ネプチューン・インシュアランス・ホールディングス (NP) 2026年第3四半期決算説明会：売上高34%増、業績予想を引き上げ

要点

• 2026年第3四半期の売上高は前年同期比34%増の5,960万ドル、調整後EBITDAは38%増の3,680万ドルとなりました。
• 売上高の伸びが人員の伸びを引き続き上回ったため、調整後EBITDAマージンは前年同期比147ベーシスポイント上昇し、約62%に達しました。
• 有効保険料は32%増の4億5,000万ドル、有効契約件数は29%増の33万6,000件超となりました。新規事業保険料は前年同期比51%増加しました。
• Neptuneは2026年通期の売上高見通しを2億400万ドルへと4回目の上方修正を行い、調整後EBITDAマージンは61%を上回ると見込んでいます。
• 2027年の初期見通しでは、住宅市場の回復、例年以上のハリケーンシーズン、買収、新製品からの大幅な寄与を前提とせず、売上高約2億3,800万ドル、調整後EBITDAマージン約61%を見込んでいます。
• 同社によると、当四半期中にAtlas+は7,000名の代理店に利用されました。Atlas+でのやり取りを伴う見積もりは、やり取りのない見積もりに比べて3.5倍の確率で成約に至りました。

主要財務データ

事業および営業業績

Neptuneは、販売ネットワークの拡大、代理店の生産性向上、ならびに継続的な製品・技術開発が成長要因であるとしています。6万5,000名以上の保険専門職が認証済みNeptuneアカウントを開設しており、アクティブな専門職数は第2四半期の3万9,000名から第3四半期には4万3,000名超へと増加しました。

当四半期中、Atlas+は7万5,000件以上の代理店メッセージを処理しました。経営陣によると、同プラットフォームは代理店が水害リスク（暴露）を把握し、引受に関する質問に回答したうえで、更新、契約変更（裏書）、失効した見積もりへの追跡などの対応事項を特定するのに役立っています。

社内の生産性も向上しました。Neptuneのエンジニアリングチームは、人員数がほぼ横ばいであるにもかかわらず、前年同期の2.5倍の成果を達成しました。カスタマーサービス部門では、契約3万件あたり担当者1名という平均人員体制となり、前年同期比で50%の生産性向上を記録しました。

当四半期末時点の従業員数は59名で、前年同期の60名から減少しました。経営陣によると、Neptuneは過去1年間で8,200万ドルの営業キャッシュフローを創出し、そのうち4,900万ドルをレバレッジ削減に、約3,200万ドルを自社株買いに充当しました。

Neptuneはより有利な条件で別の引受キャパシティプログラムを更新しました。12月1日以降、運用中のキャパシティプログラムにはコンティンジェント・コミッション（追加手数料）が含まれなくなります。経営陣は、現在の手数料収入率は短期的には妥当な巡航速度（ランレート）であり、今後の契約交渉は引受実績や保険料の伸びに一部依存すると述べました。

「Indemnity Earthquake（地震補償）」製品は、カリフォルニア州でベータテスト段階にあります。Neptuneは、クロスセル、API連携、直販（D2C）機能を含め、第4四半期に正式ローンチする予定です。経営陣は、水害保険から地震保険へのクロスセルにより成立した初の地震保険契約が決算発表当日に成約したと述べました。

クリスティアン・メレジ（Cristian Melej）氏が1月にNeptuneに入社し、2026年年次報告書（Form 10-K）の提出後にCFOに就任する予定です。ジェームズ・スタイナー（James Steiner）氏は引き続きCOOおよび取締役にとどまります。

経営陣による業績見通し

経営陣は、2027年の見通しは継続的な契約数の増加、販売網の拡大、堅調な新規事業の売上および維持率に裏打ちされていると述べました。地震保険製品の大幅な伸びや、全米洪水保険制度（NFIP）の変更には依存していません。

資本配分の優先順位は引き続き、自律的成長（オーガニック成長）、機動的な自社株買い、債務削減です。経営陣は固定的なレバレッジ目標を設けておらず、現在の1.7倍という比率は潜在的な買収に対応できる柔軟性を提供するとしていますが、これまでに検討された案件でNeptuneの選定基準を満たしたものは存在しないと説明しました。

リスクおよび注視事項

• 住宅市場は依然として低迷しており、2年連続で大型ハリケーンが発生しなかったため、気象要因による水害保険の需要は限定的なものとなっています。
• 経営陣は、強力なエルニーニョ現象が需要やポートフォリオの引受業績に重大な影響を与えることはないと予想していますが、水害リスクに対する意識の高まりが加入を後押しする可能性はあります。
• AIエージェントを介した直接的な保険購入は依然として軽微です。経営陣は、従来の代理店販売からのシフトは緩やかに進むと予想しています。
• 売上維持率は約89%でした。経営陣は、価格設定を契約維持、顧客生涯価値（LTV）、新規契約件数の間のトレードオフ関係にあると説明しました。
• 契約1件あたりの保険料は前年同期比で2%増加しましたが、経営陣は、高価格帯の強制加入水害地域と低価格帯のXゾーン（低・中リスク地域）の契約比率の変化が平均値に影響を与える可能性があると指摘しました。
• Neptuneは年間を通じて同様のコスト構造を維持する一方、水害保険の需要は冬季に低下するため、調整後EBITDAマージンには季節性があります。

アナリスト質疑応答の要点

AIと販売コスト： 経営陣は、販売手数料がNeptuneの最大の費用であり、代理店向けには年間平均約13%となっていると述べました。AIツールを通じた直接購入への長期的なシフトにより、このコストは大幅に減少する可能性がありますが、同社は移行が緩やかに進むと見込んでいます。短期的には、Neptuneは代理店の代替ではなく、代理店の生産性向上のためにAIを活用することに注力しています。

競争上の優位性： Neptuneは引受データ、4,000万件近くに及ぶ過去の見積りデータ、保険金支払実績、全国対応、引受能力（キャパシティ）関係、API連携を模倣困難な強みとして挙げました。経営陣はさらに、一部の大手代理店がNeptuneを優遇し、提供する競合の民間水害保険製品の数を削減したと付け加えました。

市場拡大： 経営陣の推計によると、Neptuneの契約のうち全米洪水保険制度（NFIP）からの切り替えはわずか約20%にとどまり、残りは既存の住宅所有者または住宅購入者による新規補償です。また、同社は新規発行された単体水害保険契約の20%〜30%を獲得していると推計しています。

Palomarの寄与： Palomarとの契約を除外した場合、第3四半期の売上高成長率は、報告された34%増に対して30%増となっていたはずです。

マージン拡大の可能性： 経営陣は、AIによる顧客サービスや業務効率の改善に伴い、時間の経過とともにマージンが上昇し続ける可能性があると述べました。ただし、現在のガイダンスを上回るマージン目標は示されませんでした。

新製品および買収： 地震保険は、Neptuneが次に本格投入する製品となります。同社は買収に対して厳選的な姿勢を維持しつつ、自社のデータやテクノロジーに沿ったその他の気候関連リスクの評価も行っています。

決算説明会トランスクリプト全文

決算説明会の完全なトランスクリプト

経営陣による説明

Operator

Hello, everyone. Thank you for joining us, and welcome to Neptune Insurance Holdings Third Quarter Earnings Call. [Operator Instructions] I will now hand the call over to Jon Carlon, Director of Corporate Development. Please go ahead.

Jonathan Carlon

Thank you, and good afternoon. With me here today is Trevor Burgess, Chairman and CEO; Matt Duffy, President and Chief Risk Officer; and Jim Steiner, CFO and COO. Before we begin, I'd like to remind everyone that today's discussion will include forward-looking statements, including, among others, statements about our expectations for our future financial performance, growth opportunities, business strategy, market trends and capital allocation plans.

These statements are based on our current views and assumptions and are subject to risks and uncertainties that could cause actual results to differ materially. We direct you to our recent SEC filings for a full description of these risks. We undertake no obligation to update any forward-looking statements, whether as a result of new information, future events or otherwise, except as required by law.

We will also reference certain non-GAAP financial measures. These measures should be considered only as supplements to their comparable GAAP measures. Additional information, including reconciliations of the non-GAAP measures to their most comparable GAAP measures can be found in our earnings release at investors.neptuneflood.com and in our current report on Form 8-K that was publicly filed with the SEC on October 20, 2026.

Now I'd like to turn the call over to Matt.

Matthew Duffy

Thanks, Jon. It's been a remarkable year for Neptune. And as we mark the first anniversary of our IPO, I want to spend some time today discussing not just the results we're delivering, but what's driving them and what they tell us about the future of the business. One of the things I've come to appreciate about Neptune is that the real strength of our business isn't in any one product, technology or distribution relationship.

It's how those elements work together. Our technology allows us to make better underwriting decisions and create better experiences for our distribution partners. That attracts more agents, more business, more capacity and helps us close the protection gap. And as the business grows, we gain greater insight, improve our technology and become more efficient. Each of those advantages reinforces the others.

What's particularly exciting today is that we're seeing the benefits of that model become increasingly apparent in our financial and operating results. We're growing revenue substantially faster than our team size. We're making our distribution partners more productive. We're attracting greater demand from risk-taking partners, and we're doing all of that at a time when the broader environment hasn't been helpful. That's important because we're not building a business that depends on favorable market conditions to succeed. The fact that it's a soft insurance market and a slow housing market hasn't impeded our ability to execute or grow. We're building a business with the capabilities to consistently take share, expand the market and compound our advantages over time.

I want to walk through the evidence of that today, including what we're seeing from our investments in AI, the evolution of our distribution network, the strength of our capacity relationships and the opportunities we see ahead. But first, let me turn to our third quarter results.

The third quarter was another record for Neptune. Revenue grew 34% year-over-year to $59.6 million. Adjusted EBITDA grew 38% to $36.8 million at a 62% margin, an improvement of 147 basis points from a year ago.

Premium in force reached $450 million, up 32%, and we now have more than 336,000 policies in force, up 29%. October 1 marked the anniversary of our first day of trading on the New York Stock Exchange, and it's worth reflecting on how far we've come in that short period. Our trailing 12-month revenue is now 44% higher than at June 30, 2025, the figure reported in our S-1.

We've achieved that growth while maintaining exceptional profitability with a Rule of 40 of 94. And the growth is broad-based. It's coming from the expansion of our distribution network, increasing productivity among our existing agents and the continued development of our products and technology.

As we get into operations, let me start with distribution because I know the impact of AI on insurance purchasing is top of mind for many investors. Our view is straightforward. Greater price transparency and easier access to insurance are opportunities for Neptune, not threats. There are more than 100 million buildings in the United States and approximately 25 million face meaningful flood risk, yet there are only around 4 million flood insurance policies.

The biggest obstacle to selling flood insurance isn't price discovery. It's that millions of property owners don't know they need it or mistakenly believe they're already covered. AI has the potential to change that by helping consumers understand their exposure and identify coverage gaps, and Neptune is exceptionally well positioned to serve that demand. Whether a consumer purchases through an insurance agent, directly through Neptuneflood.com or eventually through a personal AI assistant, our technology can support the transaction.

In the near term, we expect AI to make our existing distribution partners substantially more productive. Over time, as more consumers transact directly, there's also an opportunity to lower our distribution costs, which today represent our largest expense. We're already seeing tangible evidence of the opportunity through Atlas+. During the third quarter, 7,000 agents used Atlas+, exchanging more than 75,000 messages with the platform. And perhaps most importantly, quotes involving Atlas+ are binding at 3.5x the rate of quotes without an Atlas+ interaction. That is a powerful indication of the value we're delivering.

Let me give you an example of what that looks like in practice. We recently held a call with a large group of insurance agents. And during the discussion, one agent told us they just sold their first-ever flood insurance policy. They'd come to Neptune, run a quote and started a conversation with Atlas+. The tool gave them a sales script, helped them understand the property's flood exposure and provided the backgrounds they needed to explain the coverage to the customer.

They told us that conversation gave them the confidence to present the policy and ultimately close the sale. I love that story because it gets to the heart of what we're trying to accomplish. There are tens of thousands of insurance agents across the country who have customers exposed to flood risk, but who don't routinely sell flood insurance. Often it's because they don't have the expertise or confidence to have that conversation and Atlas+ changes that. It puts flood insurance expertise in the hands of every agent regardless of their experience.

And we recently heard from Garrett Mitchell with Rocket Flood, a HUB International company who said, and I quote, Atlas+ is making a real difference in our business. The chatbot on Neptune's quote page is like having an underwriter alongside us through the entire sales process. And with the new Atlas+ cards in just 3 weeks, we found dozens of endorsements and renewals, generating tens of thousands of dollars of premium.

Garrett's comments show that we're not simply making our partners more efficient. We're also helping drive demand for those partners so they can grow their business. More than 65,000 insurance professionals have now created verified Neptune accounts and over 43,000 of them were active during the quarter compared to 39,000 in Q2. We're not just adding distribution partners. We're making each relationship more productive and giving agents more reasons to do business with Neptune.

The same productivity gains are happening inside our organization. During Q3, our engineering team delivered 2.5x the amount of work it delivered in the same quarter last year with broadly the same number of people. And our service organization operated at an average of one human representative for every 30,000 policies in force, a 50% productivity improvement year-over-year while improving response times and customer satisfaction. These are meaningful improvements, and they demonstrate how we're growing the capabilities of the business considerably faster than its headcount, all while continuing to invest in new products, technology and underlying infrastructure.

I've spent a considerable amount of time with capacity partners over the last several months, and the quality of those conversations reflects how far the business has come. These are sophisticated counterparties with their own models, their own views of risk and many choices about where to deploy capital. They understand our portfolio, they monitor its performance and increasingly, they want to participate in its growth.

During the quarter, we renewed another capacity program on improved terms. And with that renewal, effective December 1, none of our active capacity programs will include contingent commissions. When existing partners want to grow alongside Neptune and new partners want to access the business, that's meaningful external validation of what we've built.

Let me turn to the external environment and the broader opportunity ahead. The housing market remains weak, and we've now gone through 2 consecutive years without a major hurricane, creating a substantial demand event for flood insurance. Both of those factors have historically supported new business demand and their absence has created a headwind. Despite that, our premium in force has grown at 32% year-over-year, and our new business premium during Q3 was up 51% year-over-year. That's an important indicator of the underlying strength of the business.

We're growing our distribution, taking market share and expanding the insured base without relying on an improvement in those external conditions. A recovery in the housing activity would certainly be beneficial, but it isn't required to deliver on our growth expectations. We're also monitoring the potential impact of a strong El Nino this winter. We don't expect it to have a material impact on either overall demand or the underwriting performance of our portfolio, but increased awareness of flood risk is good for the industry.

Significant weather events create opportunities to remind property owners of their exposure and of the importance of having flood insurance. We'll continue working closely with our fantastic distribution partners, particularly on the West Coast and in Hawaii, to translate that awareness into greater insurance adoption and helping to close the protection gap.

We are often asked about potential changes at FEMA around the NFIP. While we have no additional information and don't expect significant changes in the immediate term, our latest research highlights the opportunity for growth, showing that nearly half of NFIP policies are still priced below their full risk level, representing approximately $2.6 billion in annual subsidies that will continue to burn off in the coming years.

Meaningful reform would accelerate that opportunity, but it isn't a prerequisite for our success. We're also continuing to expand our addressable market through new products.

Our Indemnity Earthquake product is progressing through beta testing, and we anticipate a formal launch in Q4.

Let me finish with our outlook. Shortly after our IPO last year, we introduced initial 2026 revenue guidance of $186 million to $189 million. Today, we're raising that guidance for the fourth time to $204 million, which, if achieved, will result in approximately 28% year-over-year growth with an adjusted EBITDA margin above 61%.

We're also announcing initial guidance for the full year 2027. We expect approximately $238 million in revenue at an adjusted EBITDA margin of approximately 61%. That outlook assumes neither a recovery in the housing market nor an above-average hurricane season. It also excludes any inorganic growth and any substantial contribution from new product lines.

In other words, we expect to continue delivering strong, highly profitable growth without needing the external environment to improve. And that brings me back to where I started. The strength of Neptune isn't any single product, technology or distribution relationship. It's how those advantages work together and become more powerful as the business grows. One year after our IPO, Neptune is materially larger, more productive and more capable. Our distribution network is expanding. Our technology is improving productivity across the business. Our capacity relationships are stronger than ever. We're serving 336,000 customers nationwide, and we've demonstrated that those advantages translate into sustained growth and exceptional profitability. That's what gives us confidence in our outlook and in the significant opportunity still ahead of us.

With that, I'll hand it over to Jim.

James Steiner

Thanks, Matt. I'll spend a few minutes on the financial results, the underlying economics of the business and our balance sheet and capital allocation. Starting with the quarter. Revenue was $59.6 million, up about 34% year-over-year. Adjusted EBITDA was a record $36.8 million, up about 38% for an adjusted EBITDA margin of about 62%.

So once again, adjusted EBITDA grew faster than revenue and our margin expanded while we kept investing across the business. A few things are driving that. First, the in-force book keeps growing. We ended the quarter with about $450 million of premium in force, and each year, the renewal book contributes more of our revenue.

Our revenue retention rate over the last 12 months is about 89%. Retention matters a great deal to our financial model. Each year, we start with a larger base of policies already in force. With strong retention, much of that book keeps earning revenue before we write a single dollar of new business.

Second, the operating leverage in the business continues to show through. We don't retain underwriting risk as premium grows, we don't need to provide additional insurance capital. And much of our infrastructure doesn't need to grow in step with premium. That includes our underwriting platform, our policy administration technology, our data science capabilities and our corporate functions.

Our largest variable expense is distribution commission. Beyond that, as premium revenue grow, a meaningful share of additional revenue can convert to adjusted EBITDA while we keep investing in the platform. That's why we believe margins above 60% and continued investment in growth are entirely compatible. You can see the same dynamic in our productivity numbers.

Over the last 12 months, revenue per employee was about $3.2 million and adjusted EBITDA per employee was about $2 million. Those figures keep rising because revenue is growing much faster than the organization.

Turning to cash flow. During the quarter, we generated $25.9 million of operating cash flow. For the first 9 months of the year, operating cash flow was $69.3 million. That cash gives us a lot of flexibility in how we allocate capital. At quarter end, we had $215 million outstanding on the revolver.

Our total net leverage ratio, as defined in our credit facility, was about 1.7x compared with about 2.9x a year ago. Over the last 12 months, we've reduced leverage while also returning capital to shareholders. Our approach to capital allocation remains straightforward. The first call on capital is attractive opportunities to grow Neptune. Beyond that, we intend to continue our share repurchase program opportunistically and to reduce debt over time. We don't view those as fixed buckets. Our cash generation and capital-light structure give us the flexibility to put capital wherever we see the best expected return.

One more item. When RSUs vested on September 30, we withheld some of the shares to cover employee taxes. That avoided additional dilution. After quarter end, we paid about $14.2 million in related taxes. Because that cash went out after the quarter closed, it will show up in our fourth quarter cash flow.

With that, I'll turn it over to Trevor for closing remarks.

Trevor Burgess

We recently celebrated the 1-year anniversary of our IPO. Over this past year, I'm incredibly proud that we have grown our revenue by 34%, our adjusted EBITDA by 38%, our premium in force by 32%, our policies in force by 29%, all while our number of employees has actually fallen from 60 to 59.

In the last year, we have generated $82 million in operating cash flow and have used $49 million to delever and nearly $32 million to buy back stock. Our leverage ratio has fallen from 2.9x to 1.7x in the last year, giving us incredible flexibility to look at a wide range of strategic opportunities. I will reiterate that absent investment opportunities that will grow our business even more quickly, our bias will be to opportunistically buy back stock.

There is a paradox at the center of insurance distribution right now. The better AI gets at doing an agent's job, the more valuable the human agent becomes. The tools to disrupt how insurance is sold are here, but the typical American homeowner is now about 57.5, up more than 3 years since 2010, and homeowners are aging faster than the country as a whole. That's according to the National Association of Realtors.

Even the people buying homes today are older than ever with a median age of 59. Those homeowners aren't going to walk away from an agent they've trusted for 20 years just because an AI could do the same job. My own mother isn't going to. When advice is cheap and everywhere, trust is the scarce thing. So we don't see AI replacing the agent in the short term. We see AI working through the agent. That's why we built Atlas+. It puts the speed of AI behind the agent, so the relationship stays with the agent and the work gets faster, smarter and more accurate.

Now none of this is permanent. The futurist Roy Amara said, "we overestimate the effect of technology in the short run and underestimate it in the long run". That still holds. The gap between what AI can do and what customers are ready for will close. Today, Neptune wins by powering the agent. As buyers get more comfortable with AI, Neptune wins by serving them directly, too. Either way, we benefit as this change takes place, however quickly it comes.

I'd like to close by talking about some exciting changes to our Board and our management team. David Noble has joined our Board of Directors as a new Independent Director. I've known David for over 25 years since we worked together at Morgan Stanley. He has become one of the most celebrated executive coaches of our generation, having repeatedly been a top 50 coach and the author of the Harvard Business Review published Real-Time Leadership.

As we grow as a company, having his voice at the table is a privilege and an honor. I am also excited to announce today that our Audit Committee Chair, Cristian Melej, will join the company in January and become our CFO after the filing of the 10-K. Cristian, Matt, Jim, Jon and I first worked together in the early days of my leading C1 Bank, where Cristian would eventually become our CFO. He's an incredible accountant and thinker, and we are lucky to have him join the team.

Jim Steiner, our long-term CFO, will stay on as a Board member and as our COO. Jim has been an incredible business partner, friend and leader at Neptune, and we appreciate his tireless efforts to build the business we have today, and we look forward to continuing to work together. Thank you to our agents, to our capacity providers, our policyholders and to our employees for making this quarter possible. Thank you all for joining us, and we'll now open the call to questions.

Operator

[Operator Instructions] Your first question is from the line of Elyse Greenspan with Wells Fargo.

質疑応答

Elyse Greenspan

My first question, when you guys were talking about AI, right, you were talking about over time, just becoming more efficient. And I think you said like with lower distribution costs. Can you just give us a sense as you guys are thinking out just how beneficial AI could potentially be on your expense levels and how we could think about that benefiting your margins over time?

Trevor Burgess

Sure. Thank you for that question. First of all, I would say I think it's going to take a long time to get there. But we have some good examples in other countries where this transformation has taken place. And the United Kingdom is probably a really good example because there, 60% or 70% of homeowners buy their homeowners policies through a comparative rater. Which you can think about as being an unsophisticated form of artificial intelligence.

And there, the carriers who win the policies pay the online raters a onetime fee of GBP 50. As opposed to an average of 13% every year that we're paying an agent today. So our largest expense category would dramatically fall if consumers decided that they wanted to use AI agents such as Muse or chatbots such as ChatGPT in order to buy their flood insurance. I think it will be a very slow transition, but it would be a transition that would be very profitable for Neptune as that is our largest cost.

Elyse Greenspan

Okay. And then my second question, when you were talking about guidance, you mentioned that you guys -- it didn't include organic growth or contributions from new product lines. I guess in relation to both, I mean, can you just give us a sense of what product lines like could potentially be launched, right, over the coming year or if inorganic growth could be part of the equation and kind of supplemental to the '27 guidance you guys laid out today?

Trevor Burgess

Sure. Thank you. We're in a very, very envious position of having a very attractive leverage ratio, access to capital and tremendous free cash flow. And that allows us to think about investing in new perils such as the earthquake peril, which we will have decided we will turn from a beta product to a fully fledged product by the end of the year. That's starting to gain some steam, still early days, but starting to gain some steam.

We will look at other perils that are climate-related that are carefully aligned with the data and the tool set that we have developed at Neptune and that can be broadly applicable. The demand is really coming from 2 areas. One is our risk-taking partners who would like to do more and more with Neptune because of our success in underwriting flood and quite frankly, how we operate as an AI-native MGA, where data is incredibly transparent to them in ways that just is not how the industry currently operates.

And then second, from our distribution partners who are incredibly interested in finding more ways to do business with Neptune because of how simple our system is to access and buy product. So we are continuously looking earthquake will be the next official product, and we will look at others.

Inorganically, we have looked at a large number of opportunities as they have been presented to us or as we have seen them. And there have not been many that have met our screening criteria just given our organic growth that we are able to achieve. But if there is one that meets our criteria, we're happy that we have the capital structure to be able to execute upon that. And quite frankly, the experience as a management team to be able to execute upon them.

Operator

Your next question is from the line of Mike Zaremski with BMO.

Michael Zaremski

Maybe just poking around in a positive way about the last 12-month revenue retention figure in the very high 80s, 89, I believe. I believe it was low 90s a year ago. Obviously, your results are excellent. But just kind of curious if, given how important that KPI is, if you can kind of talk through what's going on there in terms of the moving parts?

Trevor Burgess

Yes. A lot of what we're focused on is really about the lifetime value of a customer relative to their risk. And we're finding that because of the extraordinary underwriting results that we've been able to deliver, we haven't necessarily needed to push price as much. And we think that, that has greatly contributed to the new business growth side of the house. So it's always a trade-off, right, between how much price you push versus volume in the door and understanding that elasticity.

Again, we're fundamentally pricing based upon risk, but we're able to use price as, listen, if we can lower prices for consumers and get more new people in the door, then that's something that we're going to try to do. That then obviously then translates when it comes time to renew policies, you're going to see some small variations there as maybe we don't push price quite as much.

Michael Zaremski

Understood. Okay. Got it. So you're kind of tinkering as you said you would do during the IPO with that. Okay. And maybe just a high-level question. As we've tested ChatGPT and some of the LLMs in recent months. They've gotten a lot better on the consumer side of talking to it and asking to any flood insurance, et cetera, and being able to bind a policy and it almost always gives us a Neptune quote very fast, much better process than just a couple of months ago. Obviously, direct-to-consumer is still not a big driver of distribution.

But curious, since I think we haven't used Atlas+, would you -- are there any kind of analogies or would you say that Atlas+ experience would be something like the experience we're seeing improvement on the consumer side and that agents just really get answers very fast and bind a quote much faster because of the AI underneath it?

Trevor Burgess

Yes. Atlas+ is really our product to help turn agents into super agents. And we have 2 portions of Atlas+ live right now. One is on the quote screen. And as Matt mentioned, agents who interact with Atlas+ when they're running a quote, bind at 3.5x the rate of people who don't interact. So we know that it is having a really positive effect, causation correlation, right? We get it. But still, that's a big difference, and we're excited. And when we read the questions that they're asking, a lot of it is just helping them get more comfortable selling flood. There are some agents who sell flood every single day, and they already know Neptune, and they know how to breeze through our system, and they don't need Atlas+ because they already know it.

But the vast majority of agents aren't selling flood every single day or maybe they're selling it once a week. And having the world's smartest underwriter sitting right there helping to answer the questions is a very, very valuable tool, and we see that in the data.

The second part of Atlas+ is it's only been out in the wild for about a month now and probably about 30% or 40% of our agents currently have access to it as we roll it out. And we're seeing some really exciting early-stage results. What Atlas+ actions or cards does, it tells the agent what is the next best thing that they can do, and that may be running a new quote or looking at an expired quote from last year or sending an e-mail to a customer who's past due on paying for their renewal or it may be sending them an e-mail saying, "Hey, you've got an endorsement available for pool coverage and that customer, we know has a pool, but you didn't sell in the pool coverage and click this one button to have the AI, send an e-mail out to the customer and they can pay with Apple Pay, the $30 to add the pool coverage.

We've now had over 130 positive tasks completed by consumers. That's renewals, endorsements paid as a result of this Atlas+ card. So we look forward to by the end of the year, having enough data to better understand what its impact is going to be. But beyond the direct impact of getting more consumers to pay the renewals and be properly insured with the right endorsements, it's also acting as a great training tool to agents about what is possible and reminding them of the great things that they can do within the Neptune system.

So -- we are very bullish on these investments on the agent side, just as we're investing on the direct-to-consumer side. As you mentioned, and we put out a press release about it, we are fully enabled on ChatGPT. We're fully enabled for AI agents such as Muse. If someone wants to use Muse to run an Neptune quote, they're going to be able to do that.

Operator

Your next question is from the line of Andrew Kligerman with TD Cowen.

Andrew Kligerman

First question is around carriers. And we're seeing more carriers show up in the flood market. Their volumes seem to be immaterial. But as more carriers leverage AI and digital distribution, what competitive advantages of Neptune are hardest for competitors to replicate?

Trevor Burgess

So I think there are a couple of things. First of all, we haven't seen any carriers gain any material traction that we're aware of. I know there are always Lloyd's syndicates that make an API available to the marketplace. Those APIs are really just connected to an Excel spreadsheet. It's not a sophisticated pricing tool that's looking at the nearly 40 million quotes that we've run in our history and the thousands of claims data and really understanding the fundamental risk behind this.

So I think there is a very material difference in the underwriting that's taking place, number one. And then number two is really about distribution. And it's easy to make those available on some comparator platforms. But moving into large sophisticated distributors who want to make sure that you have a history and a track record and are going to be there for a long time and have the API, the ability to do APIs directly into their quoting platforms and run thousands and thousands of quotes a day and write $15 million policies and be in all 50 states plus Washington, D.C.

We have, in fact, seen a bit of the opposite where some of our largest distributors in recent recontracting that we've done with them have actually shrunk the number of private flood offerings they have favoring Neptune.

Matthew Duffy

Andrew, I would just add on top of that as well. We quoted the new business premium growth number year-over-year. So we're up over 50% year-over-year in terms of new business premium sales. That is fantastic from a sales standpoint. It's also important to keep in mind that there's 100 million buildings out there that our goal is to get insured, right? 25 million that are at moderate to severe risk of flooding.

And so the market is huge. There's a tiny, tiny percent of take-up today, and this is Neptune expanding the market. This is just continuing to sell into those homes that have never had flood insurance before. And that is where that growth in new business is coming from, and that's what we continue to focus on.

Andrew Kligerman

And my follow-up question, and you've given a lot of good color around EBITDA growth. You've taken the guidance EBITDA margin up to above 61% from 60% to 61%. You've talked about generating more and more revenue per employee. Is it possible that in the not-too-distant future, you could be talking about EBITDA margins above 62%?

Trevor Burgess

So if you look at group Q3, we were 61.7%. So we're getting damn close. And it definitely has some seasonality to it, because as everyone remembers, we've got the same cost in Q1 as we do in Q3. But in Q1, people aren't thinking about flood insurance. So we sell a lot less in the winter time. So -- but we certainly hope that we can continue to grow margins over time. And we're seeing some amazing things through the use of AI, some of which we build and some of which we buy, right?

We're buying AI when it's something where there's sophisticated systems that can really do things like customer service where it's just a cost to us. If the most we can save is a couple of hundred thousand dollars, and that's not something I want my team focused on. We want to be focused on things that are centered around our intellectual property and centered around revenue growth.

So if I can pay a couple of hundred thousand dollars for an AI system that then can be put in front of our incoming customer service e-mails and successfully service 60%, 70%, 80% of those, we're going to be able to get a lot more efficient over time. And so I'm excited by what we're seeing with technology, our use of technology, and it's not out of the realm of the possible, but we'll work there over time.

Operator

Your next question is from the line of Gregory Peters with Raymond James.

Charles Peters

I'm going to -- I was looking at your slide on distribution, I think it's Slide 13. And at the bottom and in one of the charts, you talk about these cumulative Triton quotes reaching 38.7 million. And as I look at that number, it's substantially larger than actually the policies in force. So trying to understand what's going on with the 38.7 million because it should be nice if you could convert those all to policies in force. Maybe you have some interpretation. Is it people just getting the same quote over and over again? Or maybe you have some color on what do you think is going on there?

Trevor Burgess

Sure. I think the first thing to remember is that part of what makes Neptune very special is the technology we have to integrate with very large distribution partners directly into their quoting platforms. And when we do that via an API that enables them to do things like run a flood quote every time they're running a homeowners' quote. But that doesn't necessarily mean that all of those large distribution partners have mandated that their agents then verbalize that quote every single time that it is run, right?

And that's what we're working on, how do we change agent behavior over time? That's why we're investing in Atlas+. How do we turn agents into super agents? How do we give them tools to do more of the things we'd like them to do? And what we actually see is with our very best agents where they do offer it every single time, their quote-to-buying ratios can be 5x, 6x, 7x, 8x, an agency who doesn't have the same mandate. And so what we're working to do is change the agent behavior over time. And if we can do that, that quote-to-buying ratio will just continue to rise.

Charles Peters

Okay. I guess, the other question, just more big picture, there's been some small hurricane activity, there's flooding in Texas and New Mexico, and there's some stuff up in the Panhandle Florida. And I'm going to go on the presumption and probably little if no impact on your company, but maybe there has been some impact, maybe it affects market perception of the demand for flood? Or maybe you could just talk about some fallout that you're seeing as a result of those events as it relates to your business?

Trevor Burgess

There certainly have not been any major hurricanes this year that would be the hallmarks of large surges in demand. So the incredible results in the third quarter and the results we've had this year have really been because of our operating on all cylinders, growing distribution, working with our distribution partners, selling more, being focused on delivering great value to consumers, how do we roll out Atlas+ to make agents smarter and smarter and how they could sell. And the weather has not done us any favors this year.

[ Summer ] did us a favor by not causing a huge amount of damage, but definitely has not created that fear-based buying cycle that we see with very large hurricanes. The only slight exception would be just El Nino in general is raising awareness around weather patterns, sometimes in unique areas, right? Hawaii has had, not a big state, right, not going to make a material difference to our business, but they've had a record number of hurricanes that have gone through Hawaii, right? And California you saw Governor Newsom declared a state of emergency because of El Nino for this winter. So we're certainly watching those things, and we'll see if any of that results in some uptick in demand. So I would say we've done really well despite there not being any hurricanes and despite there not being a healthy housing market.

Operator

Your next question is from the line of Tommy McJoynt with KBW.

Thomas Mcjoynt-Griffith

A question on the metric when I look at commission income as a percentage of written premium. I understand there's been some moving pieces with your renewals of terms with different capacity programs. And you also mentioned the change in having no contingents in all terms effective on December 1. When you take those different moving pieces, we saw the third quarter a nice step-up in that metric, that commission income metric. Is that a good number to use going forward do you think? Or do you think there's still upside from some of the renewed terms playing out?

Matthew Duffy

Yes. Tommy, thanks for the question. The renewal of our -- each of our programs, each of our capacity relationships happen throughout the year. So it's not like these all come into effect at the start of Q3 or anything that would cause a particularly large jump at that point. But we continue to negotiate changes to those contracts as they come up year-over-year. One renews in Q4 that we talked about on the call. And with that renewal, we removed those contingent commissions and got some favorable increases in the ceding commission there.

So I think where we're at today is a good run rate for the near term, but we continue to produce fantastic underwriting results for our capacity partners, and we continue to provide fantastic growth from a premium standpoint -- from a written premium standpoint into those programs. And so as we continue to do so and continue to prove out the model, then as we always say, we continue to have those negotiations.

Thomas Mcjoynt-Griffith

Got it. That makes sense. And then you also mentioned your points around the guidance increase throughout the year. And so when you take the latest 2026 revenue guide versus the initial metric about a year ago, it has increased about 9%. So when you step back, what are the 2 or 3 largest drivers that you'd attribute that outperformance to and especially in the face of what's been a fairly weak housing and a relatively benign weather year?

Trevor Burgess

I'll give you 2 things, and I'll ask Matt to play cleanup there. One is Jean-Luc Eckstein, who's our Chief Customer Officer, and his team have done an absolutely fantastic job making sure that agents across this country understand Neptune, its capabilities and how we can help them grow their books in the face of a challenging and soft market for most brokers, right?

So flood can be an area of great growth and great value to their customer. So a big shout out to our distribution team for a lot of hard work this year on making sure that agents understand Neptune's value proposition. That's adding new agencies and really importantly, working with existing agencies to do more and deepen those relationships.

The second thing is around technology and Atlas+ helping -- providing that agent with an amazing tool sitting right next to them when they're running a quote, helping them to be a super agent and get their questions answered and give them the confidence necessary to sell that policy. It is an incredible tool, and we've had thousands and thousands of agents interact with that and benefit from that tool. Matt, what else would you add?

Matthew Duffy

Yes. I just add 2 things. One is from a product standpoint. So we continue to modify the product and improve the product. And so some of the changes we've made this year with increasing the limits that are available, adding those optional coverages, we sell -- every day, we continue to sell more and more higher limit policies, high premium policies and just have a better products market fit than we ever have before.

And I think that continues to be the case in the flood peril. And as Trevor mentioned earlier, we're getting more and more confident that that's the case in the earthquake peril as well. And then I'd also just mention the sort of turmoil that we've seen with the NFIP with the shutdown at the end of last year. And then with the issues again during Q1 at the start of this year, I think more and more agents through that process found the ease of use through Neptune, found our system and continue to like that process and do more and more business with us. So I think it's all of these points compounded and combining, which is what allows us to raise the guidance and also set the guidance that we've set for 2027.

Operator

Your next question is from the line of Josh Shanker with Bank of America.

Joshua Shanker

So look, the weather is not really working for you. And obviously, the housing formation is not, but you're doing great anyways. Is there any stats you have around the percentage of homes that are being bound with a new coverage, including flood compared to past years and the percentage of those homes, I guess, that are getting Neptune policy, which is probably a very high percentage. Is there a way to think about the take-up rate, how it's changed over time?

Matthew Duffy

There's not a great way to do that, Josh. There's not a great public data set around it. There's some back of the envelope math if we look at the size of the NFIP and we look at the rate at which they are shrinking over time and they continue to do so. As we sit here today, Neptune is just shy of 10% of the size of the NFIP from a policy standpoint. And also 5 years ago, we were 1%, and we're pushing 10% as we sit here today.

So we obviously continue to grow from a market share standpoint pretty rapidly as we have rapid growth and the NFIP declines. I think if we look at the sort of a market rate of turnover with the NFIP, we come back into Neptune winning 20% to 30% of all new policies that are issued from a stand-alone flood standpoint here, which continues to increase year-over-year from that same -- if we do that same equation.

Joshua Shanker

And do you have any thoughts around the percentage of policies that you're writing that are for a new home purchase versus somebody who's being shunted out of the NFIP or just changing from another carrier? Is there any way to think about that?

Trevor Burgess

A little bit. We can look at some data around who's buying us on a mandatory basis, right? And we do have some data around the timing of their mortgage and if it's a new purchase. And the estimate that we have is that only about 20% of our policies actually came from the NFIP, actually switched from the NFIP. And everybody else is net new. And that could be net new existing homeowner or net new because they're buying a home and never even went into the NFIP, right? And the breakdown between that, again, the data is not really available and it's not great, but that's about the best direction I can give you right now. We are working on some data that may help us to be able to better answer that question in the coming years.

Joshua Shanker

I may ask some more questions offline about this.

Operator

Your next question is from the line of Paul Newsome with Piper Sandler.

Jon Paul Newsome

I was hoping you could give us maybe a little bit of your most recent thoughts on the rollout mechanically of the earthquake business and in terms of how it will proceed from a state perspective and agent licensing perspective, et cetera, just to give us a sense of the pace of what might or might happen with the new product.

Trevor Burgess

Yes. So thanks for the question about the earthquake product. A couple of things. So when we launched in beta available just in one state, California, and only available via our agent portal. So an agent had to sign in and directly run a unique earthquake quote. What's next in the progression is in the past week, we've begun to be testing cross-sell, which means when someone in California, an agent in California runs a flood quote, they have the opportunity to immediately see what the earthquake quote would be, right, and without having to do any additional work.

So we actually sold our first cross-sell policy today. So that was an exciting milestone there. Next on the journey will be API availability, where we will make the earthquake product available via API to our large quoting partners. And then next will be the direct-to-consumer portal for earthquake, which allows us then to do things like include a new kind of Atlas+ action card, which would be a cross-sell from an existing flood policy to then have an agent be able to click one button and send their consumer a new earthquake quote and have that consumer fulfill entirely online if that's what they'd like to do.

So when we talk about fully launching a product, it means taking full advantage of Neptune's entire suite of technology and all that comes with it. And we're now 2 out of 3 or 4 or 5 of the steps along the way. So we hope by the end of the year, all of those steps are completed, and we'll call it a fully launched product at that point. Importantly, this is an E&S product. It is a technically deductible buydown product that fits under the export list. And so it does not require any unique licensing from the agents, but there are steps that agents have to go through to sell a surplus lines product in California, and our system is sophisticated and builds those systems and processes.

Jon Paul Newsome

That's great. Separate question. Is there -- now as we past the IPO, an optimal amount of financial leverage for Neptune in your opinion that over time, you want to be at? Or is it just sort of a -- we'll see as we go, kind of, process from a financial leverage perspective?

Trevor Burgess

We're at a pretty low level of leverage right now at 1.7x. And obviously, this is somewhat dependent upon the interest rates at the moment and other opportunities. I would say other than month-to-month opportunities to buy back stock, where I'm not a big fan of adding a ton of leverage to buy back your own stock, right? So we buy back stock with free cash flow, maybe smooth out using the revolver, but certainly not going to, at this point, lever up to buy back a lot of stock.

And so really, what having a low amount of leverage enables us to do is be very opportunistic if there are inorganic growth opportunities in front of us, right? And if there are ways that we can find companies that can really benefit from Neptune's technology, then we'll want to take advantage of that. So no set target for leverage, but I'm very happy to be in the position we're in as a company.

Operator

Your next question is from the line of David Motemaden with Evercore ISI.

David Motemaden

Just a question on the 17% revenue growth for next year that you guys are projecting. It sounds like the commission rate is stable here and could be up. I guess that would be up for the full year in '27. But maybe you could talk a little bit about what you guys are contemplating both on like the PIF growth side, but also how you're thinking about premium growth and pricing. If some of the good lifetime loss ratio results, you guys are thinking of passing that through in lower pricing to drive volume?

Matthew Duffy

David, yes, thanks for the question. So I think we continue to be more and more bullish around our ability to grow PIF to grow distribution and to drive those new business sales. I think from a pricing standpoint, we have a goal to get as many people covered as possible, right? This is an issue that America has faced for a long time is the low coverage rates that we have in the country. And so we want to get more and more people insured here.

The portfolios need to continue to make sense from a capacity standpoint for our capital -- for our capacity providers. And so that's always sort of the tension that we play within Neptune. But we expect to continue to have very good premium and revenue retention figures as we look into next year as well as strong new business growth quarter-over-quarter as we get through next year. What we're not planning on there is a particularly strong hurricane season, any uptick in the housing market, anything happening with NFIP, right, any organic -- sorry, inorganic growth opportunities or any particular growth in new products that we might put out there and really including the earthquake product that we have today.

And so I think we continue to become more and more bullish quarter-over-quarter, and I think that's reflected in the revenue growth that we've put out there. It's obviously -- as we get bigger and bigger and that revenue number gets bigger, it's harder to put up larger growth percentages, but that's what we continue to work on.

David Motemaden

Great. And then maybe just a quick follow-up. The Palomar agreement seems to have been a pretty big success. Is there any sense you could give us now that, that's sunset, how much that has contributed to the growth here? If I just sort of look at, I think you guys have added on a quarterly basis, about 20,000 units. Is there any way you could size how much Palomar is within that, on that, I guess, is now being lapped?

Matthew Duffy

Yes. So if we look at Q3 revenue growth of 34% outside of the Palomar deal, that growth would have been 30%. So we continue outside of the Palomar deal to grow very rapidly. But we have a fantastic relationship with Palomar. They're a great team over there. We're very thankful for the relationship. And I think this has been a really, really good scenario, a really good use case from both sides where Palomar and their portfolio has been able to benefit greatly over the last 12 months, and we've also been able to add nicely to Neptune's growth at the same time.

Operator

Your next question is from the line of Rob Cox with Goldman Sachs.

Robert Cox

Yes, just a question going back to the AI agent discussion and noting the press release you put out on news, I'm just curious if you have good data on how AI agents are interacting with your platform today? And if so, what you think that tells you about how this might develop over time?

Trevor Burgess

I would say it's entirely immaterial at this point and it's a good hobbyist exercise for people at this time. But if you look at the average age of a homeowner in America, this is not yet a mass market product for buying insurance. I hope someday it becomes that, but it's not impactful today at all.

Robert Cox

Yes. If I could follow up on maybe just the premium per policy trend. So up 2% year-over-year. I know you guys have increased limits on some products and you're still getting price increases. So I'm just curious what is sort of dragging down the premium per policy year-over-year as opposed to what it would be otherwise.

Trevor Burgess

You have to look at mix as well, right? So during a government shutdown year such as last year, you're going to sell a lot of A zone mandatory policies. And as that went off, you may sell a higher percentage of X zone policies, which are on average lower price. So you really have to look at mix as well before just looking at overall premium average.

Operator

Your next question is from the line of Cave Montazeri with Deutsche Bank.

Cave Montazeri

My first question is on the CFO transition. Things are doing pretty well with Jim. So I'm just curious to know if you can give us any color on why the transition and why now?

Trevor Burgess

I'll give you my view. And then if Jim wants to add anything, I'm happy to have him jump in. We've been at this for 10 years now, and it's been an incredible experience. And what we're finding is that we need -- at least for some period of time, we need more help at the senior levels. And Jim has done an incredible job as CFO, but by becoming our full-time COO is going to be able to help us to automate and transition as many of those things to technology as possible over time.

Cristian is someone that we've worked with before. He was the CFO of the bank that we all worked at and ran for years. And so he's a known quantity. He's our -- until yesterday was our Chair of our Audit Committee at Neptune. So he has a deep understanding of Neptune and its financials. And so this is a great opportunity to share the workload and broaden the base. Jim, anything you would add?

James Steiner

I'd just add, given the history that I've had with Cristian, there's really no one else in the world I would feel more comfortable handing the reins over to as my successor in this role. I also would just remind everyone that I'm staying on as the COO. There are plenty of operational things to get done around here for now in the foreseeable future, and we'll bring on the Board an oversight function and my eyes will be on the financials as yours are.

Cave Montazeri

Great. And my follow-up is on just price transparency in flood insurance. And here you're loud and clear that older people are less likely to use, like, a personal agent to buy insurance and to buy [indiscernible]. They could be using it to just shop around, educate themselves and maybe eventually let that trigger a discussion with our broker. So my question is, if that happens, how well positioned is Neptune to compete in a world where there is more price transparency?

Trevor Burgess

Boy, that would be a tremendous opportunity for Neptune and probably the biggest opportunity for us because 2 things would happen if everybody used an AI agent to get educated. The first is that 21 million Americans would find out that their house is at high risk of flooding and that they need to buy flood insurance and they don't have a policy today. So that AI is going to tell them the truth about the science, and that would be fantastic.

So job #1, please everyone use an AI agent and get educated about your risk because it will expand the market from 4 million policies today to 25 million. So that would be number one. And then number two is you have to remember that there are currently 1.7 million people who have NFIP policies who could save money by switching to Neptune today. So if they've got the AI agent out there shopping for them, Neptune is the primary beneficiary of that, right?

We have the capacity relationships to be able to grow and serve those all 1.7 million of them found out tomorrow because of an AI agent that they could save money by switching to Neptune. We're the only one with the capacity to be able to actually effectuate that change. So both of those things would be tremendous outcomes for Neptune. We have the best data. So like a Progressive, if you have the best data, you're going to win in a sophisticated open and transparent marketplace.

Operator

There are no further questions at this time. I will now turn the call back to Trevor Burgess for closing remarks.

Trevor Burgess

I want to thank everyone for joining the call today. We are thrilled by third quarter's results and the momentum that we have in this business. I think I said this last quarter, but I'm going to repeat it. I have never been more excited about Neptune, the prospects and where we can go with this company. How we are using technology is incredibly innovative. We have a lot of exposure to a lot of different insurance companies. And what we are doing is bleeding edge and is leading the industry, and I think will serve us very well as this AI transition continues to take place. So thank you all for your help. Thank you for the questions. And look forward to the quarter ahead.

Operator

This concludes today's call. Thank you for attending. You may now disconnect.