Nvidia’s stock is showing signs of life ahead of earnings next week. The move is justified and should have more room to go. Shares of Nvidia on Monday closed around $225 a share for the second session in a row — a closing level that previously hadn’t been since mid-May. It’s a sign that market sentiment toward the leading AI chipmaker has warmed following a chilly reception for much of this year. Among the reasons for the change: Nvidia’s increased financial support for key customers is looking less risky than initially feared, and a new financing initiative should make funding the AI buildout more attainable. Plus, new details on the blistering revenue growth of OpenAI and Anthropic — big users of Nvidia chips — bodes well for their ability to keep spending on compute in the future. Nvidia’s recent advance has now pushed its year-to-date gains north of 20%, outperforming the tech-heavy Nasdaq Composite , which is up about 15%, and the S & P 500 , which has climbed about 13%. “I think [Nvidia] can continue to climb,” Jim Cramer said on Monday’s Morning Meeting. The company will release its fiscal 2027 second-quarter results next Wednesday, Aug. 26. To be sure, Nvidia’s multiweek rally coincides with a broader comeback in the AI infrastructure trade after the forced unwind of the Situational Awareness hedge fund. From the recent market bottom on July 29, Nvidia shares are up about 18%, versus the iShares Semiconductor ETF’s 2% advance and the VanEck Semiconductor ETF’s almost 18% advance. Nevertheless, it’s actually encouraging to see Nvidia keep pace with these two baskets of chip peers during this stretch. That’s because the stock has trailed them for much of the year , as money rushed into memory and central processing unit (CPU) bets, and the sustainability of Nvidia’s growth faced renewed questions. NVDA SOXX,SMH YTD mountain Nvidia’s year-to-date stock performance versus the SOXX ETF and the SMH ETF. While it’s easy to suggest the blowup of a major leveraged player is what helped the stock bottom, that won’t be enough to keep the move going. A trio of funding-related updates that we’ve gotten since then, however, may well be. The latest arrived Monday, with Nvidia officially announcing its support for a massive data center project in Ohio, where ChatGPT creator OpenAI will be the tenant via a 20-year lease. The energy subsidiary of Japanese conglomerate Softbank, known as SB Energy, is leading the development and will own the facility. Nvidia will be the exclusive compute provider, according to a press release . The details revealed Monday suggest Nvidia is very sensitive to investors’ concerns regarding its own financial backing for the AI buildout. Consider: On July 26, a Sunday night, the Wall Street Journal reported on the 10-gigawatt Ohio campus, saying that Nvidia was in talks to “provide a roughly $250 billion backstop” for the project. The report, which CNBC later confirmed , was at least partially responsible for the roughly 5% decline in Nvidia’s stock the following day. It added to some investors’ concerns about “circular deals” within AI. What we learned Monday is that the backstop is actually going to be for only $105 billion, as it will focus on the initial 4.25 gigawatt build, not the full project. And, to help reduce funding costs, Nvidia plans to invest $1.5 billion into SB Energy, which could go public as soon as next month, the Journal reported Monday . Along with less at-risk capital should the project not work out, Nvidia is only on the hook for completed data centers, not those still under construction, according to the Journal. Moreover, the backing is for the assets in the data center and the facility itself — this means Nvidia is not on the hook for lease-payment obligations made by OpenAI. The Journal said Nvidia’s backing is only triggered after a series of steps. If OpenAI backs out of the lease, SB Energy will look for a replacement tenant at the same price; if SB Energy cannot locate an adequate replacement, it would look to sell the campus — and at that point, Nvidia would need to pay any difference in value, up to $105 billion. Importantly, in this way, the $105 billion backstop depends heavily on the residual value of the chips. Because Nvidia creates and executes on the product roadmap that impacts the value of its older chips, the company has unique insight into the risk being taken on here; in theory, CEO Jensen Huang and CFO Colette Kress only agreed to this because they believe there’s minimal risk that the $105 billion is ever called upon. There’s a lot to chew on, so let’s step back. Rather than providing $250 billion in backing, as was reportedly considered at first, Nvidia is providing $105 billion in backing, along with a $1.5 billion investment into the site’s owner. Moreover, that $105 billion is backed by the asset value of a 4.25 gigawatt data center, which, at an estimated $50 billion to $60 billion per gigawatt, implies an initial value of more than $200 billion. Also, the backing is only triggered should SB Energy fail to find a new tenant at the same price and then is unable to sell the site for a fair value. The point is that Nvidia’s exposure to this deal appears to be far less than feared, without much hit to the potential upside, considering it now has the SB Energy stake and the fact that the site appears to be planning to host exclusively Nvidia technology. This comes on the heels of Nvidia’s $500 billion financing initiative announced last week, in partnership with some of Wall Street’s most influential names: Apollo , BlackRock , Blackstone , Brookfield , Club name Goldman Sachs and KKR . This effort — where the firms will essentially securitize compute and offer up asset-backed securities to investors — further serves to de-risk the buildout. To be sure, while vast sums of money are still at risk, the securitization of compute means that the funding risk moves out to the institutional investors providing the money, rather than weighing on the cash flow and balance sheets of those leading the buildout, like the hyperscale players. The result is that the funding needed to build more data centers, to meet booming demand for AI compute, will still be received. But instead of companies adding tens of billions to their capital expenditure budgets, we’re starting to get the financial structure needed to tie the cash flows from these “AI factories” directly to the debt holders. Though the idea of securitizing compute is still early days — and we’ve yet to learn all the details on how Nvidia and its Wall Street partners will bring this concept to market — it certainly provides the groundwork needed to reduce the pressure on infrastructure players to fund the buildout themselves, be it with internally generated cash flows or, increasingly, with debt and equity offerings. The third piece of positive funding news to help propel Nvidia higher is the strong financial performance of OpenAI and Anthropic, the biggest and most important AI labs in the world. Late Friday, Bloomberg News reported that Anthropic’s second-quarter revenue jumped 14-fold on an annualized basis, to more than $11.5 billion, and that it recorded positive adjusted operating income. A day earlier, the news outlet reported that OpenAI was on track to do about $40 billion in annualized revenue this year — almost double its run-rate exiting 2025. The most obvious risk when it comes to funding the AI buildout the counterparty risk. If a company that takes on obligations can fulfill those obligations, there isn’t much to worry about. That, however, is a big if when the company in question is private — meaning no publicly disclosed financials — and is believed to be unprofitable and burning cash, perhaps faster than any startup in history. Here, we’re specifically talking about OpenAI. Anthropic’s finances have generally been considered healthier, though we’re still relying on press reports for these details, so the picture is incomplete. Risks around OpenAI’s financial health remain, but last week’s report should serve to reduce — though certainly not eliminate — some concerns regarding OpenAI’s ability to meet obligations and, in turn, the risk Nvidia’s financial backing is ever actually needed. Sure, you can’t pay debt with revenue alone, but rapid revenue growth is a key factor when it comes to raising debt and/or selling equity — be that in private markets or when talking about a potential future initial public offering (IPO). Both OpenAI and Anthropic have submitted confidential IPO filings with regulators, though exact timing on their potential offerings remains unclear. The bottom line? Nvidia’s stock certainly benefited from the deleveraging event sparked by the unwind of Situational Awareness. That helped investors see the sell-off in AI infrastructure stocks that began in roughly late June and continued into July in a new light — specifically, the idea that the steep declines had reflected weakening fundamentals quickly evaporated. However, the improved funding dynamics set the stage for further upside from here. (Jim Cramer’s Charitable Trust is long NVDA. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
