Q2 owns the bank's front door, and its buybacks now outrun stock pay

1. Summary

• Q3 revenue was $225.77M, up 11.93% from $201.70M, with subscription revenue up 13.88% to $188.16M ^[1].
• GAAP gross margin reached 59.7%, up from 54.0%, and operating income tripled to $33.96M ^[1].
• Trailing owner free cash flow is about $125.77M, up from $86.42M in fiscal 2025 and a loss of $39.54M in 2023, our calculation ^[1][3].
• Q2 repaid $304.0M of convertible notes and bought back $158.9M of stock this year, and now carries no debt ^[1][2].
• Enterprise value is about $3.92B, 31.14x owner free cash flow and 18.94x free cash flow, our calculation ^[1][2].
• Our view: watch. The base case sits at the top of the watch band, but the gate passes only two of four tests.

2. Thesis

Q2 sells the digital banking layer that sits between a bank's core ledger and its account holders. Every login, transfer, bill payment and fraud check passes through it.

Contracts average more than five years and existing customers spent 113% of what they spent the year before in 2025 ^[3]. If agents begin to move money for account holders, banks will want a governed, audited gateway they already trust, and Q2 is the obvious candidate for its 500 digital banking customer accounts ^[3].

Gate score: 2 of 4 pass, two partial. That is a mixed gate.

What invalidates it: registered user growth stalls below 5% a year while net revenue retention falls under 108%, which would mean the front door is losing traffic to channels Q2 does not control ^[2][3].

Product

Q2 sells five product families from one cloud platform, and the filings split the money three ways: subscription, transactional, and services and other ^[3].

Digital banking platform. This is the core product: online and mobile banking for consumers, small businesses and commercial clients on one code base, including onboarding, account views, transfers, bill pay and treasury tools for business clients ^[3]. At year end Q2 had about 500 digital banking customer accounts, with 450 using consumer functions, 300 commercial and 50 small business ^[3].

Pricing is set by the number of solutions a bank buys and the number of registered users or commercial account holders, with extra fees when users and transactions exceed what the subscription covers ^[2]. Registered users reached 28.5M on Sep 30, up 5.56% from 27.0M ^[2].

Revenue grows when a bank adds modules, adds users, or its users transact more ^[2].

Risk and fraud. About 850 customer accounts, the largest family by count ^[3]. These tools can ride with the platform or be sold alone, and they are billed as subscriptions ^[2].

3. Operations

Segments. Q2 discloses one operating segment, and all revenue comes from selling, implementing and supporting its solutions ^[2]. That means no segment margins; we read the business through the cost lines.

Costs. Cost of revenue was $90.93M in Q3 against $92.86M a year ago, even as revenue rose $24.07M ^[1]. That is the payoff from moving the platform out of third-party data centers into the public cloud, which the 10-K lists as a key operating risk ^[3].

Non-GAAP gross margin was 62.6%, up from 57.9% ^[1].

Sales efficiency. Trailing sales and marketing was about $102.09M, 11.73% of revenue, our calculation ^[1][3]. That is low for software, and it was down 7.06% in Q3 to $24.54M while bookings were described as solid ^[1].

A five-year contract with a bank sells slowly but stays sold, so Q2 spends less to keep growth near 12%. Each dollar of sales and marketing bought about $0.98 of new total ARR over the last year, our calculation from ARR growth of $100.2M ^[2].

4. Valuation

On the Oct 29 close and 61.78M cover-page shares, market value is about $4.03B. Cash and equivalents of $116.31M, no debt, gives enterprise value of about $3.92B, our calculation ^[1][2].

The $6.68M of investments is left out by method.

Trailing 12 month inputs, our calculation from the 10-K and the Q3 release ^[1][3]:

• Revenue of about $870.26M, so enterprise value is 4.50x revenue.
• Operating cash flow of $245.35M less $38.59M of capex, including capitalised software, gives free cash flow of about $206.76M, a 23.76% margin, or 18.94x.
• Stock pay of about $80.99M, 9.31% of revenue.
• Owner free cash flow of about $125.77M, 14.45% of revenue, so enterprise value is 31.14x owner free cash flow, a 3.21% yield.

Deferred revenue rose $43.34M in nine months, which flatters operating cash flow mid-year ^[2]. The fourth quarter usually reverses part of it, so we used the trailing figure rather than annualising the nine months.

The price implies owner free cash flow growth of about 10.79% a year for ten years at a 9% discount rate and 3% terminal growth, 8.22% at 8% and 13.11% at 10%, our calculation.

Five-year view, our calculations. Growth rates, margins and exit multiples are our assumptions, not company guidance. Year 5 owner FCF is year 5 revenue times FCF margin less stock pay; implied EV is that times the exit multiple.

5. Financial update

Growth. Q3 revenue rose 11.93% and 3% from Q2 ^[1]. Q2 raised full-year revenue guidance to $888.0M to $892.0M, 12% growth, and guided Q4 to $226.0M to $230.0M, 9% to 10% ^[1].

Subscription ARR grew 13.5% and total ARR 11.28%, our calculation ^[2].

Retention. Net revenue retention was 113% in 2025, up from 109% in 2024; subscription retention was 115% ^[2]. Q2 discloses retention only annually.

Margins. Adjusted EBITDA was $67.76M, a record 30.0% margin, up from 24.2% ^[1]. GAAP net income was $33.22M, 14.71% of revenue, helped by a tax charge of only $1.11M on $34.33M of pretax income ^[1].

Trailing GAAP net income is about $110.16M, 12.66% of revenue, our calculation ^[1][3]. Other income fell to $0.37M from $4.20M as the cash pile shrank ^[1].

Cash.

6. Risks

• Agents bypass the login. Q2 prices by registered user and transaction ^[2]. If account holders' agents reach banks through APIs instead of logging in, per-user pricing loses its anchor. Q2 does not disclose any agent-related revenue.
• Bank consolidation. Mergers shrink customer accounts, and the 10-Q says merger activity can cut accounts and users ^[2]. Q3 included a Tier 1 bank win that followed a merger ^[1], which shows the risk cuts both ways.
• Core vendor bundling. Fiserv, Jack Henry and FIS own the ledger and can bundle digital banking at low marginal price ^[3].
• Growth deceleration. The Q4 guide of 9% to 10% is below the 12% full-year pace ^[1]. If growth settles near 9%, our base case falls to about 1.37x, our calculation.
• Leadership transition. A co-CEO structure with the co-CEO also serving as CFO concentrates two jobs in one person until a new CFO is hired ^[1][4].
• Third-party superintelligence vendors. Q2 relies on a limited number of outside model vendors for its new products, with pricing and access risk ^[1]

7. Catalysts and conclusion

• Q4 2026 results, expected February 2027, unconfirmed date. The test is whether Q4 beats the 9% to 10% guide and what 2027 guidance implies.
• 2026 net revenue retention, in the fiscal 2026 10-K, expected February 2027, unconfirmed. A figure at or above 113% supports the base case.
• General availability of User Activity Monitoring, Q2 Code and Q2 Assistant, which Q2 says is coming in the next few months, unconfirmed timing ^[1].
• CFO appointment, timing unconfirmed ^[1].

Stock position: watch, do not add yet. Q2 is a better business than it was two years ago. Owner cash went from a loss to about $125.77M, gross margin rose almost six points in a year, stock pay fell below 10% of revenue, and the debt is gone ^[1][3].

At 31.14x owner free cash flow the price already asks for about 11% growth a year, our calculation. Our base case is worth 1.5x enterprise value, but a mixed gate holds the call at watch.

The trigger to upgrade is evidence that Q2 prices on outcomes, for example fraud losses prevented through User Activity Monitoring, or that agent traffic is being routed through Innovation Studio. Either would turn a partial gate test into a pass.

What it means for the future of software. Superintelligence agents will need a governed way into regulated money. Banks will not hand an agent the core ledger; they will route it through the layer that already handles identity, entitlements and fraud.