Antalpha Platform Holding Company

Good day and thank you for standing by. Welcome to Antalfa's first quarter 2026 earnings conference call. Today's call is being recorded. All participants are now in a listen-only mode. After management's prepared remarks, there will be a question and answer session. I would now like to turn the call over to Chris Mamoni, Managing Director of the Blue Shirt Group and representative for Antalfa's Investor Relations team. Mr. Mamouni, please go ahead.

Thank you, operator. Welcome to everyone participating in this call. Joining me today is Paul Gang, Antalfa's Chief Financial Officer. Please note the following. First, all year-over-year comparisons in today's call are for Q1 2026 versus Q1 2025, unless otherwise stated. Second, consolidated financial statements, including Aurelian, began from Q4 2025. As such, Q1 2025 comparative figures reflect AntAlpha's standalone results. Third, our remarks today will include forward-looking statements based on current expectations. These statements involve risks and uncertainties that could cause after results to differ materially. For discussion of these risks, these refer to Antalpha's filings with the SEC. We do not undertake any obligation to update forward-looking statements, except as required by law. This call also contains references to unaudited non-GAAP financial measures. Reconciliations to the most comparable GAAP measures can be found in our press release and SEC filings. Now, I'll turn the call over to Paul Liang, who will provide the Q1 operating and strategic overview, as well as the financial highlights and outlook. Paul, please go ahead.

Thanks, Chris. Good day, everyone. Thank you again for joining us. I'm Paul Leung, CFO of Enelva. Let me start with a brief framing of the quarter. Q1 2026 was a period of solid execution and business development for Enelva amid a dynamic and challenging market backdrop. for the crypto ecosystem. We deliver 52 year-over-year revenue growth and maintain our record of zero principal loss. At the same time, our long-term books saw a one-time reduction driven by substantial repayments from certain large borrowers, which notably can go in. I will address this in detail, but the key point is All borrowers will pay with no loss of principal, and we view it as a strong reflection of our borrowers' overall financial health and the soundness of our credit model. Finally, we also want two important strategic growth initiatives, the beta launch of our Web3 AI agent and the transition of our tokenization goal holdings into yield-generating deployments. I want to provide some extra context on the Cango repayment before moving to the long book. On the large borrower Cango repayment, I want to provide some extra context here before the financial results section. During the first quarter and into early quarter two, Cango in a NASDAQ listed Bitcoin miner has repaid approximately 530 million USD of its outstanding loan balance. This represents over 95% of Cango's outstanding balance as of December 31st of 2025. Cango funded the repayment through a combination of publicly disclosed Bitcoin SSLs and equity transactions. This is the type of positive positive outcome of that our credit model is designed to produce. So we were pleased with how it all played out in practice. I will now cover our loan book update and risk management activities, followed by our strategy initiatives. Then walk through the rest of the financials and close with our Q2 outlook. NRW's operating philosophy is a risk management first philosophy. We have been consistent in this approach since our inception, and it's central to how we manage the platform in Q1. Bitcoin prices were under considerable pressure in Q1, declining approximately 40% from their October 2025 peak. In this familiar environment, Our approach was deliberate. We maintained active dialogue with every client to review market conditions, stress test positions, and discuss their options as we do in our daily operation, especially every period of price volatility. We did not simply wait for the market to move. We engaged proactively. Our overcollateralization model continues to underpin the longboard. We require overcollateralization at origination, and Bitcoin mine is deposited directly into our wallet, allowing the collateral pool to build continuously. The result of this approach is a proven record. We are proud to stand behind. As of March 31st, 2026, Nalpha has recorded no loss of principal since the inception of the company. And it is the direct outcome of the prioritization of risk management above all else amidst every market with bad job. Before reviewing our loan book matrix, Let me provide some broader market context. We just discussed the devaluation of Bitcoin versus October 2025, which created a more cautious environment for new loan deployment and borrowers' activity. While the near-term sentiment for digital assets has been softer and the long-term demand backdrop remains constructive, spot Bitcoin ETF assets under management stood at approximately $102 billion as of mid-May 2026, reflecting continual institutional participation in the asset class. Historically, periods of price softness have been also coinciding with increased interest in machine upgrade financing as miners began positioning for the next cycle. We expect this dynamic to once again support the loan demand as market conditions stabilize. With that context, let me walk through the loan book metrics and in detail, starting with TVL per client, which I think it gives a clear picture of the underlying business. TVL per client increased 36% year-over-year, reflecting growth in average loan size across the client base and the continued deepening of our client relationships. This growth stems from our proactive strategy to prioritize lower risk consumers, ensuring a higher quality portfolio. Total value of loans were $1.6 billion as of March 26, 2026 and was down 3% year-over-year. This change reflects three factors. First, more measured new loan deployment in a weaker Bitcoin price environment. Second, substantial one-time loan repayment from two large borrowers. mostly from Cango, which we mentioned earlier, which will pay approximately $526 million during the first quarter of 2026 for a modest reduction of approximately 3% in the remaining portfolio on a sequential basis. It is worth we emphasize that We have never had a credit loss. We have never had a loss on principal across the entire loan book. And we enter the recovery phase of the cycle with a well-protected portfolio. As market conditions stabilize, we are positioned to redeploy capital and grow the loan book. Hash rate loans finance approximately 34.2 exahash of hashrate capacity as of March 31st, 2016, representing approximately 3.3% of global hashrate. This compares to 81.3 exahash of December 31st, 2025. The decrease was mainly attributable to CanGo's repayment, as CanGo's facilities were predominantly hashrate-backed loans. In summary, the overall health of our loan book remains quite sound. Shifting gears now to our strategic initiatives, let me cover our Web3 AI agent first, followed by our tokenization goal update. In May 2026, Nalpha launched a Web3 AI agent in public beta. I want to take some time discussing this initiative because we think it is important for the market to understand what we are building, why we are building it now, and how it interconnects into the core of Nalpha. Nina is an early stage of Web3 AI agent and part of Nalpha's broader innovation into AI-driven infrastructure. It is built on our proprietary in-house MCP, or model context protocol framework, which is designed to support intelligent routing and coordination across data and execution environments. As it is called, NINA reflects our view that The next generation of digital infrastructures should make AI and Web3 more intuitive for users to engage with, as AI becomes more deeply embedded in financial and digital systems. Users will need an intelligent interface that can help them navigate increasingly complex blockchain-based environments with greater confidence and convenience. Why now? Three factors come together to make this the right time to introduce a new solution like this. First, the capability of AI models to interpret natural language and reliably translate it into on-chain execution has only recently reached a level where user-facing product is viable. The technical barrier that previously required developer knowledge to interact with blockchain networks can now be adjusted through AI. That barrier has been broken. Second, there's no established market leader in Web3.0 AI today, in particular in Asia Pacific region. So we see a clear opportunity to build a scalable solution that is available now. First, and I want to be specific about this, NRFAS is further leveraging the domain it knows best. We have built a deep operational knowledge of Web3 infrastructure, on-chain data and blockchain systems through years of running our financing platform. Our interconnection within the Bitcoin mining ecosystem give us a unique foundation in the space that other new entrants to Web3.AI does not possess. We are applying existing capabilities and relationships to an adjacent market. And next, I will talk about the tokenization goal. As of March 31, 2026, Nalpha held 39,371 units of XAOT. In Q1, we recognized 12.9 million US dollars in unrecognized fair value gains on these holdings. In April, subsequent to the quarter end, we took the next step in our organization goal strategy by committing six 1,052 units of XAUT to the XAOEU protocol. Aurelien, our subsidiary, separately committed another additional 10,000 units to the same protocol. This is the first deployment of our XAUT holdings into a new generating arrangement, which represents a meaningful transition from holding XAUT as treasury and branched-sheet assets. to actively deploy to generate yield. And now let's move to discuss our financial performance in the first quarter. Total revenue was $20.7 million in Q1 2026, up 52% year-over-year, reflecting a continual growth in both our key revenue components. Technology financing fees on supply chain loans were $15 million, up 49% year-over-year, driven by continued strength in our HHA loan portfolio. Technology platform fees on margin loans were $5.7 million, up 62% year-over-year, reflecting healthy takeaways and continued utilization of our margin loan facilities. The net fee margin increased 21 basis points year-over-year, reflecting price discipline across the platforms. The expansion was led by our margin loan business, where net fee margin improved year-over-year and remained healthy on a sequential basis. Our supply chain loan net fee margin saw a modest year-over-year decrease, driven by a higher proportion of hatchery loans within the portfolio. Hatchery loans carry a lower branded rate than machine loans, and as the mix shifted towards hatchery loans through the period, it created some downward pressures on the branded supply chain loan margins. Now let's turn to operating expense and profitability. Total operating expense excluding unrealized gain of crypto assets were $25 million in Q1, up 102% year over year. This includes funding costs of $10.4 million, one-time restructuring charges of approximately $3.3 million, and non-cash equity-based compensation of approximately $1.3 million. Excluding these two items, non-GAAP operating expenses were $20.4 million. On funding costs, at 79% of technology financing fee on supply chain loans, funding costs increased modestly from 65% in Q1 2025. For additional context, funding costs were 70% of technology financing fee in Q4, 2025. The sequential improvement reflects a reduction in the loan base following a substantial repayment received in the period. Turning to profitability. GAAP operating income was $6.6 million, representing an operating margin of 32%. This was mainly reflects $10.9 million unreliable fair value gain from XAOT holdings, flowing through our operating results. Non-GAAP operating income, which exclude one-time restructuring costs and non-cash ESOP expense totaling $4.6 million was $11.2 million, representing a non-GAAP operating margin of 54%. Net income attributable to NALPHA was $2.7 million in Q1 2026 compared to $1.5 million in Q1 2025. As a reminder, Q1 2025 reflects another stand-alone result, as consolidation of Aurelien began in Q4 2025. Adjusted EBITDA was $13.3 million, representing an adjusted EBITDA margin of 64% compared to 18% in the prior year period. This includes approximately 12.9 million in unrealized gain on the fair value of XAOT holdings. Excluding XAOT-related gains, adjusted EBITDA was approximately 0.4 million with a margin of approximately 2%. The consolidated result I just walked through include contribution from both N-alpha prime the lending business, and Aurelien. To give investors a clear picture of each component, let me briefly separate them before turning to the valuation framework. Nalpha Prime, our core lending platform, generated standalone revenue of $20.7 million, up 52% year-over-year, and standalone adjusted EBITDA of $4.4 million, a 77% improvement from $2.5 million in Q1 last year. Standard loan gap operating loss was $2.3 million, including the $3.3 million one-time restructuring cost and $900,000 non-cash ESOP compensation in the period. If we exclude that charge, standard loan operating income was approximately $1.9 million, compared with $1.5 billion in the first quarter of last year. EBITDA is 12% for stand-alone and upper-prime lending business. Aurelien contributed to the remaining $9.3 million of operating income at the consolidated level, driven by entirely buying XAUT fair value gains. With that separation established, Let me now turn to Aurelien's result and our sum of parts valuation framework. Turning to Aurelien's balance sheet. As of 31st March, 2026, Aurelien's NAV was $116.4 million, or $3.16 per share, reflecting 33,318 units of XOT value at 4,667 per unit, net of $41.2 million in debt. Consistent with prior quarters, NRFAS' 32% economic interest in Orleans represents approximately $37 million of A tributable NAV, deducting this from Nalpha's market cap of $206 million based on last Friday's closing price, this implied value of Nalpha's core lending platform is approximately $169 million, or approximately 1.9 times of traveling 12 months' revenue of $86.8 million and $11.7 trillion 12-month net income attributable to NLFA's $14.5 million. Aurelien's value is tied to gold appreciation and is XAOT's treasury strategy. NLFA's core platform value reflects the lending business fundamentals. We present this framework so investors can better assess each on its own terms. And now turning to Outlook. We expect second quarter 2026 revenue between $11 million and $30 million, excluding the impact of the single customer can go, the year-over-year decline, will be between 7% to 22%. The main driver of this sequential change from Q1 is the reduction in our interest-varying loan base following the one-time repayment received in early 2026. Importantly, net fee margin and operating margin are expected to remain borderly stable quarter over quarter. New loan deployment activity was limited in February through April period, reflecting market conditions. We remain engaged with existing and prospective clients on new loan deployment opportunities and will provide updates as the loan book develops. Our guidance reflects the strong visibility we have in the business and assumes continued demand for crypto collateralized financing in the market environment. Border consistency is what we see today. As always, actual results may differ from our expectation and we will provide updates on our next quarterly call as appropriate. And now let me close with our three priorities for 2026. First, risk management. We will continue to apply disciplined collateral management and maintain active plan dialogue through what remains a volatile market environment. Our zero loss of principal record reflects the way we have run this business since inception. Second, core lending growth. PBL per client growth of 36% year-over-year reflects the quality and engagement of our client base. We are actively deploying capital and we may focus on growing the loan book with the right clients. The demand for crypto collateralized financing is intact and we are well positioned to continue to grow. For additional strategic growth curves, our tokenization growth strategy is advancing. XAOT is now deployed into yield generating protocols, which is a meaningful step forward. And our Web3 AI agent launched this month, applying our Web3 domain expertise to an adjacent market where we see a clear scaling opportunity. Both initiatives are at a very early stage and we look forward to updating investments as they develop. With that, let me open the call to Q&A. Operators, please go ahead.

Thank you. To ask a question, you will need to press star 1 and 1 on your telephone and wait for your name to be announced. To withdraw your question, please press star 1 and 1 again. Our first question comes from the line of Devin Ryan from Citizens Bank. Please go ahead. Your line is open.

Hey, team. This is Noah Katz on for Devin. Thanks for the questions. Lots of developments here to dive into. First, just want to touch on your announced further expansion into AI infrastructure with the Web3 AI agent, Nina. Can you help us unpack this a bit more and tell us why Nina is the right product for your AI strategy and Should we think of Nina as a user acquisition tool and a way to deepen engagement with your clients or as more of a foundational piece for future AI revenue streams? Thank you.

Yeah, thanks, Noah. Thanks for the question. So I think, firstly, we believe LLM has become the next generation of entry point. And now, so I think now we are trying turning to, I mean, all the users interact with like digital work basically through AI to some extent. So this shift kind of create a meaningful value, especially in the AI industry. As today, the users experience remain highly fragmented. Users often need to move across multiple wallets, protocols, data source, analytical tools. So it's a little bit complicated sometimes if they do this on their own, especially for those like the new users. And so we do think that it's a very highly attractive vertical opportunities for us here in the Web3 area. I think it's very difficult for most of the users to navigate around, so this create a very strong user case for AI native products. And we understand the pain point of the users. And Nina is basically designed to bridge this gap. And it provides a neutral, natural language interface to help the users to access that free information and services. So our goal is to lower the barrier to referee participants to support a broader adoption. And we do think that ANALPHA is kind of unique, have some unique competitiveness advantages. So firstly, we developed our own MCP framework, which gives us greater control over the underlying connection between user status and Web3 environments. And over time, this could become an important platform-level capability. And the second, Nina just a real pain point, as mentioned just now in the Web3. Instead of requiring users to set up and manage a lot of different tools separately, NINA basically provides a simpler and more integrated experience. And N-Alpha has accumulated deep knowledge through the years operating in this area. in this ecosystem. We understand data, the workflow, the user needs, and operational complexity of the market, which gives us a very strong foundation to be a differentiated AI product for Web3. And looking ahead, I think the first priority for us is to continue to strengthen the the function of NINA across kind of like the data infrastructure model adoption. And I think we are kind of also, we have been knowing this industry quite well, right? So we have also a lot of connection with our miners. I think it's easy for us to get some to have the resources of the computing power to support our growth. I believe that gives us a kind of unique competitiveness also. In terms of revenue, I think this is not our priority at this moment. I think we will spend more some time to understand the needs of the user. Firstly, I think we need to accumulate some meaningful users before we talk about monetization. Yeah, I hope this helps.

Yeah, that was very helpful. I was going to dive into maybe what differentiates Nina from other Web3 AI agents, but I think I think you answered it, unless you have anything else you want to add there.

Yeah, that's, yeah. Not really. I mean, compared with the other AI agents, I think, I mean, nowadays, you can find a lot of, like, genetic ones, right? But in terms of that 3D, there's not a specific product. at this moment. But we do think some startups are exploring. But we think that we are a listed company. We understand the industry well. I think we are in a very good position to promote this, the NINA, the AI agent product. But definitely, there's a lot of general products outside. use it a lot, but it's not specifically for Web3 users.

Okay, that's helpful. Thank you for answering my questions there. If I could sneak one more in, I want to touch on updates on the Clarity Act as it stands today. How are you thinking about the opportunity for AntAlpha if digital asset rules become clearer and more defined, and could clearer rules change the pace at which you continue to scale? Thanks.

Thanks.

I think it's not that, I mean at this stage I think, so definitely we need to take a second look, I mean a deeper look on how it impacts. But based on my understanding right now, it doesn't affect that much. From our side, we are basically connecting. We are in the infrastructure level. We provide financing to the miners. And it will not create a lot of impact on us at this moment, but definitely with a more on the regulation level. It helps basically the whole ecosystem or the whole state base to develop definitely is going to help us also to some extent in a broader level.

All right. Thank you. Thanks, Noah.

Thank you. We'll now move on to our next question. Our next question comes from the line of Ed Engel from Compass Point. Please go ahead. Your line is open.

Hi. Thanks for taking my question. I just wanted to drill down quickly on the impact of the Kengo loan payment. Did this, did the entire $500 million get paid off in the first quarter or was some of the impact also in the early second quarter?

Yeah, I think mostly it's paid in the first, I think we mentioned roughly like just a small portion was retained in April the second quarter.

Great. Thanks for that. And then on the XAUT yield generation deployment, just kind of curious, can you just give more color on I guess how those tokens are generating yield and I guess the expected yield or target yield that you're hoping for the allocation? Thanks.

Okay.

For the yield generation, basically, so the foundation issuing the XOT, they have some kind of like very conservative investment opportunities using the existing XOT. which, based on our estimation at this moment, is roughly between 1% to 2%. It's not that much, but anyway, it's a step forward, comparing with the pure holding of XAT in the past.

Great. Thank you for the call.

Thank you. We'll now move on to our next question. Our next question comes from the line of Dylan Hines from B Reilly Securities. Please go ahead. Your line is open.

Hey, thanks for taking the question. I'm just wondering, on top of Cangle, do you expect any other repayments coming up? And then I just have a follow-on question after that.

I think the loan balance at this moment is roughly $40 million, and we don't anticipate any repayment at this moment, but we never know. But anyway, the impact will not be significant since the balance has been reduced significantly compared with the previous quarter.

Gotcha. Thank you. And then I was wondering, you mentioned initiative in the Asian Pacific. I was wondering if you could go over again what you were talking about there.

Can you kind of ask the question again?

Yeah, there's an initiative you're talking about in the Asian Pacific regarding blockchain operators. I was wondering if you could go over that again. miss that.

You mean the AI agent or any other initiative we talk about?

I'm not sure. You said something about agent specific, but I'll have to go over it again, but okay. So nothing else for me then. Thank you.

Okay. Thanks.

Thank you.

That concludes the questions and answers period. Thank you again for joining our call today. You may now disconnect.

Thank you.