fell for nearly three years as fertilizer prices normalized, printing lower highs ($51.6 in Mar-2023, $36.9 in Jul-2025, $31.0 in Mar-2026) and finally a five-year low of $19.80 (10-Jun-2026). At $21.13 (2-Jul-2026) it is roughly –70% off the 2022 peak and –43% off the July-2025 52-week high, with a 52-week range of $19.80 → $36.94. This is the signature of a low-beta (0.78) commodity/value instrument in a multi-year downtrend, not a franchise — the FactorsToday leaderboard shows a lifetime maximum drawdown of –95% and a negative Sharpe ratio on every horizon.

Cycle narrative. (1) The 2020–21 climb is macro, not company-specific: Mosaic is a levered play on crop prices and planted acreage, both of which boomed post-COVID. (2) The April-2022 five-year high was a pure supply shock — the sanctioning of Belarusian and Russian potash (together ~40% of seaborne supply) sent potash toward ~$1,200/t and handed Mosaic a record year (2022 EBITDA $5.72B, EPS $10.57). (3) That windfall reversed brutally over two-plus years: Belarus/Russia exports found workarounds, farmers destocked and under-applied, potash and phosphate prices fell, EBITDA collapsed to $1.65B (2024) and GAAP EPS bottomed at $0.55, and the stock round-tripped to ~$23. (4) 2025 was a price- and volume-led recovery (EBITDA back to $2.13B, EPS $1.70) helped by Mosaic’s phosphate-asset investment program lifting production rates. (5) From mid-2025 the stock rolled over again on renewed farmer-affordability pressure, a difficult Brazilian credit environment, and — critically — a 2026 geopolitical squeeze that lifted finished phosphate prices but drove sulfur and ammonia raw-material costs up faster, compressing stripping margins so severely that Mosaic began curtailing its own phosphate output. (6) The recent bounce off the $19.80 low is the Iran/Strait-of-Hormuz fertilizer-supply-shock trade plus a fresh Rothschild Buy initiation ($30 target), partly offset by Morgan Stanley trimming to Equal-Weight ($26). Each price move is a FACT (AZI CSV); each attributed cause is INTERPRETATION.

1. Executive Summary

The Mosaic Company is the world’s second-largest potash producer and one of the largest integrated phosphate producers, organized in three segments — Phosphates, Potash, and Mosaic Fertilizantes (Brazil production + distribution, built from the 2018 Vale Fertilizantes acquisition). FY2025 revenue was $12.05B, adjusted/GAAP EBITDA ~$2.13B, and GAAP diluted EPS $1.70, recovering from a 2024 trough ($11.12B / $1.65B / $0.55) but still a fraction of the 2022 super-cycle peak ($19.13B / $5.72B / $10.57). The stock, at $21.13, trades at 0.55x book, 0.68x tangible book, and 0.55x trailing sales — all near the cheapest 5% of its own decade-long range — and at roughly 8x trough / ~6x mid-cycle EV/EBITDA on a $13.7B enterprise (net debt ~$5.24B).

The bull case in one line: irreplaceable hard assets — #2 global potash (Esterhazy, Belle Plaine, Colonsay, Carlsbad), an integrated Florida/Louisiana phosphate complex, a co-controlling stake in the Canpotex export JV, and a Brazil distribution network — bought below replacement cost and below liquidation value at the bottom of the cycle, with capital discipline (capex cut, portfolio pruning) improving free cash flow. The bear case in one line: a capital-consumptive, price-taking commodity producer that earns below its cost of capital in the trough and roughly at it mid-cycle (through-cycle ROIC ~mid-single-digits), with the weakest-quality segment mix among the fertilizer majors (heavy phosphate, no retail stabilizer, Brazil country risk), a large phosphogypsum environmental/ARO liability, and a new-supply overhang (BHP Jansen) capping its best business.

The honest verdict is that both are true: this is a genuinely cheap stock on assets and a genuinely low-quality business on returns. Unlike Nutrien — a cost-advantaged trough survivor priced mid-cycle — Mosaic is a higher-cost, phosphate-heavier version of the same commodity exposure, which is why the market marks it at half of book while Nutrien trades above it. The variant question is not “is it cheap” (it is) but “is book value real and is the cycle turning” — i.e., whether $38 of book survives a prolonged phosphate stripping-margin squeeze and a Belarus/Russia/Jansen-normalized potash market, or whether the 2026 geopolitical tightness marks a genuine cyclical trough. The body below argues the asset value is largely real and the downside is bounded by the cost curve, but that the business quality does not justify a premium to that asset value — making this an asset-value/mean-reversion situation, not a quality-compounding one.

(Sections 2–11 and the risk matrix synthesize the primary-source analysis that follows.)

2. Business Overview

What it is. Mosaic is the world’s #2 integrated phosphate producer and a top-tier global potash producer, formed in October 2004 from the combination of IMC Global and Cargill’s fertilizer business; headquartered in Tampa, Florida; 13,249 employees at 12/31/2025, serving customers in ~40 countries (FY2025 10-K, Item 1). It mines phosphate rock in Florida, Brazil and Peru, mines potash in Saskatchewan (and, until April 2026, New Mexico), and runs the largest fertilizer distribution network in Brazil. Three reportable segments — Phosphates, Potash, Mosaic Fertilizantes — plus “Corporate, Eliminations & Other” (China/India distribution, Mosaic Biosciences, and the Ma’aden equity stake).

FY2025 consolidated (FACT, 10-K). Net sales $12,052.4M (+8% YoY); gross margin $1,901.9M (15.8%), up from 13.6% in FY24; operating earnings $821.5M; net earnings to Mosaic $540.7M / $1.70 diluted EPS vs. $0.55 in FY24. GAAP net income was flattered by a $271.7M FX transaction gain and a ~$317M unrealized Ma’aden mark-to-market — non-operating and non-run-rate. The segment gross-margin walk, not GAAP EPS, is the honest read.

Phosphate — cyclical, currently the weak link (FACT). Net sales $4,576.5M; gross margin $437.3M — a 9.6% margin, down from 13.1% (FY24). Finished-product volume 5,945kt (DAP/MAP 2,935kt; MicroEssentials + feed 3,010kt), down 8% YoY; average finished selling price $667/t (+13%), DAP fob mine $670/t. The margin compression is a pure raw-material squeeze — ~$285M higher sulfur & ammonia, ~$140M higher conversion, ~$60M turnarounds; sulfur cost jumped to $237/long ton (+80%). Assets: Florida — New Wales (largest), Bartow, Riverview; Louisiana — Faustina (makes its own ammonia) and Uncle Sam. Phosphoric-acid capacity 4.5Mt but only 2.9Mt produced — chronic under-utilization the multi-year story management is now fixing. Mosaic makes ~72% of NA finished concentrates and ~10% of global phosphate; Florida rock ≈ ~47% of NA rock.

Potash — the crown jewel (FACT). Net sales $2,661.7M; gross margin $870.0M — a 32.7% margin (up from 26.9% FY24). MOP volume 8,262kt + Performance/Other 706kt; average finished price $266/t (+13%). This segment is only ~22% of net sales but ~46% of segment gross margin — Mosaic’s economic center of gravity. Assets: Esterhazy (Saskatchewan shaft, flagship, carries chronic brine-inflow risk); Belle Plaine (solution mine, low-gas-cost); Colonsay (explicitly the high-cost swing mine); Carlsbad, NM (K-Mag, sold April 2026). Mosaic ≈ ~34% of NA potash production.

Mosaic Fertilizantes (Brazil) — high revenue, thin margin (FACT). Net sales $4,847.3M (biggest revenue segment); gross margin $492.0M — 10.1%. This is a distribution/blending business: it buys 7,587kt of purchased nutrients for resale plus local production (5 rock mines, 4 plants, ~73% of Brazil concentrate production), terminals and blending in Brazil/Paraguay. Acquired as ex-Vale Fertilizantes in 2018.

Recurring vs. cyclical. Essentially none of Mosaic’s revenue is contractually recurring — it sells undifferentiated commodities (DAP, MAP, MOP) at spot/short-tenor prices set on global benchmarks. The only pockets of pricing stickiness are MicroEssentials (patented micronutrient-enhanced phosphate) and the Brazil distribution margin, neither a subscription-like annuity. Verdict: a price-taker whose earnings swing violently with the cycle; through-cycle ROIC ~3% (FY25) vs. 22% (2022) / 1.5% (2024) is the definitive tell.

3. Industry Dynamics

Mosaic sits in two structurally different industries, and conflating them is the single biggest analytical error here.

Potash — a genuine oligopoly, with a looming supply wave. Global potash supply is ~80% controlled by five players — Nutrien, Mosaic, K+S, ICL, and the Russia/Belarus complex — with two historical export cartels (Canpotex, Nutrien + Mosaic, ~40% of seaborne supply; and BPC). Resource is concentrated in Saskatchewan’s Elk Point Basin (~25% of global production). This is a textbook economies-of-scale + resource oligopoly: greenfield mines cost $8B+ and a decade, a steep cost curve, and coordinated export marketing — which is why Mosaic held a 32.7% potash gross margin in a soft year. But two Marathon capital-cycle red flags loom: BHP’s Jansen mine reaches first production mid-2027 (Stage 1 ~4.15Mt/yr at a guided $114–130/t cash cost; Stage 2 now deferred to ~FY2031 with a $2.3B impairment), and Belarus/Russia normalization adds latent supply. The capital-cycle lesson — high returns attract capital that mean-reverts — argues potash margins are more likely to compress than expand over the next 3–5 years. Near-term, though, 2026 is unusually tight (Canpotex sold out through June, China record Q1 imports, strong SE Asia palm-oil economics) — a favorable window sitting in front of a structural supply build.

Phosphate — fragmented, more cyclical, currently supply-starved. Phosphate is less consolidated than potash — OCP (Morocco) dominates with ~31% share and ~70% of global rock reserves; other majors include PhosAgro, China’s producers, and Saudi Ma’aden. Mosaic is only ~10% of global output. Economics turn on the “stripping margin” — finished price minus purchased sulfur and ammonia (~0.4t sulfur + 0.2t ammonia per tonne DAP). In 2025–26 that margin was crushed from both ends: sulfur spiked toward ~$1,150–1,200/t because ~50% of seaborne sulfur and much ammonia originate in the Middle East/Black Sea and the Strait of Hormuz conflict choked flows. The offsetting positive: China banned phosphate exports through August 2026, removing ~7–9Mt (~20–30% of global trade), so finished DAP is elevated ($670–780/t) even as producer margins are thin — an unusual, unsustainable configuration.

Industry verdict. Potash: structurally good, cyclically late — a real oligopoly with high barriers, entering a less-good part of its capital cycle (Jansen + latent Russia/Belarus). Phosphate: structurally mediocre, cyclically tight — fragmented, dominated by lower-cost OCP, hostage to two volatile purchased inputs Mosaic does not control; current tightness is real but geopolitical and temporary.

4. Competitive Position

Potash — a real but second-tier cost/scale advantage (Greenwald: economies of scale + resource). Mosaic’s potash moat is genuine — irreplaceable Saskatchewan reserves, Canpotex export scale, low-cost solution mining at Belle Plaine — and it shows up in the financials (32.7% gross margin, the highest-return segment). Remove the resource/scale advantage and the economics collapse to price-taker levels; that satisfies the moat test. But Mosaic is not the cost leader. Nutrien runs a cash cost widely cited under ~$60/t on ~20Mt of capacity vs. Mosaic’s ~11.5M; Mosaic’s cost is raised by Colonsay, its explicit high-cost swing mine, kept running only because demand is strong. Mosaic is a second-quartile potash producer inside a first-quartile industry — advantaged versus non-Canadian producers, disadvantaged versus Nutrien.

Phosphate — integrated, but no durable moat (stated directly). Mosaic’s Florida operations are vertically integrated (rock → acid → granulation) and it enjoys an advantaged raw-material logistics position (~80% of US ammonia internal/gas-linked; ~80% of sulfur as molten product from nearby Gulf refineries). But that edge is relative and temporary — it means Mosaic’s stripping margin compresses slightly slower than peers when sulfur spikes, not that it has pricing power. There is no durable phosphate moat. Mosaic is a price-taker on a global DAP benchmark set by OCP and Chinese policy; its Florida rock is finite and depleting (mining just moved to the higher-overburden South Fort Meade extension, pushing Florida mining cost to $63/t); acid capacity has run chronically below nameplate. The only genuine differentiator is MicroEssentials (patented premium phosphate) and the small, fast-growing Biosciences line — real but not enough to convert a cyclical converter into a moated business. The FY25 collapse to a 9.6% gross margin, driven wholly by input costs Mosaic cannot control, is the proof of no moat.

Mosaic Fertilizantes (Brazil) — distribution scale, weak returns. Mosaic owns the largest fertilizer distribution/blending/port network in Brazil (73% of local concentrate production, terminals, Paraguay reach) — a market-access advantage with scale and light switching-cost stickiness. But the segment earns a ~10% gross margin and sub-hurdle returns on much of its production — the just-idled Araxá/Patrocínio SSP assets “have struggled to meet our internal hurdle rates for some time,” triggering a $442M charge. It is a market-access advantage, not a profit engine.

Head-to-head. Versus Nutrien, Mosaic is structurally weaker — Nutrien pairs lower-cost/larger potash with a ~$20B retail network (a genuine, less-cyclical downstream earnings stream Mosaic entirely lacks). Versus CF Industries (nitrogen), CF enjoys a durable North American natural-gas cost advantage; Mosaic has no analogous structural feedstock edge in phosphate (it buys sulfur and ammonia). Verdict: not a durable-moat compounder. The only advantage tied to a financial outcome that would deteriorate without it is the potash resource/scale position — real but not best-in-class, and bracketed by a lower-cost rival above and a new-supply wave ahead. This is a well-run, cheaply-valued cyclical, not a widening-moat franchise. Phosphate has no durable moat; potash’s is real but second-tier.

5. Growth History and Forward Opportunities

History: cyclical price, not organic units. Mosaic’s revenue history is a price chart, not a volume-growth story. Phosphate finished volume has been flat-to-down (~1.65Mt/qtr for three quarters into Q1 2026), and the multi-year revenue swings track DAP/MAP and MOP benchmarks, not tonnes. There is essentially no secular unit-volume growth in the core — global nutrient demand grows ~1–2%/yr and Mosaic’s share is roughly stable.

(a) Phosphate production recovery. The stated target is 1.8–2.0Mt/quarter of finished product vs. a ~1.65Mt run-rate; three of four US plants are already at target phos-acid rates. This is recovery of lost capacity, not new growth — high-quality once stripping margins normalize, but currently deferred via curtailment. (b) Potash cost reduction. The Esterhazy hydrofloat ramp and Belle Plaine’s cheap gas “are expected to drive cost meaningfully lower,” offsetting Colonsay — genuine, durable margin (not volume) improvement. © Mosaic Biosciences. Revenue “expected to double again in 2026” with 8–10 new products — real but immaterial off a tiny (undisclosed) base; treat as optionality. (d) Rare earths (Rainbow Rare Earths / Uberaba). A March-2026 JV (Mosaic 51%) to recover REEs from an existing phosphogypsum stack: PEA after-tax NPV $916M @10%, IRR 45%, ~$217M average annual EBITDA (Mosaic’s 51% ≈ ~$110M), construction targeted 2027. The most interesting new vector — it monetizes an environmental liability into critical-minerals revenue and could replicate across Florida stacks — but early-stage and modest vs. a $2.1B-EBITDA company. (e) Ma’aden / MWSPC. A low-cost Saudi phosphate equity stake — a passive, modest earnings/dividend contributor and a potential monetization source, not a controllable growth lever.

Verdict: low-quality, cyclical growth with a thin layer of genuine optionality. The bulk of near-term “growth” is price recovery plus recapturing lost operating rates — mean-reversion, not compounding. The two structurally interesting items (rare earths, biologicals) are real and high-return but immaterial for 2–3 years. An investor is buying a cyclical trough recovery, not a growth story.

6. Financial Quality

Mosaic’s financials tell one story loudly: this is a high-amplitude cyclical whose reported returns are almost entirely a function of the fertilizer price, not of anything management does. The five-year record makes the point better than prose.

*ROIC’s book_val_per_sh reads higher ($44.7) on a different share/equity convention; common equity ($11.8B ex-minority) over 317.8M shares is ~$37–38/share, tangible ~$35. Figures: ROIC.ai statements, reconciled to 10-K.

Margin structure is entirely cycle-driven. Gross margin swung from 12% (2020 trough) to 30% (2022 peak) and back to ~16% — a 18-point range with no secular trend. There is no operating-leverage “compounding”: incremental margins are high on the way up (49% in 2025, 65% in 2023) and deeply negative on the way down precisely because price, not volume, drives the P&L. FACT: across the full 2020–2025 window Mosaic’s average ROIC is ~8–9% and its median is ~7% — i.e., roughly its cost of capital, with the 2022 spike doing most of the averaging work. INTERPRETATION: strip out the one super-cycle year and this is a business that earns below its cost of capital more often than above it. That is the defining financial fact, and it is why the equity deserves — and gets — a discount to book.

Cash generation is real but lumpy, and 2025 was distorted by working capital. Operating cash flow fell to $825M in 2025 from $1,299M in 2024, despite higher EBITDA — a $961M working-capital build (largely a $761M inventory increase from deferred Q4 phosphate demand) drained cash. INTERPRETATION: this is a timing item, not deterioration; Q1 2026 management explicitly flagged selling 1.9Mt of phosphate above production to release that inventory and improve working capital, so 2026 OCF should recover mechanically as inventory unwinds. Free cash flow (ROIC’s FCF-to-equity proxy) was ~$1.47B in 2025; on a cleaner basis, OCF of $825M less ~$1.4B capex was negative pre-working-capital-normalization — the tell that at trough prices Mosaic barely self-funds its capital program. The 2026 capex cut ($250M, to $1.25B) is a direct response.

The balance sheet is investment-grade but not a fortress. Q1 2026: total debt ~$5.76B, cash $282M, net debt ~$5.24B against ~$2.1B TTM EBITDA — ~2.5x net leverage at the current (near-trough) EBITDA, which would balloon toward 3x+ in a true trough and compress toward ~1x at mid-cycle. Net-debt/equity is ~44%. Current ratio 1.25x. This is a manageable but pro-cyclical balance sheet: leverage rises exactly when the business is weakest, which is the standard commodity-producer trap and a real constraint on counter-cyclical buybacks. Quality-of-earnings flags — GAAP EPS is noisy in both directions: (i) the 2025 effective tax rate of 52.7% (and 59% in 2024) is elevated by foreign-mix and non-deductible items, depressing GAAP EPS relative to pre-tax income; (ii) but FY2025 net income was simultaneously flattered by a ~$271.7M FX transaction gain (vs. a $685.8M loss in 2024) and a ~$317M unrealized mark-to-market gain on the Ma’aden equity stake — both non-operating and non-run-rate; (iii) recurring asset impairments ($99.9M in 2025, plus the $442M Araxá/Brazil charge) are a run-rate feature of this portfolio, not one-offs; (iv) depreciation (~$1.05B/yr) is a real, capital-intensive maintenance burden — this is not an asset-light cash machine. The segment gross-margin walk, not GAAP EPS, is the honest read of this business. Critically, book value carries a $2,601M asset-retirement obligation (Florida/Louisiana phosphogypsum-stack closure, water treatment, monitoring — flagged as a critical audit matter), a permanent, growing, largely non-discretionary claim on cash that the market under-weights when it anchors on 0.55x book. Verdict: economics do NOT improve with scale in any durable way; the business is a price-taker whose returns oscillate around its cost of capital, and whose balance sheet is investment-grade but pro-cyclically levered. The financials support an asset-value thesis, not a quality thesis.

7. Capital Allocation

If the business quality is the reason not to pay up, capital allocation is the reason to keep position size modest — across the full cycle it has been below average, and the improvement is recent and unproven.

Buybacks were textbook pro-cyclical — bought high, stopped at the low. The repurchase record is the clearest indictment of Mosaic’s through-cycle discipline (figures from the SHARE REPURCHASES notes across four 10-Ks):

Across 2021–2024 Mosaic spent ~$3.06B to retire ~67.7M shares at a dollar-weighted ~$45, the heaviest single-year outlay ($1.665B in 2022) going in at ~$54 including an accelerated share repurchase priced at $64.37 — within a few dollars of the all-time high. Those shares are worth ~$1.43B at today’s $21.13 — an unrealized ~$1.6B destruction of capital versus what was paid — and the Board sized its $3.0B of 2022 authorizations to the peak. Then it went to zero repurchases in 2025 with the stock near its cyclical low. This is buy-high/stop-at-the-low in its purest form; the one defense (2025 cash was needed for the balance sheet and Brazil restructuring) is real but does not rescue the timing.

The dividend, by contrast, has been managed prudently. Dividends paid rose $76M (2020) → $198M (2022) → $280M (2025); per share ~$0.88 in 2025, a ~4.2% yield. Kept deliberately small relative to through-cycle cash generation (~$280M against >$1B of normal operating cash flow), it is comfortably covered even when it exceeds a trough GAAP EPS (the ~118% payout on 2024’s depressed number). The discretionary buyback, not the dividend, is where capital was misallocated.

M&A scorecard — Vale Fertilizantes (2018) has not earned its cost of capital. Mosaic bought Vale Fertilizantes in January 2018 for ~$2.5B (cash + ~34.2M shares), creating the Brazil segment. It has since absorbed a $588.6M goodwill impairment in 2019 (one year after closing) and a further $96.3M in 2025 plus smaller true-downs — ~$685M combined. Two write-downs on the same franchise, one shortly after purchase and one seven years later, are strong evidence the deal was struck near a phosphate peak at the wrong price; the strategic logic (Brazil is the world’s largest fertilizer import market) was defensible, the price and timing were not. The 2025–26 Araxá/Patrocínio idling and mine divestitures (Taquari −$66M, NM/Carlsbad −$185M impairments on sale) are Mosaic exiting sub-scale assets below carrying value — cleaning up prior over-investment.

The one genuinely good action: MWSPC → Ma’aden. On 24-Dec-2024 Mosaic exchanged its 25% MWSPC (Saudi JV) stake for 111.0M Ma’aden shares valued ~$1.5B, booking a $522.2M gain, now carried at fair value with a further +$317.4M 2025 mark. It converted an illiquid, non-controlled JV into a marked-to-market ~$1.8B+ liquid (if restricted, five-year lock-up) equity stake worth materially more than the exchange value — a self-funding, non-dilutive source to reduce net debt as restrictions lapse. This is the best capital-allocation action in the file, and it is already executed.

Incentives lean to EBITDA/volume, not returns on capital. The 2025 annual bonus keys on Free Cash Flow, adjusted EBITDA (“Price Normalized Earnings”), SG&A, and Safety; the LTI is 40% time RSUs + 60% relative-TSR performance units (gated by cumulative positive adjusted net earnings). Positives: naming the metric “Price Normalized Earnings” strips commodity-price windfalls, and the 2025 cash bonus paid ~$0 in the down year — genuine downside flex. Negatives: there is no explicit ROIC/ROE metric — precisely the discipline a serially over-investing cyclical most needs; CEO comp rose to $10.41M in 2025 even as shareholders lost money (equity grants sized off target, not entry price); and personal alignment is thin — CEO Bodine at 2.9× salary and CFO Siani Pires at 1.6× are both below ownership guidelines, and all 18 directors/officers together own <1% (988,177 shares).

The insider tape is negative. Of 195 Form 4s, there are only two open-market purchases — a director’s 15,600 shares @ $63.49 at the 2022 peak, and a token 685 shares @ $25.53 in Nov-2025. There is no cluster of conviction buying at the $21–25 trough — the classic bullish tell for a cyclical bottom is absent. Instead, CEO Bodine sold his entire direct common stake (180,708 shares @ $31.56) in May 2025. For a stock that screens as a possible cyclical low, insiders are conspicuously not buying with their own money.

Verdict: a below-average through-cycle capital allocator showing recent, unproven improvement. Pro-cyclical buybacks (~$1.6B underwater), a value-destructive Brazil acquisition (~$685M impaired), no ROIC in the comp plan, thin insider ownership, and no trough buying are the negatives. The prudent dividend, the value-accretive Ma’aden conversion, and the disciplined 2025–26 pivot (capex cut to $1.25B, divestitures, deleveraging bias) are the offsets. The forward swing factor is whether this team deploys the Ma’aden runway and free cash flow into a counter-cyclical buyback at half of book — or idles the authorization again.

8. Changes and Headwinds — Last Two Years

The price round-trip is the master fact. FY2025 EBITDA was $2.13B and EPS $1.70 versus a 2024 trough EPS of $0.55 — a ~3x earnings swing driven almost entirely by the fertilizer-price cycle, not volume. The shares round-tripped from the $70.81 April-2022 peak to a five-year low of $19.80 (June 2026). This is a classic Marathon capital-cycle unwind: the 2021–22 grain/energy shock pulled forward a wave of pricing that has fully mean-reverted, and the market now prices Mosaic near book on the belief that mid-cycle economics are structurally worse than the 2022 mirage implied.

A genuinely bifurcated 2026 shock. The current disruption is bullish for finished-product price and bearish for producer margin simultaneously. Management (Bodine): “Roughly 20% of global phosphate, 1/3 of urea, 1/4 of ammonia and 1/2 of seaborne sulfur volumes originate in the Middle East.” Finished DAP guidance is $760–780/t for Q2, yet stripping margins are compressed because sulfur and ammonia spiked harder — Q1 stripping margin ~$400/t (sulfur $379/t), but the marginal cargo cost is sulfur ~$1,200/t and ammonia ~$800/t, at which “the marginal stripping margin is below variable costs.” This is why higher headline prices have not rescued the stock — the incremental tonne is unprofitable, forcing curtailment rather than harvest.

Production curtailments — a defensive pivot. Mosaic is “partially reducing production rates at Bartow and Louisiana” — about half of Louisiana’s ~1.4Mt and half of Bartow’s ~2Mt annualized capacity — and scaling back Brazil. With China’s export ban and global sulfur scarcity, “there is not going to be enough phosphate to meet global demand.” The curtailments are “temporary” and “can be quickly unwound,” but the timing is hostage to the Strait of Hormuz.

The US phosphate asset-investment program is the one clean success. After two years of reliability capex, “3 of our 4 facilities operating at targeted rates,” and Q1 phosphate sales of 1.9Mt were “the highest quarterly sales volume total for the segment in 5 years.” Real, controllable progress that finally addresses the multi-year US operating-rate problem — but the raw-material squeeze mutes the payoff near-term.

Brazil restructuring — a strategic retreat from bad assets. Mosaic idled SSP production at Araxá and mining at Patrocínio ($442M charge, $328M non-cash) and is assessing a sale, and sold Carlsbad, NM (April 2026) — disciplined capital reallocation out of sub-hurdle, high-ARO assets, modestly accretive to segment margin and future ARO. Cost/capex defense: 2026 capex cut $250M to $1.25B; a $50M annualized workforce reduction; $120M of phosphate inventory released in Q1. Leadership: Bruce Bodine is President & CEO (transition from Joc O’Rourke completed 2024); Luciano Siani Pires (ex-Vale) is CFO — a more portfolio-surgical posture than the prior harvest-the-cycle era. Sell-side split: Morgan Stanley cut to $26 (Equal-Weight, 30-Jun-2026); Rothschild Redburn initiated Buy at $30 (26-Jun-2026, arguing margins are “unsustainably low but set to recover”). Verdict: net neutral-to-mildly-strengthening for a patient thesis, weakening for a near-term one. The controllable items genuinely improve mid-cycle earnings power and balance-sheet resilience; they are swamped near-term by an exogenous margin shock management cannot control or time.

9. Risk Analysis (Risk Matrix)

The three that matter: #1, #2, #6. #1/#2 are the cyclical engine of the current drawdown and, by management’s own framing, temporary. #6 is the structural one — the ~$2.6B ARO is a permanent, growing, largely non-discretionary claim on cash that the market under-weights when it anchors on 0.55x book. Risk of a total loss is low (real assets, investment-grade balance sheet, positive mid-cycle FCF); the realistic downside is a value-trap of years of sub-cost-of-capital returns with the stock oscillating around book, not a zero.

10. Valuation Discussion (Embedded Expectations)

The one-sentence version: the market is pricing Mosaic as a below-cost-of-capital commodity producer at the bottom of its cycle — which is exactly what it is — and giving essentially no credit for either asset value above ~half of book or a cyclical recovery.

Own-history context (the highest-signal datum). On AZI’s own-history percentiles Mosaic trades at the 4.6th percentile on P/B (0.57x), the 4.3rd on P/S (0.55x), and the 40th on P/E (9.1x trailing) — composite 16th percentile. The split is diagnostic: cheap-on-assets/sales, mid-on-earnings. INTERPRETATION: the P/E is not low because earnings are near a cyclical low and the market refuses to capitalize them at a high multiple (correctly — you never pay a high P/E for a commodity trough). The P/B and P/S, which are far less cycle-distorted, are the real tell: the equity is priced at roughly half of book and half of sales, near the cheapest it has ever been. The only comparable episode was the 2015–16 and early-2020 troughs.

Cross-check on multiples. EV/EBITDA sits at ~8.0x on TTM EBITDA of ~$1.7B (a near-trough number) and ~6.0x on FY2025’s $2.1B; its own history ranges from 3.2x (against 2022’s peak EBITDA) to 10.4x (against 2020’s trough EBITDA), so ~6–8x is mid-range on the multiple but the denominator is depressed. On mid-cycle EBITDA of ~$2.5–3.0B (2021 was $3.4B, 2023 $2.2B), the same $13.7B EV is ~4.5–5.5x — genuinely cheap for the asset base. EV/sales of 1.1x is near the low end of the 0.9–1.5x range.

Embedded-expectations / scenario frame. Holding the ~$13.7B enterprise value and ~$5.24B net debt over ~318M shares:

INTERPRETATION: at $21 the market is underwriting something close to the bear case — permanent trough-ish EBITDA of well under $2B — and pricing the equity below tangible book ($35). An asset-value floor cross-check tells the same story: even a 20–30% haircut to book (for phosphogypsum ARO risk and Brazil) leaves book near $26–30, above the current price. The variant is not the multiple; it is whether the denominator (mid-cycle EBITDA) is ~$1.8B (bear, price is fair) or ~$2.5–3.0B (base, price is 50%+ too low). The evidence — a functioning potash oligopoly, a structurally under-applied global soil-nutrient balance (management’s agronomic point), and Mosaic’s own cost-reduction and portfolio-pruning — leans toward the base case being more likely than the bear, but the phosphate stripping-margin squeeze and Brazil are genuine reasons the market is discounting it. No price target; the analysis frames what the price implies. Verdict: priced for permanent trough; asymmetry is favorable if book value is real and the cycle mean-reverts, which is a commodity-cycle bet, not a quality bet.

11. Variant Perception

Consensus view. Sell-side is split and unenthusiastic — Morgan Stanley Equal-Weight ($26 target, cut on 30-Jun-2026), Rothschild fresh Buy ($30, 26-Jun-2026) — clustering in the mid-to-high $20s, i.e., “cheap but no catalyst, wait for the cycle.” The market treats Mosaic as a low-quality, phosphate-heavy commodity producer with Brazil baggage and a squeezed near-term, worth less than book because its returns don’t justify book.

The strongest bull case. You are buying the world’s #2 potash franchise plus an integrated phosphate complex plus a half-share of Canpotex for ~half of tangible book at a cyclical bottom. Global soil-nutrient balances are being drawn down by years of under-application (there is no substitute for phosphate and potash in agriculture); the potash oligopoly is intact; Belarus/Russia have re-entered but not collapsed pricing; and the 2026 sulfur/ammonia squeeze that is crushing phosphate stripping margins is itself curtailing global phosphate supply, setting up a sharp price recovery when raw materials normalize. Management is finally acting like a cyclical steward — cutting capex, pruning Brazil, and (the swing catalyst) it could buy back a lot of stock at half of book. Add hidden optionality: rare-earth recovery from phosphogypsum (Rainbow Rare Earths JV), Mosaic Biosciences biologicals doubling, and monetization of the MWSPC/Ma’aden stake.

The strongest bear case. Book value is too high and the business earns below its cost of capital. Phosphate is a structural price-taker with a massive, growing phosphogypsum-stack closure/ARO liability that the market rightly discounts; Brazil is a serially value-destructive, FX- and credit-exposed distribution business bought at the top (Vale Fertilizantes, 2018) and now being impaired and idled; and the crown-jewel potash business faces BHP’s Jansen greenfield ramping from 2027 into a market where Belarus/Russia supply has fully recovered — a genuine oversupply overhang that caps the one good segment. Through the cycle this company has destroyed or barely preserved capital, its buybacks were pro-cyclical (heavy near the 2022 peak, light at the trough), and “cheap on book” is a value trap if the assets keep earning 3% ROIC. The –95% lifetime drawdown and negative Sharpe on every horizon are the tape telling you this is a falling knife, not a coiled spring.

The 3–5 assumptions that matter most, and what falsifies each:

1. Is mid-cycle EBITDA ~$1.8B (bear) or ~$2.6B+ (base)? Falsify base: two more years of sub-$2B EBITDA with potash net-realized prices stuck below ~$220/t. Falsify bear: potash + phosphate net realizations and volumes drive EBITDA back above $2.5B by 2027.
2. Is book value real? Falsify: a large phosphogypsum-ARO or Brazil impairment that cuts tangible book toward the current price. Confirm: assets sold/monetized (Araxa, MWSPC) at or above carrying value.
3. Does the potash oligopoly hold against Jansen + Belarus/Russia? Falsify: potash reverts toward $200–250/t as Jansen ramps. Confirm: Canpotex/oligopoly discipline holds pricing in the $250–300s.
4. Is management a good enough cyclical steward to buy stock counter-cyclically? Falsify: another idle authorization while the stock sits at half of book. Confirm: a real buyback executed sub-$25.
5. Does the phosphate stripping-margin squeeze normalize? Falsify: sulfur/ammonia stay elevated into 2027, keeping phosphate curtailed and unprofitable at the margin. Confirm: raw materials normalize and the curtailed capacity restarts into higher finished prices.

The factor-positioning read (FactorsToday). Mosaic loads Value +0.46, OilPrice +0.43, GoldPrice +0.32, DividendYield +0.22, SmallSize +0.23 — and carries no Momentum and no Quality loading whatsoever (L1-sparse model zeroes them). Its factor-nearest peers are Alcoa, Olin, Dow, Westlake — deep-cyclical materials/chemicals, not quality compounders. The risk-adjusted record is uniformly poor (lifetime Sharpe ~0.03, y1 Sharpe –0.98, y5 –0.17; lifetime max drawdown –95%). INTERPRETATION: consensus is positioned in this name as an out-of-favor, low-momentum value/commodity instrument — which is where the variant opportunity lives if the cycle turns, and where the value trap lives if it doesn’t. The tape is not pricing any recovery; that is the asymmetry and the risk in one sentence.

12. Fact vs. Interpretation

13. Open Questions

1. What is normalized/mid-cycle EBITDA? The entire thesis hinges on whether it is ~$1.8B or ~$2.6B+. Needs a segment-level bridge on potash net-realized price × volume and phosphate stripping margin at normalized sulfur/ammonia.
2. Is the $2,601M phosphogypsum-stack ARO adequately reserved, and what is its cash-flow timing? The liability is on the balance sheet (so book already carries it), but the cash timing and the risk of upward revision (New Wales sinkhole and Piney Point legacy are reminders of tail risk) is the real question on whether “book value is real.”
3. Will management buy back stock counter-cyclically at half of book, or idle the authorization again? The capital-allocation tell.
4. What is the realizable value of the MWSPC/Ma’aden stake and the Araxa/Patrocinio assets held for sale? Above or below carrying value?
5. Is Mosaic Biosciences / rare-earth optionality real value or narrative? Biosciences “doubling” off a tiny base; the Rainbow Rare Earths phosphogypsum project is early-stage.
6. How high is Mosaic on the potash cost curve vs. Nutrien? Colonsay is a high-cost swing mine; is blended cash cost bottom-quartile or second-quartile?

14. What Must Be True

For the bull (asset-value/mean-reversion) case — falsification test each:

• Book value is real and floors the downside. Falsified if a phosphogypsum-ARO or Brazil impairment cuts tangible book to ~$21 or below within 18 months.
• Mid-cycle EBITDA recovers to ≥$2.5B by 2027. Falsified if EBITDA prints below $2B for FY2026 and FY2027 with no line of sight to recovery.
• The potash oligopoly holds pricing above ~$250/t through the Jansen ramp. Falsified if potash net realizations fall below ~$220/t on Jansen/Belarus/Russia supply.
• Management acts counter-cyclically (buyback sub-$25, capex discipline, accretive asset sales). Falsified if the buyback stays idle and Brazil re-impairs.

For the bear (value-trap) case — falsification test each:

• The business permanently earns below its cost of capital (ROIC stuck at 2–4%). Falsified if ROIC sustainably exceeds ~9% for two consecutive non-spike years.
• Phosphate is structurally unprofitable at the margin (sulfur/ammonia stay elevated). Falsified if stripping margins re-expand above ~$500/t on raw-material normalization and curtailed capacity restarts profitably.
• Book value overstates realizable asset value. Falsified if Araxa/MWSPC and any other divestitures clear at or above carrying value.
• New supply (Jansen) breaks potash. Falsified if potash pricing proves resilient through 2027–28.

15. Source Appendix

Primary filings (SEC EDGAR, CIK 0001285785; via SEC EDGAR):

• FY2025 Form 10-K, filed 2026-02-27 (mos-20251231.htm) — segment MD&A, Properties, ARO Note ($2,601M), goodwill impairments, share-repurchase note, MWSPC/Ma’aden exchange.
• FY2024, FY2023, FY2022, FY2021 Form 10-Ks (2025-03-03, 2024-02-22, 2023-02-23, 2022-02-23) — multi-year segment/volume/price trends and buyback history.
• DEF 14A proxy statements 2023–2025 (and 2026 proxy referenced) — incentive-metric design, NEO comp, ownership guidelines.
• Q1 2026 earnings call transcript (2026-05-11; ROIC.ai MCP) — sulfur/ammonia marginal costs, stripping margins, curtailments, Canpotex/Colonsay, Araxá idling, capex cut, Rainbow REE, Biosciences.
• Form 4 corpus (195 filings) — insider-transaction read.

Quantitative data: ROIC.ai MCP (income statement, balance sheet, cash flow, profitability/valuation ratios, enterprise value, per-share data); AZI valuation_index own-history percentiles and 5-year adjusted-close price CSV; FactorsToday factor model (/stock-loadings, /leaderboard, /related-stocks, /stock-info). All reconciled to filings; ROIC/AZI/FactorsToday are third-party aggregated data, not primary.

Peer / industry cross-reference: public Nutrien (NTR) filings for potash-oligopoly, Canpotex, Jansen and Belarus/Russia framing; UBS fertilizer/potash sector primers (2006–2009, framework context only).

External (accessed 2026-07-04): BHP Jansen cost/timing (Globe and Mail; mining.com; mining-technology.com); OCP phosphate share (Univ. de Navarra Global Affairs, 2025); potash oligopoly structure (Wiley, 2025); China phosphate/acid export suspension (S&P Global; Exiger); 2026 Strait of Hormuz crisis and fertilizer impact (Wikipedia; Carnegie Endowment; farmdoc); fertilizer prices (World Bank; DTN; InvestingNews); Rainbow Rare Earths / Uberaba (Nasdaq; Northern Miner); sell-side actions (Markets Daily / Morgan Stanley $26; StreetInsider / Rothschild Redburn $30).

Facts are cited to primary filings or named third-party sources; interpretations are labeled as such. Price data as of 2026-07-02 close ($21.13). No price target and no buy/sell recommendation appears in this institutional body; the sole opinion is the labeled Author’s Take.

APPENDIX A — Standard Diligence Questionnaire — The Mosaic Company (NYSE: MOS)

Report date 2026-07-04; price $21.13. Fact / Interpretation / Assumption labeled where it matters. Supplemental to the memo.

General

What thoughtful questions have other investors asked about this company? The live debates (from the Q1-2026 call and sell-side): (1) How much stranded sulfur sits behind the Strait of Hormuz and how fast do Gulf refineries resume? (JPMorgan’s Zekauskas — management conceded “I don’t think there is good visibility”). (2) Is the Q2 phosphate stripping-margin guide (>$400/t on a 60%-committed book) achievable if sulfur keeps rising? (3) Is Q2 potash shipment guidance soft? — management clarified K-Mag (~175kt) was moved out of the potash line, so underlying demand is “pretty normal.” (4) What is normalized mid-cycle EBITDA? — the crux of the whole thesis. (5) Will management buy back stock at half of book, or keep the authorization idle?

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Low. (Fact) FY2025 EBITDA $2.13B and EPS $1.70 sit far below the 2022 super-cycle peak ($5.72B / $10.57) and modestly above the 2024 trough ($1.65B / $0.55). Q1-2026 shows phosphate margins re-compressing on the sulfur/ammonia squeeze, so near-term earnings are near-trough, not mid-cycle.

Driven by the external environment or internal actions? Overwhelmingly external (fertilizer prices, sulfur/ammonia costs, crop prices, geopolitics). Internal actions (US phosphate reliability program, Brazil pruning, cost-out, capex discipline) are real but second-order to the commodity cycle. (Interpretation)

How stable are revenues? Highly unstable — revenue swung from $8.7B (2020) to $19.1B (2022) to $12.1B (2025). It is a price chart, not a volume series; unit volumes are roughly flat. (Fact)

Outlook for products/services; how big is the market — growing, shrinking, domestic or international? Global crop-nutrient demand grows ~1–2%/yr (population, diets, soil-nutrient depletion); there is no substitute for phosphate and potash. The market is global (Mosaic sells to ~40 countries; Brazil, India, SE Asia, North America are key). Structurally the market grows slowly; Mosaic’s share is roughly stable. (Fact/Interpretation)

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Potash: becoming more competitive medium-term as BHP’s Jansen (mid-2027) and normalized Belarus/Russia supply enter. Phosphate: currently less competitive (China export ban, sulfur-driven curtailments) but structurally fragmented and OCP-led. (Interpretation)

How profitable is the business (ROIC, ROE)? Poor through-cycle: ROIC 3.0% (2025), 1.5% (2024 trough), 22% (2022 peak), ~12.8% (2021 normal) — averaging roughly its cost of capital, i.e., value-neutral through the cycle. ROE similar (3.8% / 1.2% / 27.3%). (Fact)

How profitable is the industry — competitors, barriers to entry? Bifurcated. Potash is a high-barrier oligopoly (five players ~80% of supply; $8B+/decade to build a mine; Canpotex) — genuinely profitable for low-cost producers. Phosphate is lower-barrier, fragmented, and hostage to purchased sulfur/ammonia — structurally less profitable. (Fact/Interpretation)

Can the business be easily understood? Yes — it mines two rocks and sells them at global benchmark prices, plus distributes in Brazil. The complexity is in the cost curve and the raw-material (stripping-margin) mechanics, not the model. (Interpretation)

Can it be undermined by foreign low-cost labor? Not labor — but yes by foreign low-cost resource: OCP’s Moroccan rock, Russian/Belarusian potash, and Gulf/Saudi (Ma’aden) integrated phosphate are the competitive threats, all cost-curve driven. (Interpretation)

Do brands matter? Barely. MicroEssentials (patented premium phosphate) and Aspire/K-Mag carry modest premia and grower preference; MOP/DAP/MAP are undifferentiated commodities. (Fact)

Nature of competition / customers’ switching costs? Competition is on delivered cost (price + logistics). Switching costs are low for commodity nutrients; the Brazil distribution network and MicroEssentials agronomy provide light stickiness, not lock-in. (Interpretation)

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The Ma’aden equity stake (111M shares, carried at fair value ~$1.8B+, appreciating) is a real, somewhat-hidden liquid asset. Irreplaceable Saskatchewan/Florida reserves are carried at historical cost, likely below replacement value. (Fact/Interpretation)

Off-balance-sheet liabilities? The major on-balance-sheet long-tail item is the $2,601M asset-retirement obligation (phosphogypsum-stack closure, water treatment) — a critical audit matter. Environmental tail risk (gypstack failure — New Wales sinkhole, Piney Point precedent) is the off-balance-sheet catastrophic exposure. (Fact)

How conservative is the accounting? Mixed. Recurring impairments ($620M+ in 2025) and the large ARO are conservatively recognized, but GAAP EPS is distorted both ways (FX gains, Ma’aden marks, elevated tax rate). Segment gross margins are the cleaner read. (Interpretation)

How CapEx-hungry is the business? Very — sustaining capex ~$1.25–1.4B/yr against ~$1.05B annual D&A; this is a capital-intensive extractive business, not asset-light. (Fact)

Capital Allocation & Management

How much FCF, and how is it used? Operating cash flow ran $825M (2025, working-capital-depressed) to $3.9B (2022 peak); normal ~$1.5–2.5B. Uses: capex (~$1.25B), dividend (~$280M), and historically buybacks (now idle) and debt paydown. Trough FCF barely covers capex + dividend. (Fact)

Significant acquisitions recently? No recent M&A; the defining deal was Vale Fertilizantes (2018, ~$2.5B), since ~$685M impaired. Recent activity is divestiture (Carlsbad NM, Taquari, Araxá idling). (Fact)

Buying back shares? Not currently — zero in 2025 after ~$3.06B in 2021–2024 (~$45 blended, pro-cyclical, ~$1.6B underwater). The forward buyback is the key capital-allocation swing factor. (Fact)

Issuing shares to insiders? Modest — SBC ~$31M/yr; share count fell from ~379M (2020) to ~318M (2025) via buybacks. Not a dilution story. (Fact)

Compensation policy of directors/management? Annual bonus on FCF/adjusted-EBITDA (“Price Normalized Earnings”)/SG&A/safety; LTI 40% RSU + 60% relative-TSR. No ROIC metric. 2025 cash bonus paid ~$0 (genuine downside flex); CEO comp $10.41M rose YoY despite share declines. (Fact)

Motivations of management? Alignment is thin: all insiders own <1% (988,177 shares); CEO (2.9×) and CFO (1.6×) below ownership guidelines; CEO sold his entire direct stake at $31.56 in May 2025. Incentives tilt to EBITDA/volume over per-share/return-on-capital. (Fact/Interpretation)

Valuation & Market Data

ADR, MLP, or K-1 issuer? No — a US C-corp common stock (NYSE: MOS), standard 1099 dividend. (Fact)

Dividend policy? Small, rising, well-covered: ~$0.88/share (2025), ~4.2% yield, ~$280M/yr; ~46% payout on 2025 EPS (but a normal ~15–30% on mid-cycle earnings). Sustainable through the trough. (Fact)

How profitable is the business? See ROIC above — mid-single-digit through-cycle returns; low quality. (Fact)

Net income diverging from cash from operations? Yes, and both ways: 2025 OCF ($825M) fell below net income ($541M… no — OCF exceeded NI, but was depressed by a $961M working-capital build); GAAP NI is noisy from FX/Ma’aden marks and impairments. Normalized cash earnings are the better anchor. (Fact/Interpretation)

Risks & Downside

What would cause the stock to decline? A protracted sulfur/ammonia squeeze keeping phosphate unprofitable; potash oversupply (Jansen + Belarus/Russia) breaking pricing; a Brazil or ARO impairment cutting book; farmer demand destruction; a China-export-ban lift pressuring phosphate. (Interpretation)

Risk of a catastrophic loss? Low-probability/high-severity: a phosphogypsum-stack failure or spill (environmental catastrophe), or a large upward ARO revision. Not a solvency risk under normal scenarios. (Interpretation)

Chance of a total loss? Very low. Real, irreplaceable hard assets; investment-grade balance sheet (~2.5x near-trough leverage); positive mid-cycle FCF; trades below tangible book. The realistic downside is a value-trap of years of sub-cost-of-capital returns, not a zero. (Interpretation)

Recent News & Events

Has the business environment changed recently? Yes — the 2026 Strait-of-Hormuz conflict (blocked since 28-Feb-2026; widened route/reopening narrative from 27-Jun) drove a bifurcated shock: higher finished-fertilizer prices but crushed phosphate stripping margins, forcing Mosaic to curtail Bartow/Louisiana. China’s phosphate-export ban (through Aug-2026) supports prices. (Fact)

Significant acquisitions? No — the direction is divestiture/idling (Carlsbad sold Apr-2026; Araxá/Patrocínio idled; Rainbow Rare Earths JV signed Mar-2026). (Fact)

Change in accounting policies? None material identified. (Fact)

Recent changes — new markets, facilities, management? Bruce Bodine CEO (from 2024), Luciano Siani Pires CFO (from 2025, ex-Vale); US phosphate reliability program (3 of 4 plants at target rates); 2026 capex cut $250M to $1.25B; Rainbow Rare Earths (Uberaba) and Mosaic Biosciences as new optionality. (Fact)

APPENDIX B — Source Appendix — The Mosaic Company (NYSE: MOS)

Report date 2026-07-04. Primary sources over secondary; every non-obvious fact in the memo traces to one of the below. Price data as of 2026-07-02 close ($21.13).

Primary — SEC filings (EDGAR, CIK 0001285785; via SEC EDGAR)

Primary — quantitative data services

• ROIC.ai MCP — income statement, balance sheet, cash flow, profitability ratios (ROE/ROA/ROIC/margins), enterprise value ($13.7B), valuation multiples, per-share data; FY2020–FY2025 + TTM. Reconciled to filings (third-party aggregated, not primary).
• AZI — valuation_index own-history percentiles (P/E 40th, P/B 4.6th, P/S 4.3rd, composite 16.4th); 5-year adjusted-close price CSV (5y high $70.81 18-Apr-2022; 5y low $19.80 10-Jun-2026; 52-wk $19.80–$36.94); news feed (Morgan Stanley $26, Rothschild $30, Hormuz items).
• FactorsToday — /stock-loadings (Value +0.46, OilPrice +0.43, GoldPrice +0.32, DividendYield +0.22, SmallSize +0.23; no Momentum/Quality; Materials β 0.63–0.70, R² ~31%); /leaderboard (lifetime max DD −95%, negative Sharpe all horizons, y1 return −42%); /related-stocks (AA, OLN, DOW, WLK — deep-cyclical peers).

Peer / industry framework

• Public Nutrien Ltd. (NTR) filings (40-F/annual report) — potash-oligopoly structure, Canpotex, BHP Jansen, Belarus/Russia, Iran/Hormuz framing (co-owner of Canpotex; the closest listed peer).
• General fertilizer/potash industry value-chain framework (public sector literature).

Secondary — external (accessed 2026-07-04)

• BHP Jansen cost/timing: theglobeandmail.com; mining.com; mining-technology.com
• OCP / global phosphate share: en.unav.edu (Univ. de Navarra Global Affairs, 2025)
• Potash oligopoly structure: onlinelibrary.wiley.com (2025)
• China phosphate/sulfuric-acid export suspension: spglobal.com; exiger.com
• 2026 Strait of Hormuz crisis & fertilizer impact: en.wikipedia.org; carnegieendowment.org; farmdocdaily.illinois.edu
• Fertilizer prices (DAP/MAP/potash/sulfur): blogs.worldbank.org; dtnpf.com; investingnews.com; cobank.com
• Rainbow Rare Earths / Uberaba economics: nasdaq.com; secure.northernminer.com
• Sell-side actions: themarketsdaily.com (Morgan Stanley $26, 30-Jun-2026); streetinsider.com / investing.com (Rothschild Redburn Buy $30, 26-Jun-2026)

Facts cited to primary filings or named third-party sources; interpretations labeled in the memo. ROIC.ai, AZI, and FactorsToday are third-party aggregated/estimated data, not primary — every material figure reconciled to the filing.