Fidus Keeps Saying Deal Flow Is Picking Up. The Portfolio Says Otherwise.
A clean credit book and a covered dividend can't hide a flat $1.4B portfolio, $35.5M of fresh ATM equity, and equity realizations sliding into 2027.
FDUS · Earnings Call · 2026-11-06
The portfolio that stopped growing
Fidus Investment Corporation has spent a year telling investors the lower middle market is about to reopen. In the third quarter, the arithmetic finally caught up with the narrative: “Originations in the third quarter totaled $51.8 million ... Proceeds from repayments and realizations were $52.4 million” — Edward H. Ross, Chairman and Chief Executive Officer · 2026-11-06. Assets in minus assets out, roughly zero. The portfolio closed the quarter at $1.4 billion fair value, 102.8% of cost — the same place it has been sitting.
That is not, by itself, a bad quarter. Adjusted net investment income of $0.54 per share comfortably covered the $0.43 base dividend, and NAV per share was $19.47. But it is a telling one. Fidus raised $35.5 million of net proceeds through its ATM program in the same three months that its asset base went nowhere, which means shareholders were diluted into a portfolio that was shrinking on a net basis. Management is explicit about why: it wants dry powder and a fortress balance sheet. “our goal is not to go too far over 1:1. We do care about our investment grade rating.” — Edward H. Ross, Chairman and Chief Executive Officer · 2026-11-06 Net debt to equity is 0.9x; statutory leverage ex-SBA debentures is 0.5x. Liabilities to assets have drifted up to 50.8%, roughly six points higher than a year ago — but the company is still deliberately under-levered relative to its own 0.9x-1.1x target band.
Five quarters of "picking up"
Here is the tension worth sitting with. Across the prior four calls, the same phrase keeps recurring. In November 2025: deal flow is "pretty good." In February 2026: deal flow is "a little more modest" but expected to build through the year. In May 2026, with the Middle East conflict on the tape: “we still have confidence in a pickup in activity, but the pace will be somewhat dependent upon a reduction in the current level of uncertainty in the world today” — Edward H. Ross, Chairman and Chief Executive Officer · 2026-05-08. In August 2026: “we are seeing a higher level of deal flow today than we were 60 days ago” — Edward Ross, Chairman and Chief Executive Officer · 2026-08-07.
And now, in November 2026:
things started to pick up midyear. From a deal flow perspective, we did not have anything closed in Q3, really. Had 1 holdover from a commitment from Q2.
That is the cleanest admission yet that a rising deal flow tide has not reached Fidus's closing book. The explanation offered is a genuine one — buyers and sponsors thawing out on geopolitics — but it has been the explanation for five straight quarters, and the portfolio has yet to show it. Edward Ross also pins the mix to the old economy rather than the new one: business services, niche manufacturing, industrial services, not software. Given that peers reporting in the same window — Main Street Capital on the same day, TSLX two days earlier — are all leaning on the same middle market vocabulary, this is a sector-wide waiting room, not a Fidus-specific thaw.
Zero non-accruals, one $11 million goodbye
The credit story is genuinely clean and it is the strongest part of the quarter. At quarter-end there were no debt investments on non-accrual, 88% of the debt book is first lien, weighted average loan-to-value is about 41%, and the effective yield on debt investments ticked up to 12.6% from 12.5%. Against that, Fidus booked roughly $11 million of net realized losses on Virtex Enterprises — a name that had already been marked down, and which the prior call described as exiting close to its mark. So the loss was reserved, not a surprise. But it is a reminder that in a portfolio of ~86 companies with equity in 85% of them, the tail is where the earnings volatility lives.
The recurring analytical question is artificial intelligence, and Mickey Schleien has now asked a version of it three times in four quarters. Fidus's answer has subtly shifted from defense to offense: the software and tech-enabled book now averages $43 million of revenue and $7 million of EBITDA at 15% margins, and the pitch is that portfolio companies are leveraging AI to improve products rather than being eaten by it. Christopher Nolan probed whether law firms and consulting firms — the archetypal AI-displaceable services — sit inside that business services bucket. They do not, which is the right answer, but the fact that the question now arrives unprompted every quarter is itself the signal.
What shareholders actually own now
Strip out the noise and Fidus is a ~11% annualized distribution story with a deferred capital-gains kicker. Total investment income rose to $45.9 million, up $2.4 million sequentially, on higher average debt balances; the True Environmental prepayment fee lifted fee income $0.9 million while dividend income fell $0.9 million — a roughly neutral wash that Robert Dodd correctly characterized as normal-ish. The $0.11 supplemental piece of the $0.54 declared dividend is literally 100% of the prior quarter's surplus adjusted NII over the $0.43 base, so it flexes with the fee calendar.
The kicker is the equity portfolio: about $159 million marked near 50% above cost, which is real embedded value that the income statement has not yet seen. But management is deliberately not forcing it. “we do see some potential realizations on the equity side in 2027 ... the M&A market is not at that level” — Edward H. Ross, Chairman and Chief Executive Officer · 2026-11-06. Q4 gains, per the call, will be "very limited." So the NAV growth engine is parked.
Which brings it back to valuation. Price to book sits at 0.98x, right at NAV, down about 5% year over year, and the stock is 15.9% below its February 2025 peak of $23.37 while basically flat over the last 90 days. At roughly book, ATM issuance is nearly neutral to NAV but still dilutive to per-share earnings unless the proceeds get deployed. That is the whole bet in one line: effective revenue of $37 million, up 13% year over year, on a four-year climb of 88%, against a price-to-earnings multiple of about 9.4x.
Fidus did what a disciplined lender should do this quarter: protected the book, kept credit clean, covered the dividend, and held leverage below target while waiting. What it has not yet done is prove that the M&A thaw it has promised for five quarters shows up as portfolio growth. Until it does, the shareholder return is the coupon, and the capital gains remain a 2027 promise.