Tech Stocks – Investment Ideas & Trades
Q2 Earnings Season
Neo AI clouds are currently the highest beta play on the AI investment cycle. This is where AI startups are renting GPU capacity to train and infer their models, and clearly revenue growth is exploding here. This is Nebius’ CEO on the recent call:
“We more than doubled our revenue for the whole group from Q1, and this quarter, we also became EBITDA positive in our core AI infrastructure business ahead of our previous projections. We could grow faster, but we were oversold on all of our supply of previous generation Hoppers, and we decided to wait for the new generation of GPUs to come. New Blackwells are coming to the market in masses, and in parallel, we are dramatically increasing our data center capacity. That's why we are increasing our ARR guidance for the year-end from the previous $700 million to $1 billion, to now $900 million to $1.1 billion.
By the end of this year, we expect to have secured 220 megawatts of connected power that is either active or ready for GPU deployment, and this expansion includes our data centers in New Jersey and Finland. In addition, we have nearly closed on 2 substantial new greenfield sites in the United States and overall, we are in the process of securing more than 1 gigawatt of power by the end of 2026. Also, we continue to significantly expand our customer base. We started to gain real traction on the enterprise side, adding large global technology customers such as Cloudflare, Prosus and Shopify. And we still remain a leading new cloud provider for so-called native AI tech startups.”
A potential drawback for the neo-clouds is that their business will be highly sensitive to the AI startup cycle – as long as money is pouring into AI startups, neo-clouds will do well. But these types of gold rush moments don’t tend to last. The bull case for the neo-clouds is that a number of their clients will become big winners in the AI era, grow into large enterprises, and then run their inference workloads on the neo-clouds. Obviously, there is a risk that more workloads move to the big clouds over time due to the vast amount of services and tools available on these platforms such as data management, analytics, data streaming, all sorts of compute, etc. Data gravity and reduced software engineering complexity are two reasons why a lot of companies opt for one or two clouds.
While it’s good news for Nebius that they’re also winning large enterprises such as Cloudflare, again, there will be a lot of startup exposure here. Startups are renting GPU capacity at Cloudflare for inference and Cloudflare is diverting this traffic to Nebius as they don’t have enough capacity. That being said, it’s definitely good news that they’ve attracted Shopify and Prosus as customers. This is Nebius’ CRO on why they’re able to win large customers now:
“As we bring on larger clusters, we are able to bring on new large customers who want to purchase greater and greater capacity. This allows us to expand and diversify our customer base. The demand environment in the second quarter, as you can tell from our results, was very strong. As we brought on more capacity, we sold through it. And by the end of the quarter, we were at peak utilization. If we had more capacity in the second quarter, we probably would have sold more as well. At the same time, we were able to improve the maturity of our platform, which has contributed nicely to increasing our competitive win rate, all of which is continuing on into this quarter.”
As Nvidia Blackwell Ultras will get delivered in Q4, Nebius is seeing strong revenue acceleration next year, and this will show up in ARR already this year:
“The increase in our ARR guidance reflects the strong demand we are seeing in the expected delivery of additional GPU capacity later this year, particularly the Blackwell Ultras. Because much of this capacity will come online by the end of the year, the impact will show up more in ARR than within annual revenue. That timing dynamic is why we are holding our 2025 revenue guidance steady. That said, this late year ramp will create a strong foundation heading into 2026 and will support meaningful revenue acceleration next year.”
More good news is that Nebius is looking to win major contracts from frontier AI labs and hyperscalers, as hyperscalers are still not able to sufficiently get their hands on Nvidia capacity. This is Goldman on this opportunity:
“Notably, management emphasised that this outlook does not include any major contracts from frontier AI labs or hyperscalers, both of which are areas where the company is actively engaged in discussions; any such wins would represent incremental upside beyond the current guidance. Looking ahead, we model revenue of $5.9bn by 2030 in our base case, and $8.4bn in a bullish scenario, contingent on market adoption, enterprise penetration and further customer wins across global AI infrastructure markets.”
CoreWeave is much larger with already $1.2 billion in quarterly revenue while Nebius is merely at $105 million. Nebius’ balance sheet is definitely healthier than Coreweave’s – while Coreweave is already $13.5 billion in net debt, Nebius still has $500 million in net cash. But at least CoreWeave is telling investors how much cash they’re burning every quarter – for example, $2.7 billion last quarter alone. For Nebius, we can look at the company’s cash and debt balances over time to calculate the number – i.e. $958 million last quarter and $3.7 billion over the last 12 months.
We’re not uninterested in these businesses – if long term, neo-clouds can attract a lot of sustainable inference and training workloads, clearly this is attractive business. However, in the meanwhile, we’re not fully convinced on the economics. For example, Nebius mentioned it takes 2 to 3 years to break-even on the gross profit level for their Hoppers, which obviously doesn’t include their SG&A and R&D expenses, both of which are large. On an analysis from Deutsche on CoreWeave, the positive NPV on their GPU investments only really came through in year 5. To be honest, numbers like these don’t strike us as great economics. And with these companies’ currently huge cash-burn, we’re remaining on the sidelines in these names. Nebius is already valued at $17 billion for a company which will be doing $567 million in revenue this year while having burned $3.7 billion in cash over the last 12 months.
We still think the risk-reward makes much more sense in a name like Microsoft, where Azure is probably the best business in the world. We’d much rather pick up safe, double-digit returns here than chase current momentum in the neo-clouds where long term returns and economics remain questionable.
For premium subscribers, we tour 7 investment ideas in the tech sector, all of which we’d rather hold than the neo-clouds. For each one of them we’ll do a similar analysis as above, drawing from interesting insights and data points we came across over the last week, and with thoughts on whether the risk-reward is looking attractive here.