Innventure, Inc. (INV) Q4 2025 Earnings Report, Transcript and Summary

Good day, and thank you for standing by. Welcome to the Innventure Fourth Quarter 2025 Earnings Conference Call. [Operator Instructions] Please be advised that today's conference is being recorded. I would now like to hand the conference over to your first speaker today, Lucas Harper, Chief Investment Officer. Please go ahead.

Thanks, operator, and thank you all for joining us for Innventure's Fourth Quarter 2025 Earnings Call. My name is Lucas Harper, Innventure's Chief Investment Officer. And joining me from the company are Roland Austrup, Chief Growth Officer; and Bill Haskell, Chief Executive Officer; and Dave Yablunosky, Chief Financial Officer. Earlier today, we issued a press release announcing our financial results, which are available on our Investor Relations website, along with the supplemental slide presentation. As referenced on Slide 6, we will be discussing non-GAAP financial measures during this call. The most directly comparable GAAP financial measures and a reconciliation of the differences between the GAAP and non-GAAP financial measures are available in our earnings release and supplemental slide presentation on our website. In addition, certain statements being made today are forward-looking statements that are based on management's current assumptions, beliefs and expectations concerning future events impacting the company. These forward-looking statements involve a number of uncertainties and risks, including, but not limited to, those described in our earnings release, Form 10-K for the period ending December 31, 2025, and other filings with the SEC. The actual results of operations and financial condition of the company could differ materially from those expressed or implied in our forward-looking statements. And now I'd like to turn the call over to Roland Austrup.

Thank you, Lucas, and thank you to everyone joining us today. I am Roland Austrup, Chief Growth Officer. Before I begin, I want to note that this April, we will host an Innventure CEO Call featuring deep dive commentary from Accelsius CEO, Josh Claman, AeroFlexx CEO, Andy Meyer; and Refinity's CEO, Bill Grieco. There will be more details to follow, and I strongly encourage our shareholders to tune in. Now let me start by saying something plainly. This is the earnings call we have been building toward, not because of a single milestone, not because of a single announcement, but because for the first time in Innventure's history, every part of this platform is firing at the same time, and the results are undeniable. There is a difference between a company that tells you it is going to do something and a company that has done it. There's a difference between a thesis and a proof, and what we are presenting to you today is proof. This is not one milestone. It is not one announcement we're addressing up for you. Let me give you the headline and then I'll give you the proof. The headline is this: Innventure has crossed the threshold from potential to performance. And the proof is in third-party validation, commercial bookings at scale, operational expansion, execution milestones delivered across our family of operating companies simultaneously. What you are seeing in the fourth quarter of 2025 and the opening months of 2026 is not incremental progress. It is decisive across the board inflection in the trajectory of this company. This is what an industrial growth platform looks like when it starts to run and it is what differentiates Innventure from single asset or venture style stories. Since inception, we have deployed approximately $160 million of balance sheet capital into our operating companies. That capital has generated roughly $860 million in net asset value, including approximately $460 million distributed directly to shareholders through PureCycle. That track record matters, but what matters more is what is happening now. The platform is beginning its transition from being capital funded to be commercially self-funding. And the evidence is clear. In the first quarter of 2026 alone, our operating companies generated more than $50 million in new bookings in a single quarter. That is a commercial inflection point by any measure and a powerful leading indicator of forward revenue and enterprise value creation. Across our operating companies, the momentum is unmistakable. Accelsius is scaling with the speed and urgency the AI infrastructure build-out demands backed by institutional validation from Johnson Controls and Legrand and a growing pipeline of commercial deployments. AeroFlexx has entered prestige beauty, one of the most demanding markets in the world and expanded global manufacturing capacity to meet accelerating demand. Refinity moved from formation to pilot scale validation in just over a year, demonstrating the speed, repeatability and discipline of the Innventure model. Three companies, three proof points all at once. This is not a coincidence. This is architecture, the architecture of a platform business delivering on its promise. This momentum underpins our expectation of reaching consolidated cash flow positivity in 2028, driven by Accelsius expecting to achieve cash flow positivity this year. Each operating company is now directly raising capital, we're reducing the need for Innventure's balance sheet and fundamentally changing the financial character of this business, exactly on schedule. Our model has always been well defined. I know there are investors on this call who have been patient. I know there are investors who have been waiting for us to stop talking about what we are going to do and start showing what we have done. We appreciate your patience. I want you to hear me clearly now. The waiting is over. The results are here. They are accelerating, and the best chapters of the Innventure story are the ones we are writing right now. With that, let me pass the call to Bill Haskell to walk through each operating company and provide the specifics behind this acceleration.

Thanks, Roland. I want to start with Accelsius, and I want to start with something I think people in this market is still underappreciated. The world has decided that launch artificial intelligence. Not eventually, now. Every major technology company, every sovereign wealth fund, every hyperscaler on the planet is in a race to build compute infrastructure at a scale that has no historical precedent. And here is the part that most people have not yet fully internalized. You cannot run that infrastructure without solving the thermal problem, you cannot. The chips that power AI generate heat and densities that make traditional air cooling physically insufficient. This is not an engineering preference. It is completely thermodynamics. That is the market Accelsius is scaling into. And Accelsius is not scaling into it theoretically and is scaling into it with over $50 million in contracted backlog secured in the first quarter of 2026 alone, all tied to greenfield next-generation data center development, acted by an initial order for the first deployment by DarkNX, a vertically integrated and funded AI data center developer with a healthy tenant pipeline and the ability to build the liquid cooling, ready capacity and an accelerated time line. Now I want to be honest with you about something because I think honesty on earnings calls is more valuable than cheerleading. Data center construction is experiencing real global supply chain on strengths. Our distribution equipment, switchgear, memory and milling mechanical systems. These constraints can affect the timing of delivery and revenue recognition even when customer demand and purchase orders are firmly in hand. So while we expect to recognize the majority of the contracted backlog as revenue this year, the exact cadence is difficult to forecast with precision. Our expectation today is that revenue will be heavily back-end weighted in 2026. But I want to be very clear about what that means and what it does not mean. It does not mean demand is uncertain, it does not mean our technological improvement. It means the physical world has supply chains and supply chain set constraints. The important signal is not the quarter-to-quarter timing. It is the bookings. It is the customer commitments. It is the scale of demand we are now seeing. Those are the leading indicators of long-term value creation and those indicators are unambiguous. Based on our current trajectory, Accelsius is on a path to exit 2026, cash flow breakeven defined by cash from operations. This implies a December 2026 annual revenue run rate of approximately $100 million. And importantly, we believe Accelsius cash on hand is sufficient for the company to reach this cash flow positive threshold. Think about what that means, a company that just a short time ago was still in the early field deployment is now approaching self-funded commercial scale. Let me contextualize this further. -- because the market is telling you something important that you should be paying attention to. The recent acquisitions of CoolIT and Boyd had roughly 8 to 9x revenue and nearly 30x forward EBITDA and make it unmistakable that the industry is moving decisively toward direct-to chip liquid cooling. And here is the critical distinction. Both CoolIT and Boyd remains single phase today. Accelsius is already commercially deployed in the 2-phase architecture that the market is converting toward. Two-phased cooling is not an incremental improvement on a single phase. It is a fundamental architectural advantage. Because of the phase change that occurs, it removes far more heat with far less energy, enabling rack densities and thermal performance that single-phase water systems simply cannot reach. Industry analysis consistently show that directorship cooling is one of the fastest-growing segments of the data center thermal market, with forecasts ranging from low double digits to mid-30% compound annual growth rates over the next decade. The earliest adopters are exactly where we are seeing our strongest traction today. Greenfield, high-performance computing and AI focused data centers, where air cooling cannot keep up with the heat box of modern GPUs and accelerators. But here is what I want investors to understand about the size of the opportunity. The first one is already here, new builds, HPC, AI infrastructure. But the second wave, and this is potentially even larger is legacy data centers. Even in facilities where air cooling is technically adequate today, operators are recognizing the 2-phased school income locks higher rate lease, greater compute per square foot and significant energy savings that allows them to densify and sort of expand to deploy more complete power that new construction to reduce the energy overhead of air-based cooling. We believe that the necessity of 2-phased cooling for HPC and AI workloads, combined with the compelling economics for non-HBC environment, will cause direct to chip 2-phased cooling to become the dominant architecture for both new and retrofit data centers over the next 3 to 5 years. Accelsius is now widely recognized as a leader in direct-to-chip, 2-phased cooling, a position reinforced by our strategic investors, Johnson Controls and Legrand. Their involvement is not passive. It is a strong validation of both our technology and our commercial readiness like 2 of the most respected names in global building systems and data center infrastructure. In December 2025, Accelsius closed the second tranche of a $65 million Series B investment, led by Johnson Controls and Legrand, valuing the company at approximately $665 million post money. I want to emphasize this, that valuation was set by 2 global industrial companies deploying their own capital. not paint venture, not by internal Accelsius financial models, but by external institutional investors with deep domain expertise writing real checks. That is the kind of validation that is very difficult to dismiss. Let me turn to AeroFlexx, which operates in a completely different market but demonstrates the invention model just as clearly. There is a problem in packaging that at everyone acknowledges but almost no one has solved. The world produces an enormous amount of single-use rigid plastic packaging. Everyone agrees that is wasteful. Everyone agrees the supply chain or inefficient. And yet, the alternatives have historically forced a trade-off. You could have sustainability or it could have performance in consumer appeal, but you cannot have both. AeroFlexx changes that equation. Founded in 2018 around liquid packaging technology sourced from Procter & Gamble, AeroFlexx is an integrated packaging and filling platform that improves the consumer experience, simplify supply chains reduces virgin plastic usage and enhances e-commerce economics. Its differentiation comes from delivering all of this value simultaneously. The current side recyclable package. It uses up to 85% less virgin plastic than rigid bottles, a flat back format that enables up to 10x greater shipping efficiency, lower total cost of ownership by consolidating the supply chain and consumer testing that consistently shows a clear preference versus traditional packaging. This is not a trade-off. This is a better product. As of the fourth quarter, AeroFlexx has delivered 6 consecutive quarters of revenue across pet care, baby care, personal care, household products and industrial applications. And what is notable today is that AeroFlexx is transitioning from early market validation to large-scale adoption and volume production units. During the first quarter of 2026, AeroFlexx announced a global partnership with Aveda, part of the Estee Lauder Companies. Aveda is the first global prestige brand to adopt AeroFlexx refill packaging format with select products expected to date with early 2027. Let me tell you why that matters beyond the headline. Prestige beauty is one of the most demanding markets in the world. the brand standards, the performance requirements, the aesthetic expectations is are extraordinarily high. When Aveda backed by Estee Lauder chooses our platform, that is the statement about the maturity and credibility of our technology. Aveda is 1 of 4 anchor customer relationships that now underpin AeroFlexx's commercial momentum across distinct end markets. The other anchors include a multinational consumer packaged goods company with a signed multi-brand multimillion unit agreement, a major producer of industrial fluids and packaging services where commercialization is advancing through both equipment and back sales with the first purchase order already received in production beginning next month. a large beverage and food service partnership that was made at AeroFlexx entry into fluting beverage, the largest portion of its addressable market. Taken together, these 4 anchor customers valid the platform across premium beauty, household and personal care, industrial applications in food and beverage, and each has the potential to support line extensions geographic expansion and follow-on programs as AeroFlexx becomes more deeply integrated into long-term packaging strategies. AeroFlexx near-term commercial pipeline stands at just on the $30 million including an approximately 1/3 in final negotiations. We have not provided any guidance on the timing of revenue conversion, but the realization of these opportunities is incorporated into our assumptions or AeroFlexx to reach cash flow positivity in 2028. The company's opportunity set is broader and more diversified than at any point in its history. AeroFlexx is also in the process of launching a direct capital raise at the operating company level, targeting strategic investors that also serve as commercial partners. As our operating companies mature, they are increasingly able to raise capital independently, reducing the need for parent level funding. That is the model working exactly designed. Let me turn to Refinity, and I'll tell you candidly, this may be the most compelling industrial opportunity we have ever launched. Here is the problem. The world produces hundreds of millions of tons of plastic weighted every year. A meaningful portion of that waste has no viable recycling pathway to today. It goes to landfills, they go to incinerators, it goes into the environment. At the same time, petrochemical companies are spending enormous sums buying fossil feedstocks, ethane and naphtha to produce ethylene and propylene, 200 carbons that represent a $350 billion global market and are essential to producing polyethylene, polypropylene and a wide range of specialty materials. Refinity connects those 2 problems. It takes the portion of plastic waste stream that today have no viable recycling pathway and convert it into high-value chemical building blocks, ethylene and propylene the petrochemical companies are already buying. The substitutional loan creates a compelling economic incentives and ability to hedge against fossil price swings while meeting circularity commitments. Across the value chain, circular materials command a 30% to 50% price premium with the highest premiums closest to the consumer. This is not a sustainability story that require you to or economics. This is a sustainability story where the economics of the reason it works. Refinity's primary commercialization strategy is built around integration, co-locating plants directly at petrochemical sites such as the Dow steam cracker. This eliminates testation cost feeds directly into existing infrastructure, reduces CapEx for both Refinity and the customers -- it accelerates adoption by fitting seamlessly into the way these companies already run their assets. Beyond its core ethylene and propylene focus, Refinity sees significant opportunities in producing customized circular hydrocarbon products, including specialty high-value lubricants and sustainable aviation fuel or SAF. One of our independent directors is a C-suite executive in the aviation industry, and we have come to appreciate that SAF has become one of the most critical pathways radiation to meet its Net Zero commitments with demand going far faster than supply and U.S. production expected to scale dramatically over the next decade. Refinity's recently licensed technology from a U.S. national lab for catalytic conversion of its mixed ethylene and propylene product to SAF and SAF to crystal liquids and intends to demonstrate this conversion process late this year. The SAF market alone is growing at 38% to 50% annually and is anticipated to reach $40 billion by 2034. The ability to disrupt a $350 billion commodity market while also accessing high-growth specialty sectors like lubricants and SAF underscores just how significant the total addressable market is for Refinity. Now here is the process to get your attention. Refinity was formed in December 2024. Less than 1 year later, the team produced its first metric ton of circular product from real-world mixed plastic waste yields typically above 60% to 70% with minimal chart. That compares to about 25% conversion for competing technologies. For our technology of this complexity at speed is exceptional. Since then, Refinity has filed multiple patents on a DuoZone reactor design, expanded its IP portfolio with licenses from the U.S. University and a national lab and advanced engineering toward a 10-kiloton per year demonstration plant targeted for completion in 2028. A commercial scale plant of around 150 kilotons per year is planned for early next decade aligned with the chemical industry's expected return to growth. Refinity is hitting KPIs ahead of schedule. It is solving a real cost problem for petrochemical customers and is positioning itself to scale just as the industry reentrant and growth phase. This is the Innventure model: rapid formation, accelerated validation and a disciplined progression towards commercialization in a massive market with structural economic drivers. Before Dave gets into the financials, I want to leave investors with 3 clear takeaways. First, invention awards. PureCycle approved it and Accelsius is validating it again at a faster pace. This is not theoretical, it has been demonstrated to us. Second, we are not dependent on a single operating company. We now have 3 businesses executing simultaneously, each with independent third-party validation that is diversification with conviction, not diversification as a hedge. And third, I think this is the one that the market has been slow to absorb. The platform is transitioning structurally from capital consuming to increasingly self-funded. Operating companies are raising their own capital. They're converting commercial traction in the revenue, the architecture of this business is changing and it's changing exactly the way we expected it would. I want to say something about valuation because I think it needs to be said plainly. We believe our current share price does not fully reflect the value of meta shares. The $665 million third-party valuation of Accelsius was set by institutional investors deploying their own capital, adding 2 strategic investors to the cap table and securing more than $50 million in contracted backlog. We believe the value of Accelsius alone has materially increased, and that does not include the value of AeroFlexx or energy. We're not going to complain about the market, but we are going to stay static. In fact, suggests there is a significant gap in we were our shares create and what this platform is worth. Our focus remains on execution. We believe that if we continue to execute, value will ultimately be recognized and we intend to continue executing. When we look across our family of operating companies today, our confidence in Innventure's path to consolidated cash flow positivity in 2028 is grounded in execution, not aspiration. Accelsius is scaling into production deployments in a market where liquid cooling is becoming mandatory with third-party institutional validation and a clear line of sight towards self-funded growth. AeroFlexx has moved beyond pilot programs into repeat revenue, anchor customers and global manufacturing scale while transitioning to direct capital formation at the operating company level, and Refinity is rapidly validated its core technology, established a clear commercialization road map and has become the process of funding its next days independently. Taken together, these developments reflect a platform that is structurally maturing with the operating company is increasingly funding their own growth, corporate capital requirements declining and commercial activity translating into revenue. This is exactly how the Innventure model is designed to work, and it underpins our confidence in the enterprise's long-term financial trajectory. With that, I'll turn the call over to Dave to walk through the financials.

Thanks, Bill, and good afternoon, everyone. I'll walk through our fourth quarter and full year results, but let me begin with the most important thing I could tell you. The financial profile of this company is changing, not gradually, structurally and the numbers I'm about to give you are evidence of that change. 2025 was an important proof point year for Accelsius. Revenue increased from $0.3 million in 2024 to $1.6 million in 2025, driven by successful proof-of-concept deployments with early customers. At the consolidated level, Investors 2025 revenue was $2.1 million, up from $1.2 million in 2024. Fees from our management of the Innventure ESG fund along with intercompany eliminations, were $0.5 million compared to $0.9 million in 2024. But the real step change happened in the first quarter of 2026. Accelsius generated more than $50 million in contracted backlog. These are production volume orders, not pilots, not trials. This provides strong visibility into meaningful revenue scaling in 2026. And as Bill mentioned, we expect Accelsius to exit December 2026, and with positive operating cash flow, implying an annual revenue run rate of approximately $100 million. We also expect revenue to be heavily weighted to the back end of this year. General and administrative expense. Before I get into the specifics, I want to explain something about how our cost structure has evolved because it gives important context. We have included a slide in our earnings presentation that illustrates this in granular detail. Historically, prior to the operating companies reaching commercialization, Innventure funded essentially all G&A costs from the topco level: personnel expense, special services, operating expenses, all centralized, all flowing through Innventure's consolidated P&L. That's now changing. While cost at associates and Refinity will continue to flow through the consolidated financials, a growing portion of the total operating expenses will be funded directly within those businesses. rather than buy in venture. At the topco level, our focus is increasingly on a lean corporate cost structure, funding only what's required to operate Innventure topco. Now let me give you the numbers because they're significant. G&A has decreased sequentially every quarter since Innventure went public. Consolidated G&A declined from $29.7 million in the fourth quarter 24 to $11.5 million in the fourth quarter 25, a 61% reduction. That reflects disciplined cost management across Innventure, Accelsius and Refinity, as well as the tapering of noncash expenses associated with our public listing. Professional service fees shows the same trajectory; $3.5 million in the fourth quarter 25, down 42% from their peak of $6.1 million in the first quarter, $25 million as we brought key functions in-house at lower cost. At the parent company level, Innventure's fourth quarter 25 cash G&A expenses were $5.7 million, down over 55% from $12.9 million in the fourth quarter of last year. That's not noise. That's a structural change in how this business operates. Looking ahead to 2026, we expect the venture Topco G&A to follow a trend similar to the last 3 quarters of 2025. A few income statement highlights. Excluding the $347 million noncash goodwill adjustments and other minor noncash items, adjusted EBITDA for 2025 was a loss of $78.8 million. As we look forward, the combination of a significant contracted backlog, our expectation that Accelsius will reach a revenue run rate of approximately $100 million and exit 2026 cash flow positive gives us confidence that there will be a substantial improvement in the reported adjusted EBITDA in 2026. Moving to cash and liquidity. On a consolidated basis, we ended 2025 with $65.4 million of cash, restricted cash and cash equivalents, up from $11.1 million at the end of 2024. Additionally, in January 2026, we strengthened our balance sheet with a $40 million registered direct offering as Innventure became shelf eligible. Shelf eligibility is an important milestone. It gives us efficient access to public market capital on substantially better terms than what was available previously. Just as importantly, we repaid the entire remaining $5.6 million balance on our convertible ventures, which simplifies our capital structure. Let me walk through why we believe our cost of capital will improve significantly going forward. One, we believe Accelsius is now effectively fully funded and entering rapid commercial scaling with the over $50 million in contracted backlog. Two, fourth quarter '25 G&A is down 61% from fourth quarter '24, with further efficiencies expected as we take advantage of productivity improvements. Third, shelf eligibility, which reduces reliance on higher cost financing alternatives. As our operating companies, particularly at Accelsius begin generating cash, we expect this to further extend our cash runway and move Innventure towards a self-funding ever remodel. While it is too early to discuss the details of the ongoing capital needs for Refinity and AeroFlexx. The disciplined cost actions I discussed earlier gives us visibility into in ventures needs. At the Innventure level, we estimate 2026 capital needs to be materially less as our operating companies become self-funded. This reflects a near parent company structure as expenses continue to shift to the operating companies. On the balance sheet, by way of explanation, our $28.7 million in investments represents our $19.5 million equity method investment in AeroFlexx and $9.2 million in AeroFlexx debt securities. And following the goodwill write-downs earlier this year, $23 million of goodwill still remains on our balance sheet. On the cash flow statement, you could see many of the noncash items that appear in our income statement. The cash used in investing activities primarily reflects funding to AeroFlexx and capital expenditures at cells. Let me close with this. There are companies that talked about inflection forms. And then there are companies that cross them in venture is crossing born right now. rapid commercialization, a dramatically improved cost structure, efficient access to capital, operating companies that are beginning to fund their own growth. These are not just aspirations we are sharing with you. They are facts supported by every number I just walked you through it. We believe this combination positions us to scale with far greater capital efficiency and to create meaningful long-term value for our shareholders. And every investor on this call understand we are not slowing down. We are accelerating. With that, we'll open the call for questions.

[Operator Instructions] And our first question comes from the line of Chip Moore of ROTH Capital Partners.

So maybe if I could start on Accelsius, the $50 million plus in contracted orders here in Q1. It sounds like DarkNX is a significant chunk of that, but maybe you can talk about the types of customers in there, other customers and what you're seeing there and then pipeline as well?

Yes, sure. So I would say this, Chip, we have literally hundreds of people in the pipeline of customers in the pipeline that are all kind of moving forward through. I mean the beginning of that is starting to drop through, as you can see. So it's fairly chunky right now. But I think what you'll see going forward is we'll have a larger framework of customers. I mean we have delivered to dozens of customers to date. So I think what you're going to see is many more groups of purchase orders fall with increasing numbers as they go forward. It's tricky marketplace, as I think we all know, just because of, again, some of the supply chain issues that have been talked about on this call and people are seeing in the marketplace. So that affects some of the timing of various both purchase orders and the prospective deliveries of those. And while I'm not predicting that we'll have any material delays in delivery, it's not something within our control. I mean, ultimately, these are things that will be determined by the pace of the build-out of the various data centers and our customers' sort of supply chain constraints. So that's kind of where we stand at the moment. But we'll have a broader and broader, more diversified pipeline of contracts as we go forward is my belief.

Yes. That's helpful, Bill, and it makes a lot of sense. Obviously, a lot of things out of your control like many. And I guess for the deliveries to dozens of potential customers. Would you describe that more as sort of piloting phase? And how long do you think people want to have a look at the technology before they get fully comfortable?

Yes. Well, so last year, virtual all of our deliveries were kind of one-off pilots where people were just evaluating the technology. I think we've moved past that for most all of the customers that are in the pipeline at this stage. So the way I would frame it is this: if we were sitting here a year ago, our average proposal outstanding was probably worth a couple of hundred thousand dollars and now we're not virtual -- not everyone, but most of the purchase orders are either 8 or even 9 figures in terms of scale. I would say most are probably in the 8-figure range. So that just shows you a significant transition from evaluation units to real commercial production orders.

Yes. Definitely. That's helpful. And maybe just one more on Accelsius. For me, to your point, on cadence being tougher to predict near term and some of the things that did control but it sounds like you have reasonable visibility into maybe $25 million-ish of revenue in Q4 if something that your control doesn't get held up. Is that the right way to think about it based on what I said?

Yes, like you said, that's the kind of run rate we indicated that would make the company cash generative. And I think Josh came out a couple of months ago and said that was our expectation that we would reach cash flow positivity by year-end, and that is still our belief.

Yes. Okay. And just one more before I hop back in queue. AeroFlexx, a lot of momentum, data, obviously, announcement recently, but now you're talking about a $30 million pipeline and some of that getting close. Just a little more detail there. And I think I heard you say that you might be looking to do arrays with some strategics there. Just any more color you can provide.

Yes. I would say this, now that we've gotten to the point where we've proven out the technology and proven out the recyclability of the AeroFlexx package, and it's gone through its own evaluation unit phase just as we did in Accelsius. Now we're starting to see, again, same thing, commercial-sized proposals that have been asked for. And so Aveda is really a framework deal that we think can be quite significant. I don't have a number of scale yet of what that can grow to but they're a very large luxury brand within Estee Lauder, as you may know. And what we believe, based on conversations we've had with lots and lots of customers is that they'll start with a product 1 SKU, I'll call it, that they'll go out and run and assuming that's successful, they will kind of broaden the reach of that packaging solution to other brands within the same platform. So again, Aveda is one, but they're a big one. And as we mentioned, it's a very challenging customer in the sense that, again, the bar is very high across the board because aesthetics is very important. And then so they want unique shape, some different kinds of packaging and labeling that is more difficult than, say, industrial where you look -- putting a lubricants and barring chain oil, which is an opportunity for us and other things of that category.

Yes. No, that's great. One last one for me before I hop back in queue, I think probably more for Dave, but the transparency around G&A and some of the expenses. I really appreciate that and the slide in there. I guess the question would be, is there much more low-hanging fruit there? Do you think G&A continues to come down? And how much more optimization do you think you could see there?

Chip, thanks for the question. And certainly, we're always focused on G&A. We're always looking at ways to be more efficient, more effective, get more done with less. So while I don't want to give forward guidance on where I think that might be just going to be, I can assure you, I can assure you, it's on our radar, and we're always looking at ways to operate more efficiency. But we're pretty proud of the 5 consecutive quarters since we went public. So that's how I'd add into that.

Yes. I would just amplify that a little bit, Chip, in the following way. I mean if you we talked about when we went public that we needed to be a public company ready kind of overshoot and we relied very heavily upon outside vendors to help. And now we've brought a lot of that functionality in-house. So we've weaned ourselves off some of those outside services will quite expensive, but that wasn't all realized by the end of December. There was still some carryover. So that continues to reduce which is where I think you'll see some improvements in our cost structure going forward.

Our next question comes from the line of Nehal Chokshi of Northland Capital Markets.

I think it was a well said narrative on the transition of Innventure and proof points that the unique VC model is working. So well done there. There are some questions, though, that I think need to be asked. So I got a bunch, and let me just go through them real quickly. The COGS to revenue ratio continues to inch up I understand that we're still in basically pilot base. But at what point in time -- why is it continuing to inch up? And at what point in time do you expect that to start to get normalized?

Yes. So I would say we're building some things to inventory based on projected orders, Nehal, so the cost of goods is ahead of delivery, right? So I think that's a the primary issue, right, where we have customers that know what they're going to want in terms of product mix. And so we've developed some inventory which, of course, all that cost goes in, but the revenue is yet to be realized in certain areas.

Yes, I'll jump in, and that's accurate, right, really cost element to COGS as well. So you got to keep that into account. It's not all variable.

Certainly, later this year, that will -- that should even out as we start delivering at scale, you'll see all of that really flush out.

Okay I mean when I look at the COGS relative to revenue. I mean it's roughly scaling, but it's increasing a little bit, right? It's not like a massive increase. So it almost looks like the variable cost structure is close to 100% if I were just to look at what the pure analytical lens of COGS to revenue ratio. So can you help me understand like what percent of these COGS is actually fixed versus rambled then?

Care to answer that one, Dave?

I do. It gets into the margins and the cost structure of Accelsius and probably don't want to go into too many details on there. But we're amortizing intangible assets, right? So for the R&D development took place and the other things that are attributable to cost of goods sold, that amortization is going through. It's fixed. It doesn't vary with each unit produced. And the second thing is there's been a shift in Accelsius right to the higher-capacity cool units to a different MR250s and demand by the customer. So that generated a little bit more cost than just doing straight math on units and per unit cost. So those will be the 2 things I'd point you to.

But it's a very nice margin business. I'm not going to give you the projected margins at this stage later as we -- as revenues get to scale, I think we'll be able to share more in terms of that. but the margins are attractive margins in this business, in my view, on a comparative basis to kind of other vendors out there.

Okay. And then -- so the fixed cost element within these COGS that has been going up each quarter then. Is that correct?

Well, well, again, I mean I think the fixed portion is fixed. I just think there's a lot of different things happening as we're scaling as we're getting customer orders and costs are getting booked to COGS. And again, I think as we scale, then you start to see it more normalized where you could say, hey, burn units produced. This is the cost per unit you'll start to make more sense. When your number's at this level, I think you have to be careful drawing those kind of conclusions.

And we are a brand-new manufacturing facility, too. I don't aware of that in the hall, but we've opened another -- I think it's 25,000 in the facility there in Austin in addition to the -- where we had before. So we just materially increased our manufacturing footprint. So there are obviously some costs associated with that.

Okay. Okay. still going on this slide here. Inventory was down about $5 million Q-over-Q on less than $2 million of revenue recognition in the quarter. Can you help me understand that there?

Yes, I can answer that. Again, as we transition to different products, there was some inventory write-downs. And that's flowing through COGS as well. So there was a little bit of obsolescence, a little bit of manufacturing costs, some more heads allocated to cost of goods sold. It's all kind of related, but that's why you saw that inventory.

Yes. And I can give you some more to that. So this market has evolved very rapidly. And our initial belief was that a sort of 70-kilowatt rat was sort of 10x the average rack size and that was going to be kind of where the market was headed. -- really leapfrogged over that. And so we have about 150-kilowatt or a 250-kilowatt product as well. And that's really more of a sweet spot of what the market seems to warrant. So that's where the obsolescence really came in is just writing off a lot of that 70-kilowatt inventory.

Got it. That makes a lot of sense. That's very helpful. Okay. A couple of other questions, and then I'll see the floor here. You said that DarkNX is funded. Can you give us a -- it's hard to find information on this company. Can you give us a sense as to where these funding sources are coming from for DarkNX?

Roland, do you want to fill that one?

Sure, I can, Nehal. I mean all I can tell you is, I mean, coincidentally, of course, I'm in Toronto, which is where they are. So I've met with them, a handful of them already got to know them. I didn't go through the formal qualification that was done by Accelsius themselves, and I believe JCI did that as well because they're part of the chillers for that facility. All I can tell you is they are funded and they represented to me directly that they're funded, but it was formally done when they were qualified by both Accelsius and Johnson Control. A key part of that was determining that they did have funding. I don't know the source though, to tell you truth.

Okay. All right. And then my last question, and I'll get back in the queue for the rest of my questions here. So companies are raising company capital independently, which then means that Innventure will get diluted relative to these operating companies. So does that also represent a change in philosophy on whether to fund the operating companies or not rather than just evidence of operating company maturation? I do agree that there is evidence of operating company maturation, but I'm also saying, hey, does it also represent a change in funding philosophy as well?

Yes. So let me field that one for you, Nehal. So in the early days of Innventure, most of the companies were funded not off the balance sheet of Innventure, but there was subsequent funding we had a fund that code invested within Innventure. We have some outside investors that funded a lot of those, so we ended up with relatively small stakes in each of PureCycle and AeroFlexx. When we evolved with the conglomerate model, our goal was to own more. But the balancing at the trade-off there is until we, Innventure, are cash generative at the opco level, which we projected for 2028, you'd have to take permanent dilution at the Innventure level, which would affect not only the companies that we currently have, but future companies going forward. So the thought was if these companies are in a position, you're mature enough to be able to raise capital independently, let's raise some capital for each of AeroFlexx and Refinity directly in the marketplace. And yes, we'll take some dilution there. but we saved the permanent dilution at the Innventure level for our shareholders in doing that. But when we're cash generative at the top level of veto-proper level, they would love to be able to fund as much as possible of our own balance sheet to retain full ownership. So it's kind of a trade-off between taking care of investor positions today and taking a little bit of dilution at the opco level versus kind of having to suffer permanent dilution for all future companies. And as we mentioned, certainly, Accelsius already has the requisite capital it needs to get to a cash generative position. So we're not funding anything more from there, and they're not raising any further capital. So it's really at the AeroFlexx and Refinity level. And AeroFlexx, as you know, is not a consolidated asset. We own a minority stake in it. So we're a little less sensitive to dilution at the AeroFlexx level. And we believe now that it had done this deal with beta and is seeing contraction that raising the smallish amount of capital they need going forward is imminently doable.

Our next question comes from the line of Aashi Shah of Sidoti.

My question is related to Accelsius and the $50 million bookings in the first quarter. They are all tied to the greenfield data center. When do you expect meaningful traction from brownfield deployments and that could accelerate adoption much quicker than greenfield in my assumption because they're going -- the growth over there is slower and riskier as to when the data centers will be ready. So any thoughts on when the brownfield deploy -- like when you start seeing traction from brownfield deployments?

Yes. It's a tricky question to answer. I'll do my best, Aashi, and thanks for the question. So we do have customers, obviously, that have the existing data centers, and some of those that would like to retrofit at least part of their data centers with a different technology. But at the moment, as you know, liquid cool data center is a relatively small number. It's, I don't know, 5% or something like that of data centers have a coin today. growing very rapidly and evidenced by a lot of the M&A activity in the sector. But we are talking to customers that have many data centers, some new builds and some existing. And the nice thing about the technology is that it drops in very nicely into an existing data center because each rack is self-contained in the sense that you have a cooling solution and a CDU all contained together with one or multiple racks. So you can replace a rack, a row whole facility or any combination thereof in a relatively straightforward way. So I don't just hard to know what the definitive answer to your question is, but I would expect some blend of that even later this year as things move forward. Just with the ones that have the most acute need of those that are mirroring up specifically for HPC and AI workloads, where they're acquiring the latest and greatest as possible. But as we migrate to agentic computing, you're going to see, I think, a big change. 2026 is really supposed to be the inflection point where we go from 80% of the computing being for training these large volume models to 80% being sort of consumer directed for generative AI. This is what's happening. People are starting to use these various AI agents, and it's a very steep curve of adoption. I mean most of it now uses it every day now, and you're going to see that continuing to grow. So if they switch over from again, these LLMs and the compute horsepower goes the other way, then it's more of a CPU game versus a GPU game. And then you're going to have clusters of large computes for CPs in, I'd say, big clusters. And that will also require liquid cooling just because of the density that they want to be able to put forth in one data center. So I mean, just in the last 2 years watching this market evolves, it's gone well ahead of the pace that it had anticipated. So I think we're going to see continue to see a shift. And I think that will drive more to the brownfield sites as trying to get back around to answer your question here because a lot of those...

I can probably add to that a little bit.

Yes, go ahead.

Yes. Because I mean I pretty long conversations about that with Accelsius and kind of the view I get. When you have the CEO call, I think it's a good time to ask that question exactly, by the way, to Josh. But right now, there's a need for greenfields to adopt the technology. So need drives adoption there. But in the real legacy center, really what they want to see is they want to see a mature industry develop that there's plenty of supply. And I think where you're going to see the adoption inflection point is when you see that there's a robust industry supplying or more of the greenfield developments, then the brownfields will become comfortable, I think, making switchovers.

It makes sense. Understood. And another question I had was on the Aveda partnership. Can you just give us a ballpark on what the expected annual volume is going to be for that launch in 2027? And if this is going to be like a pilot program or is it a full commercial rollout?

Yes. So we've been dealing with the data for quite a long time. I don't have the direct answer to your question. I'm not ducking that, I just don't know the answer to the question. But Aveda has big brands, big luxury brands and the reason it's taken until 2027 before they roll out the scale is to figure out which particular product lines they want to use in the packaging making sure we tailor the packaging to what they want to see in the labeling and all the various things that they require. So I just don't know the answer, Aashi, but we'll be learning more over the next 6 months in terms of the details of what we would expect to see.

Yes. I mean you can look at the potential, too, Aashi, I mean, there's not a lot of published data available on Aveda. But if you do any search out there, you'll find that Aveda is in the tens of millions of packages a year is what the brand is estimated to have in the marketplace. So it could be significant. But as with any brand rollout, you don't expect to get the whole thing, but it's definitely not a pilot launch. It's a global commercial launch that's targeted for 2027.

And our final question comes from the line of Nehal Chokshi of Northland Capital Markets.

Yes.So another part of the narrative here is improving corporate governance. And you guys did announce that you'll be, I think, nominating some independent Board of Directors and some existing, I guess, so-called insiders are going to be stepping down. Could you just say what percent is independent now? And what percent do you expect to be independent once these changes are made?

Sure. So today, we have 5 independent directors and 4 executive directors. And what we're targeting is to go to 7 independents and 2 executive directors. And so we have our AGM in June. So I would anticipate in that window of time that we will have migrated to 7 independents. That's the target.

Great. Also, can you give an update on the Accelsius pipeline? I believe a quarter ago, you said it was a little bit over $1 billion.

Yes. I mean I don't have any updated information directly. I will tell you that the pipeline has a lot of different levels to it. There is, I'll call it, high conviction things in the pipeline, there's things that we think are probable, and there are things that we think are possible and then there are new things coming in all the time. So I don't have a competitive number, but we can figure that out and certainly when we do the CEO call, that's a fair question to ask them, Josh, but I just don't have the number in front me. I just haven't seen it yet.

Got it. Understood. And then how much of the greater than $50 million of Accelsius bookings in 1Q 2026 correspond to in terms of megawatts to coal?

Say that again?

The $360 million of Accelsius bookings in 1Q '26, what does that correspond in terms of megawatts to coal? That's a good.

Yes. I would just say this because I think we're trying to be a little bit careful about ascribing dollars per megawatt, which I think is probably where you're headed. I will say that we've part of that, a fraction of that is a small portion of the DarkNX prospective build-out, what had been announced before was 300 megawatts. And then there was another announcement saying that they had the funding for the first 2 phases of 65 megawatts each. That is not what our bookings represent today a smaller fraction of that. So again, I think when we do the CEO call, maybe we'll provide a little bit more granularity on just how many megawatts we've been contracted to roll out. But it's -- and it's going to grow. It's going to grow pretty materially between now and the next couple of quarters. again, a lot of things in the pipeline that we think will close over the next couple of quarters. But we haven't announced that some of we'll see if we can get clear information from that for the call that we do with the companies.

Nehal, that's the whole point of doing a separate CEO call was that we want to have the CEOs to be able to go into greater granularity here on the earnings call, we're really just trying to sort of at the macro overview of where it's going and that we're having an acceleration across all companies and a dramatic decrease in G&A, where you can get into the technicals, I think, is going to be on the CEO call.

Okay. All right. I'll save my additional questions for them.

This concludes the question-and-answer session. Thank you for participating in today's conference. This concludes the program. You may now disconnect.