Neptune Insurance Holdings (NP) Q3 2026 earnings review
Neptune earns more per policy, but its plan halves growth
Neptune sold a record amount of flood insurance and kept more of each premium dollar, but its own plan has growth slowing sharply. Revenue rose 34% because the insurers behind its policies paid the company a higher commission on each one it placed. Neptune raised its 2026 target again and set a first target for 2027; by our math both leave growth in the mid-teens.
| Revenue | $59.6M +34% from a year ago |
|---|---|
| Adjusted EBITDA margin | 61.7% 60.2% a year ago |
| Full-year revenue outlook | Raised $5M $204M, about 28% growth |
| First 2027 revenue outlook | $238M 17% above the 2026 plan |
⚖️ Verdict: 🔴 Bearish
The story got worse, because the company has now put a number on the slowdown it had only implied: about 17% growth next year, after the Palomar policy transfer that lifted sales for a year ended in the quarter. The good news: a higher commission rate lifted margins, Neptune repaid $25 million of debt, and the August plan proved cautious.
The question now is whether the 2027 plan is as cautious as this year's turned out to be. Either Neptune keeps lifting its targets through the year, or the Palomar transfer and renewal losses were doing more of the work than the record quarter shows. The fourth quarter, the first without Palomar sales, gives the first reading.
🐂 Bull Case
Insurers Now Pay Neptune A Higher Commission
Neptune does not insure anyone itself. It sells flood policies written by 45 insurance companies and keeps a commission on each premium dollar it places. That commission rate is the main lever on its revenue, because the fees per policy hardly change. Neptune said the quarter benefited from "increased commission income as a percent of written premium".
By our math the gap between what insurers pay Neptune and what Neptune pays the agents who sell its policies widened by almost two percentage points of premium in a year. That is why the adjusted profit margin, which leaves out stock pay and one-offs, rose even though overhead grew faster than sales.
- Commission income: 34.6% of written premium, from 33.4% a year ago and 33.5% last quarter
- Agent commissions: 13.1% of written premium, from 13.6% a year ago; Neptune called this "normalized commission expense as a percent of written premium"
- Agent commissions as a share of revenue: 29.2%, down 2.0 points
- Adjusted EBITDA margin: 61.7%, up 1.5 points
This matters because every extra point of commission rate is almost pure profit: the insurers pay it and no new cost comes with it. The step arrived faster than the gradual gain management described after renewing its two largest programs in June, so the better terms show in full this quarter. The remaining programs renew in October and December.
What to watch: commission income as a share of written premium in the fourth quarter, after the October and December renewals. A rate above 34.6% would mean the June step began a trend rather than a one-off reset.
Neptune Repaid Debt Instead Of Buying Shares
Three months ago Neptune had drawn almost all of its credit line, partly to fund share buybacks, and its balance sheet still shows a stockholders' deficit from last year's preferred dividend. This quarter it bought no shares and used operating cash to repay debt.
- Debt: $215 million, from $240 million at June 30
- Credit line unused: $45 million, from $20 million
- Net leverage: 1.7 times adjusted EBITDA, from 2.1 times
This matters because a company one year past its listing should not borrow to retire stock. Neptune said it will "balance debt reduction with opportunistic share repurchases", so the stated policy is unchanged, but this quarter the cash went to debt and interest expense fell 35% from a year ago.
What to watch: the debt balance and any buyback in the fourth quarter. Another quarter of repayment with no purchases would mean the June buyback belonged to the private-equity share sale rather than to a new habit.
Neptune's August Plan Proved Cautious Again
Neptune gives a full-year revenue target and leaves investors to work out the quarters. By our math the target it set in August pointed to about 20% growth in the second half. The third quarter came in at 34%, so the plan was cautious, and Neptune lifted the full-year target to $204 million.
Without the raise, the August target would have left about $46 million for the fourth quarter, a rise of only 5%, so part of the increase simply recognises what the third quarter delivered. It is the third raise this year, and the moves have been steady rather than shrinking.
- Full-year target by vintage: $186–189 million, then $195 million, then $199 million, now $204 million
- Size of each move: +$7.5 million, +$4 million, +$5 million
A beat against its own plan says the plan was conservative, not that the slowdown in next year's target is wrong, because Neptune sets each plan knowing what it has already sold. What to watch: the fourth quarter against the roughly $51 million the new plan implies. A beat of this quarter's size would mean the 2027 target carries the same caution.
Positives this quarter didn't test
Four drivers of the standing case got no new numbers, because Neptune reports them on the call or in the quarterly filing rather than in the release. Each is listed with the reading that would move it.
- AI in engineering: Neptune said "AI is allowing our engineering team to accomplish substantially more without a corresponding increase in headcount"; it released 14 versions of Triton, its pricing and underwriting software, against 23 last quarter, and gave no output figure.
- Agent activation: more than 55,000 agents had accounts at June; the release gives no update, and the second phase of Atlas+, the AI sales assistant that management said would arrive this quarter, is not mentioned.
- Product breadth: 11 products, from 9 at year-end, with earthquake cover the newest; premium by product is not broken out.
- Capacity: 45 insurers back the programs, unchanged from June, with coverage above $150 billion from $122 billion at year-end; the October and December renewals will show the terms.
🐻 Bear Case
Neptune's Own Plan Has Growth Halving Next Year
Neptune says it sees "significant opportunity ahead" and talks of compounding growth. Its first plan for 2027 says revenue of $238 million, or 17% growth. This year's plan works out to about 28%. By our math the fourth quarter already runs at the lower pace: the raised 2026 target leaves about $51 million, 16% above a year ago.
Two things explain part of the step-down. The Palomar policies that Neptune has been moving onto its own book since October 2025 finished transferring, and Neptune said the extra sales "concluded in Q3". The fourth quarter also compares against last autumn's 43-day federal shutdown, when the government flood program could not write new policies and agents turned to Neptune.
But those two effects do not explain all of it, because premium in force, the yearly premium on every active policy, grew almost as fast as revenue and is the base for next year's renewals. Either Neptune has set a cautious plan, as it did twice this year, or growth without Palomar is well below the headline, and it has never disclosed that rate.
What to watch: fourth-quarter revenue growth, the first period with no Palomar sales in it. A rate near the 16% the plan implies would mean the slowdown is in the business, not in the plan.
Fewer Renewal Dollars Return Each Year
Neptune's revenue depends on policyholders renewing, because each policy lasts a year and Triton re-prices it at renewal. Revenue retention, the share of last year's policy revenue that comes back on renewal, was 89% over the last twelve months. It was 92% in 2025 and has edged down each quarter since.
The two measures behind it, how many policyholders accept a renewal offer and how much premium those renewals carry, appear only in the quarterly filing. Last quarter premium retention fell almost 8 points from a year earlier because Neptune offered smaller price increases, so the in-force book compounds more slowly than its size suggests.
What to watch: premium retention in the filing for this quarter. A rate still near 92% would mean the smaller renewal increases are a lasting choice, and new sales must do more of the work each year.
Overhead Is Growing Twice As Fast As Sales
Neptune lists "continued emphasis on operating leverage" among the reasons profit rose. The expense lines say otherwise. General and administrative cost, the overhead of running the company, rose 82% from a year ago, more than twice the pace of revenue, and employee pay rose 30%.
- General and administrative: 6.5% of revenue, from 4.8%, up 1.7 points
- Employee pay: 3.6% of revenue, about unchanged
- Stock pay, left out of adjusted profit: 11.8% of revenue
Neptune blamed "increased expenses related to public company costs and technology development", which is fair one year past a listing. But by our math the overhead growth took about 1.7 points off the margin, so the entire gain came from the commission rate, and stock pay is a real cost that the adjusted figures leave out.
What to watch: general and administrative cost in the fourth quarter. A figure still rising faster than revenue would mean the public-company step-up is not a one-year effect and the 61% margin target rests on the commission rate alone.
Risks this quarter didn't answer
Five standing risks got no new reading, because the release carries no call commentary and the quarterly filing is not yet out. Each is listed with the disclosure that would move it.
- Dependence on the insurers that carry the risk: the lifetime loss ratio still reads 19.5% as of June 30; a quiet hurricane season means it faced no test, and the October and December program renewals will show the terms.
- State scrutiny of AI pricing: no comment on the Texas insurance regulator's bulletin or on who oversees Atlas+; the call transcript is where it would appear.
- FEMA flood-map modernisation, the largest lever in the company's own story: no update; any timetable from the agency would move the case.
- Housing market: Neptune again cited the slow real estate market as a drag on selling; home-sales data, not Neptune's, settle it.
- New private-flood competitors: not mentioned; a disclosed loss of agents or policies would be the reading.
👓 Other Themes
A Quiet Hurricane Season Slowed Selling
Flood insurance sells best when storms are in the news. Neptune said the quarter carried a "residual slowdown as a result of below-average 2026 Atlantic hurricane season" and an "ongoing slow real estate market reducing selling opportunities". Working the other way, agents moved business to Neptune "driven in part by NFIP uncertainty", the doubt over the federal flood program's future.
💲 Other KPIs
Up 31.7% from a year ago to a record, and up $31 million from June against $24 million in the same quarter last year. Premium in force is the yearly premium on all active policies, so it is the base for next year's renewal commissions. Growth has held near 32% for three readings while policies in force grew 29%, so the average policy carries slightly more premium.
Up 78% from $38.9 million a year earlier, when the figure absorbed listing costs. By our math the quarter alone brought in about $25.9 million, more than net income, because stock pay is a charge that costs no cash. That cash funded the quarter's debt repayment with room to spare.
Rose from 25.6% a year ago. Tax expense grew 89% on pretax income up 60%, so a growing share of the profit gain stops at the tax line. Neptune uses a 27.6% rate for its adjusted figures and says the GAAP rate carries items that are not deductible or get special treatment.
Up 0.1% from last quarter. Restricted stock vesting added 1.4 million shares at quarter end, with 0.5 million withheld to cover employees' taxes. Neptune bought back no shares. A further 3.8 million unvested units remain, so dilution continues at a slow pace unless buybacks resume.
🔮 Guidance
Raised. The plan now has 2026 revenue of $204 million. That is about 28% growth, and by our math it leaves about $51 million for the fourth quarter. That is 16% above a year ago, well below this quarter's pace, because the fourth quarter compares against last autumn's shutdown and the Palomar sales have ended. Third raise this year: +$7.5M, +$4M, +$5M.
Raised. The plan now has the adjusted EBITDA margin, profit before interest, tax, amortisation and stock pay as a share of revenue, above 61% for the year. By our math a 61% year needs a fourth-quarter margin of about 62.4%, higher than any quarter so far, so the plan assumes the commission-rate gain holds while costs stay flat on lower seasonal revenue.
New. The first target for 2027 is $238 million, 17% above this year's plan. That is roughly half this year's growth rate. Neptune has lifted its annual target three times in the past year, and if the pattern holds, by our math 2027 revenue would end near $259 million. But the starting point for 2026 was 25% growth and the starting point for 2027 is 17%.
New. The plan keeps the adjusted EBITDA margin at 61% for 2027, the same level as this year's floor, so Neptune expects profit to grow at the pace of revenue rather than faster. By our math that is about $145 million of adjusted EBITDA. The plan assumes no further gain from the commission rate and no relief on the public-company overhead.
❓ Key Questions
What did Palomar add, and what is growth without it?
The transferred Palomar policies lifted sales for a year and ended this quarter. Neptune has never disclosed their premium or revenue, so nobody outside the company can tell whether the 2027 plan is cautious or simply organic.
Why did the commission rate jump a full point?
Commission income reached 34.6% of written premium from 33.5% last quarter, far more than the gradual gains management described. Readers need to know whether this is program terms, product mix or a one-off true-up before counting on it.
What are policy and premium retention this quarter?
Revenue retention is down three points from 2025, and premium retention fell almost 8 points last quarter. The filing will show whether smaller renewal increases are still the cause and whether policy retention held at 86%.
Does the 2027 plan assume any change at FEMA or the NFIP?
Management has said earlier plans assume nothing from the federal flood program. If the 2027 target also assumes nothing, flood-map modernisation and NFIP uncertainty are upside; if it assumes something, the 17% is less cautious than it looks.
How many agents use Atlas+, and what did phase two deliver?
The second phase of the AI sales assistant was due this quarter. The release gives no agent-account count, no usage figure and no binding rate, so the distribution leg of the case is unreadable until the call transcript or filing arrives.