Rocket Lab reported third-quarter results after the US close on 9 November 2026, and the two numbers that define the investment case point in opposite directions. Revenue was a record USD 267.8m, up 73% year on year and above the company's own USD 250m to USD 265m guidance. The GAAP net loss widened to USD 79.7m, or USD 0.12 a share, against USD 18.3m in the same quarter of 2025 — materially wider than the roughly USD 0.06 to USD 0.07 loss analysts had modelled. The market could not decide which mattered more: the stock opened at USD 54.99 and traded as low as USD 54.96 on 10 November, some 6.6% below Monday's USD 58.86 close, before finishing the session at USD 62.71, up 6.54%.
So, on the evidence of this quarter, is RKLB stock a good investment? What the numbers support is a company whose demand signal is now very hard to argue with: a record backlog of USD 3.57bn, up 52% in a single quarter, with roughly USD 1bn of that from new launch contracts signed in the three months, plus positive GAAP operating cash flow of USD 7.2m — a first for the business at this scale. What would have to go right is narrower and more specific: Neutron, the medium-lift vehicle that converts that backlog into the kind of revenue per launch the valuation implies, has to fly, and it has to fly roughly on the schedule now in place after slipping out of 2026. Everything else in the story is downstream of that one engineering milestone.
What actually happened
The revenue line beat the company's own guidance range, which is a more demanding test than beating consensus. USD 267.8m represents 73% growth against a year-ago quarter of roughly USD 155m, and it came with the company guiding to a further record in the fourth quarter — revenue of USD 270m to USD 290m, with gross margin expansion.
The loss needs unpacking rather than accepting at face value, because its composition is unusually lumpy. The USD 79.7m GAAP net loss includes a USD 27.3m loss on extinguishment of debt, which the quarterly filing attributes to the termination of the bridge loan commitment arranged for the Iridium transaction rather than to anything in the operating business. The year-on-year comparison is distorted further by tax: the company booked a USD 3.5m tax provision this quarter against a USD 41.1m tax benefit in the same quarter of 2025, a swing of nearly USD 45m that flattered the year-ago figure and has nothing to do with trading. On the company's own underlying measure the direction of travel is the opposite of the headline — adjusted EBITDA loss narrowed to USD 18.0m from USD 26.3m. The GAAP loss widened; the underlying loss did not.
It is worth being careful about the per-share figure, because coverage has been inconsistent about its basis. The USD 0.12 loss per share corresponds to the GAAP net loss of USD 79.7m. Some write-ups present it as an adjusted number. It is not, and the distinction matters when it is being compared against an analyst estimate that may itself have been struck on an adjusted basis.
The backlog is the cleanest datapoint in the release. USD 3.57bn, up 52% sequentially, is the largest in the company's history, and the composition matters as much as the total: approximately USD 1bn of new launch contracts in a single quarter indicates that customers are committing to vehicles and slots well ahead of delivery. The company's own release headline led with the backlog growth and the positive GAAP operating cash flow rather than with the revenue record, which is a reasonable reading of where the news value sat.
Why the market reacted the way it did
The intraday round trip on 10 November — down about 6.6% at the morning low, closing up 6.5% — is worth more analytically than either endpoint. It is the market publicly working through the trade-off in a single session.
The morning selling was a straightforward reaction to the headline: a loss close to double what was expected, against a company whose shares have already lost close to half their value over six months. The afternoon reversal looks like the result of the composition of that loss becoming understood, combined with the backlog figure being digested properly. A USD 27.3m charge for terminating an acquisition bridge facility is, in context, a financing event rather than evidence of operating deterioration, and a 52% sequential backlog build is not the kind of number that sits alongside a demand problem.
There is also a sequencing point that explains some of the confusion in the wires. Copy filed while the stock was still near its lows reported the quarter as sending the shares down; that was captured before the reversal and is contradicted by the close, which independent price sources agree finished at USD 62.71, up 6.54%. Anyone reading the 10 November coverage in the wrong order would come away with the opposite impression of how the market judged this quarter.
Crucially, the Neutron delay was not news in this release. The slip of Neutron's inaugural flight from the fourth quarter of 2026 into 2027, after a manufacturing problem with the first-stage tank, was disclosed on 14 October and the shares fell sharply on it at the time. By 9 November it was in the price. What the market was assessing on 10 November was whether the commercial momentum shown in the backlog was strong enough to carry the story through a longer wait for the vehicle.
The Openbook read
The five-factor profile on Rocket Lab splits harder than almost anything else we cover, and averaging the factors into a single score destroys the information. The spread between them is the thesis.
Growth is the strongest reading, and it is now corroborated on two independent axes rather than one. Revenue growth of 73% is a delivered number. The 52% sequential backlog build is a forward number. When those two agree in direction, the growth signal is considerably more reliable than revenue growth alone, which can be flattered by timing. The fourth-quarter guide to a further record extends it to a third data point.
Profitability is weak and should be scored weak, but the mechanism matters for anyone trying to forecast it. More than a third of the quarter's GAAP loss is a single financing charge, and the adjusted EBITDA loss of USD 18.0m on USD 267.8m of revenue — narrower than the USD 26.3m a year earlier — is the better read on the underlying margin, against a company guiding to gross margin expansion. The direction of travel on gross margin is positive; the operating leverage is not there yet because Neutron development is being funded out of the current cost base. Profitability improves mechanically if and when launch cadence rises on a vehicle with better unit economics — which is, again, the same single dependency.
Solvency is the swing factor and the most improved. Positive GAAP operating cash flow of USD 7.2m is the first genuine evidence that the operating business can fund part of itself, and the debt extinguishment charge relates to a terminated acquisition facility rather than to liability accumulation. Set against that, the company is funding a major vehicle development programme through a delay, and the Iridium transaction adds integration cost and complexity. Solvency here is better than it was a quarter ago and still entirely dependent on capital-markets access remaining open.
Momentum is the most misleading factor on this name right now. A 6.5% close looks like positive momentum; a share price down close to half over six months does not. Both are true. Short-horizon momentum has turned up off a low base while the medium-term trend remains firmly negative, and a factor model that reads only one of those will mis-score the stock. The honest reading is that momentum is inflecting, not established.
Reward/Risk is where the split resolves, and it is binary in a way most large-cap names are not. The backlog gives a reasonably well-defined upside path: contracted revenue, improving gross margin, operating cash flow already positive. The downside is not primarily financial — it is schedule. Another Neutron slip does not just defer revenue; it compresses the window in which the current backlog converts at the assumed economics, while the development cost continues to run. Readers can compare how that spread sits against other high-growth, loss-making names on our screener.
The read-across
The most direct implication is for the listed small and mid-cap space complex. Rocket Lab is the sector's reference name for whether public markets will fund a long-dated hardware programme through a visible delay. The 10 November reversal is a modest answer in the affirmative: the market was willing, on the day, to weight a contracted backlog more heavily than a widened loss. Peers with thinner order books and similar cash burn should not read that as a general re-rating — the distinguishing feature here was USD 3.57bn of backlog, and the read-across travels only as far as a comparable demand signal does.
For launch competitors, the Neutron timeline slipping into 2027 extends the window in which medium-lift demand continues to concentrate with the incumbent heavy-lift provider. That is a commercial fact rather than a share-price one, since the dominant player in that segment is private, but it affects how customers schedule missions across 2027 and therefore how much of Rocket Lab's backlog is bookable into near-term quarters.
For Iridium and the satellite-operator cohort, the transaction costs disclosed in this quarter are a reminder that vertical integration between launch and constellation operation is actively being attempted rather than theorised. If it works, it changes the comparable set for both: a launch provider with an owned constellation is a different business from a launch provider.
What to watch next
The fourth-quarter and full-year 2026 results, due in late February or early March 2027, are the first test of the USD 270m to USD 290m revenue guide and of the promised gross margin expansion. Watch whether operating cash flow stays positive — one quarter is a datapoint, two is a trend — and whether the backlog holds above USD 3.5bn once the quarter's new signings are netted against revenue recognised.
Before that, the operational milestone to track is Neutron hardware. The company has been publishing progress on major structural components, including the Stage 1 forward dome, and the sequence that matters runs through stage testing to a static fire. A firm launch date, when it is set, is the single highest-impact disclosure available to this stock in 2027; continued vagueness about the quarter would be its own signal.
Finally, watch the Iridium integration for the first disclosure of what the combined entity's revenue and cost base actually looks like. Until that is quantified, the transaction costs are visible in the loss while the benefits are not visible anywhere.
Discussion
Log in to join the discussion