Paymentus Is Winning the Argument and Losing the Multiple

The case that gets stronger while the stock gets cheaper

Paymentus put up its third consecutive beat-and-raise of 2026 and, for the first time, attached a number to the ambition it has been gesturing at for four quarters. Revenue of $372.2 million (+19.8%), contribution profit of $118.5 million (+20.6%) and adjusted EBITDA of $49.7 million (+38.6%) were all records; the EBITDA line equates to a 41.9% margin on contribution profit, up 540bp. But the sentence that mattered was the one that came with a disclaimer welded to it.

if I were to model the business out and simply add 4 components together... it is quite possible that in the next 18 to 24 months from today, we have the potential to reach approximately $2 billion of revenue and backlog combined.

Note the careful accounting: revenue and backlog combined, not revenue. CFO Sanjay Kalra spent his first answer of the Q&A dismantling any reading of it as a forecast — “this is in no shape or fashion any guidance of any particular year” — Sanjay Kalra, CFO · 2026-11-02 — which is precisely the guidance methodology this management team has trained the Street on for three years. That discipline is not new. What is new is that the CEO is now willing to do public arithmetic at all.

The cost center becomes a P&L line

The genuinely new fact in this quarter is buried in a bookings disclosure: Q3 bookings included the company's first sizable interchange monetization agreement. For years Paymentus has framed interchange as a vendor expense to be attacked rather than a fee to be earned. Dushyant Sharma made the framing explicit: “we have always thought of interchange as a cost center not as a toll after on a highway” — Dushyant Sharma, CEO · 2026-11-02. Converting that cost into a revenue stream is the mechanism by which earnings could compound faster than revenue — the same logic he has repeated since at least the May call, when he described the long-term goal as "convert interchange expense into interchange revenue." The difference now is a signed contract, not a slide. Engineers of this story should note the second leg: BillWallet doubled to 200,000 users, sits in roughly a third of U.S. ZIP codes, and — per the prior quarter's Q&A — converted at high rates with zero marketing spend. That is a distribution proof point, not a product announcement.

Yet both remain "a couple of years out," in the CEO's own words. Which is why the interview-worthy tension in this call is not the products — Agentic commerce has been on this transcript since 2025 — but the cost of building them.

Q4 EBITDA is guided down, and the Street noticed

Full-year adjusted EBITDA guidance rose to $186–191 million, implying ~37% growth and a Rule of 40 score of 61–62. But the Q4 EBITDA range of $45–50 million is sequentially lower than Q3's $49.7 million, breaking a pattern of uninterrupted dollar progression. Tien-Tsin Huang asked directly whether something outsized was being spent. Kalra's answer was the most honest thing on the call: “Think we should leave some room to actually spend more and convert our pipeline to bookings for outer years.” — Sanjay Kalra, CFO · 2026-11-02 Read that against the prior quarter's language, where he told investors to “look at our guidance philosophy over time and see the results we have delivered” — Sanjay Kalra, Chief Financial Officer · 2026-05-04 rather than over-read one quarter. The pattern is consistent — guidance is a floor, execution is the product — but a sequentially declining EBITDA dollar figure is a different animal than a conservative revenue range, and it landed on a stock with no room for ambiguity.

The mix mechanics explain some of it. Gross margin has drifted down as large billers and card-heavy payment mixes scale — 26.1%, versus 32.6% at the 2020 peak. Contribution margin, however, ticked up to 31.8%, and revenue per transaction held at $1.70 with contribution profit per transaction flat at $0.54. That is the tell: this company is not buying growth with price. Kalra has said for two quarters that customer mix is the swing factor, and the mix is now being managed, not absorbed.

Why the multiple won't cooperate

Here is the actual story. Over the last 90 trading days Paymentus shares returned +49.6% — a violent recovery off a drawdown that had taken the stock 47% lower over a 20-week stretch. But the tape still shows a 21.4% drawdown from the August 4 peak of $44.60, and the valuation math has collapsed. Price to revenue sits at 2.2x, down 43% year over year and down 80% from the 11.1x of 2021. Price to free cash flow ex-SBC is 22.4x — down 44% on the year. Meanwhile revenue compounded 374% over six years.

So the market is repricing a compounder as a cyclical. Part of that is the broad tape: the global 90-day decliners are almost entirely AI infrastructure — data center capacity, high bandwidth memory, co-packaged optics, AI cloud services — and Paymentus is up nearly 50% against that current. Being long an AI-adjacent story in a quarter when AI infrastructure de-rated by 30–45% is its own kind of statement. What the market has not yet paid for is the possibility that disintermediation risk runs the other way: Sharma argued, unusually, that consumer AI agents threaten billers, and that a wallet preserving biller identity is the hedge — not the victim. Adyen called the same thing "the right rails for Agentic Commerce" this week; Carvana and GoDaddy both leaned on agentic language. Paymentus is the only one framing it as a defensive moat rather than an offensive land grab. That is either the sharpest insight on the call or the most elaborate justification for R&D spend of $15 million a quarter.

The bottom line

The operating story is intact and accelerating: guidance raised, backlog described as diversified across utilities, insurance, government, property management, telecom, banking and education, DSO down ~13% year over year, and $409.6 million of cash with no leverage worth mentioning. What changed this quarter is that management converted two years of conceptual talk — interchange, wallets — into one signed contract and one user statistic. What did not change is the central friction: “we are fully aware and we are expecting a dogfight in the retail. Commoditized retail market” — Dushyant Sharma, CEO · 2026-11-02 — a CEO volunteering a warning about the fight ahead while his stock trades at 2.2x sales and 22x free cash flow. If the $2 billion arithmetic is even half right, the multiple is the mispricing, not the business. If the interchange deal is a one-off and Q4's EBITDA dip extends into 2027, then two years of personal agent rhetoric will have been an expensive way to defend a utility-billing franchise. That is the bet the tape is currently refusing to make.