Summary Hut 8 Corp. is pivoting from volatile Bitcoin mining toward a scalable, contract-driven AI and energy infrastructure platform. FY25 revenue rose 45% to $235.1M, with Compute contributing $202M and gross margins expanding to 54%, yet net loss reached $248M due to crypto exposure. A $7B, 15-year AI lease with Fluidstack and an 8.5 GW pipeline signal a shift to utility-like, stable cash flows, but execution risk remains high. HUT stock valuation is rich at 17–25x EV/sales, reflecting high expectations for successful transition and future profitability despite near-term volatility. Thesis Hut 8 Corp. ( HUT ) is trying to transition from a volatile Bitcoin miner to a long-duration, AI and energy infrastructure platform. Now, I see it as a good thing, but the aim has to be more towards predictable, contract-driven cash flows. With the support coming from high-efficiency and their vertically integrated data centers, which take advantage of some disciplined project-level financing. The upside here is that if Hut 8 executes on their multi-gigawatt pipeline and flagship AI deals we saw in 2025, whilst keeping the build cost-efficient, it could convert future revenue into utility-like, infrastructure-style earnings. This would be needed to justify the premium valuation despite near-term losses. Hut 8’s management is well aware that AI-driven compute demand is still pretty robust despite the broader market noise we’re seeing. It puts management’s focus on relationships with strategic partners like Fluidstack/Anthropic. They need to use these partnerships to align their capacity expansion with actual customer demand going forward. On the financial side, the FY25 earnings showed us some of this operational strength, with revenue growing 45% year-over-year to $235.1 million, and their Compute segment alone now driving $202 million of that growth, which is good to see. Gross margins managed to expand to 54%, which goes to show the improving unit economics, which I’ll explain. However, the bottom line here is still extremely pressured. There was a net loss of $248 million, largely from about $220 million in unrealized Bitcoin (BTC-USD) losses, compared to the significant gains we saw back in 2024. So the key worry for me is that whilst Hut 8’s core operations with the new AI-focused infrastructure are improving markedly, the headline earnings are still being distorted by crypto exposure. So I think this will mask the underlying operational progress and long-term potential of the business in 2026 as well. So I do expect more volatility in the stock price; however, in the longer term, there is clearly a lot of upside. FY25 earnings review As for end-of-year earnings , revenue grew quite meaningfully to $235.1 million, up about 45% since last year, which we can put down to being mainly driven by Compute, which brought in $202 million alone. Hut 8 Corp. So they’re scaling their activity in ASIC mining and cloud services. However, their profitability, on the other hand, swung sharply negative, with a net loss of $248 million compared to a $331 million profit back in 2024. Now, the big driver here wasn’t core operations, but rather large unrealized losses on some digital assets, about $220 million compared to massive gains we saw the year prior. Hut 8 Corp. This tells me Hut 8’s earnings are still heavily tied to Bitcoin price movements and accounting revaluations, and as you know, these can obscure underlying business performance. Even Adjusted EBITDA turned negative at -$135 million, whilst last year we saw a $556 million profit in Adjusted EBITDA. That goes to show just how much last year’s profitability depended on favorable crypto market conditions rather than purely operational strength. On the operations side, we are seeing the company make some pretty aggressive moves that could reshape its long-term profile. The standout is the $7 billion, 15-year AI infrastructure lease with Fluidstack. It's being backstopped by Google and would signal a pivot into high-demand AI data center capacity. Now, this would definitely add a more stable, contract-driven revenue stream than mining. Hut 8 has also gone about streamlining its capital structure. They’re selling a 310 MW power portfolio and launching American Bitcoin as a separate vehicle. So effectively that it should isolate the volatile mining exposure. They also hold an 8,500 MW development pipeline and access to structured financing, including the likes of JPMorgan and Goldman Sachs. So, going forward, I see Hut 8 is positioning itself more like a hybrid of a power developer and digital infrastructure provider. The tradeoff here is quite clear: near-term financials will look weak and volatile, but the company is trying to lay the groundwork for long-duration, infrastructure-like cash flows tied to increasing AI demand. It makes the risk going forward execution, and turning that massive pipeline and flagship projects like River Bend into predictable earnings before capital intensity and their hefty crypto exposure continue to weigh on results this year. Cost efficiency and capital structure With the ongoing losses, Hut 8’s capital structure transformation is going to be very important going forward. I see it as the backbone of management’s long-term strategy. What they’re doing is deliberately shifting away from traditional corporate-level leverage toward a lower or non-recourse, project-level form of financing, which I think is good to see. Debt is being secured against their individual assets and their contracted cash flows rather than the broader balance sheet. First, this approach should protect equity holders somewhat from downside risk, but the second thing here is that it also enables significantly higher leverage at the asset level. We just saw some finance structures reach roughly 85% loan-to-cost. Now, that level of leverage would be pretty difficult to justify without strong counterparties and predictable revenues, which is why I see directly tying into the company’s parallel push toward long-duration, investment-grade contracts. Hut 8 Corp. Hut 8 is starting to lock in these contracts, and lenders are now gaining confidence in the stability of future cash flows. So, going forward, we should see a lower cost of capital and a more efficient recycling of equity across the multi-gigawatt pipeline. The overall aim here would be to translate this disciplined financing approach into a much wider investment-grade profile. That should help compress financing costs and reinforce that cheaper capital enables more competitive project bids/higher returns. Elsewhere, Hut 8 is also trying to cut the cost side of the equation. What we want is each deployed megawatt to generate superior economics. The Vega data center design is a decent example of this in practice. They’re set to deliver about 180 kW per rack, which was actually ahead of Nvidia’s prevailing assumptions at the time. So this set-up would imply a significantly higher compute density per unit of infrastructure. Thus, essentially increases revenue potential per site. The key idea there is compute density, in how much computing power they can fit into a given space. At 180 kW per rack, I see Hut 8 running far more GPUs per rack than traditional setups, meaning more compute output from the same physical infrastructure. At the same time, they reported a $455,000 per MW build cost. That’s actually notably efficient for high-performance compute infrastructure, and I see it as being a broader emphasis on vertical integration with an in-house design and value engineering. What they’re trying to do is control more of the development stack, so from design to construction. In doing that, Hut 8 is essentially reducing reliance on third parties and shortening the build timelines. It would also standardize deployments, which would contribute to faster time-to-revenue and lower capital intensity. I think of it as Hut 8 is effectively turning data center development into a repeatable, almost manufacturing-like process. We also have to tie these cost advantages back to the financing model, in that they somewhat enable the good financing terms. You see, lower build costs improve project-level returns, which should then support the higher leverage they’re seeing. The good part is that this will reduce equity requirements and amplify the overall return on invested capital, but the downside is execution risk. Valuation In terms of Hut 8’s valuation , I think the company is being priced far more on future potential than current fundamentals, which is consistent with what we saw in the earnings, in that there’s clearly a strong narrative, but weak near-term profitability. On that note, the stock looks rather expensive, with a trailing EV to sales ratio at 24.9x and a forward ratio of 17.2x versus a sector median around the 3x range. It's a hefty premium, even for high-growth infrastructure or AI-adjacent names. Elsewhere, the price-to-sales sits at 21x and goes to show how much investors are willing to pay a steep price relative to current revenue, especially considering that a large portion of that revenue still comes from a very volatile compute/Bitcoin segment rather than long-term contracted infrastructure. What I see the market pricing in here is a successful transition into a high-quality, AI infrastructure and power platform. Eventually, we should see durable, long-term cash flows. The premium would also imply expectations that projects like the $7 billion Fluidstack deal, their broader 8.5 GW pipeline, will convert into stable, somewhat utility-like earnings streams over time. In other words, investors may be valuing Hut 8 less like a cyclical crypto miner and more like a next-generation digital infrastructure developer, closer to a data center REIT or power developer, which is good to see. But the risk embedded in this valuation is execution, and to justify these multiples, Hut 8 must not only grow revenue pretty rapidly but also shift that portfolio mix toward contracted, predictable cash flows. We would also need to see positive EBITDA in the near term; otherwise, the current premium leaves little margin for error. Risks As you know, there are a few risks tied to this transition. Execution risk is the big one I mentioned. It’s high for large-scale projects like the Fluidstack deal, where delays or cost overruns could strain their finances. The business is also very capital-intensive and has to rely on timely access to power and structured financing to scale. This is where energy cost spikes could constrain operations going forward. We’re already seeing competing data center players such as Applied Digital ( APLD ) invest in their own power supply. Finally, the shift away from Bitcoin revenue always adds transition risk. In a downside scenario, AI adoption or project execution falters, and Hut 8 could see gaps in cash flow that were previously cushioned by mining profits. That would magnify the near-term volatility we’re already seeing. Looking ahead So the long-term strategy is more structured around a clear three-phase roadmap here. The main focus is to transform the company from a Bitcoin miner into a scalable energy and AI infrastructure platform. 2026 execution is going to be critical. With a focus on delivering River Bend on time and on budget, whilst also converting their multi-gigawatt pipeline into contracted revenue. Successful execution here would validate the company’s transition. Again, my long-term buy case here is that Hut 8 can generate high-margin cash flows despite near-term earnings volatility.
