The $100,000 Check and the $12 Million Stock

A Vancouver-based public company announced today that it had invested $100,000 in a startup building data centers in space. The startup is valued at $2.3 billion. The public company is valued at about $12 million.

That gap is not the point. The point is what sits between them: a publicly traded vehicle designed to give ordinary investors exposure to private venture deals they cannot access any other way. Whether that vehicle creates value or merely sells the idea of value is a question worth asking — not just about this company, but about the whole model.

Planet Ventures trades on the Canadian Securities Exchange and on U.S. OTC markets. It describes itself as an "investment issuer" — a public shell that raises capital to buy stakes in private startups across space, energy, gaming, and mining. It makes roughly $2 million a year and carries negative enterprise value, meaning its cash on hand is nearly equal to what the market says it's worth. On paper, you are not buying a business. You are buying a portfolio of stories.

The Starcloud investment illustrates the model in miniature. Planet Ventures placed $100,000 into an SPV — the ST-0504 Fund I — that participated in Starcloud's $250 million Series A extension at a $2.3 billion post-money valuation. The round was led by Manhattan West and included NVIDIA, Cisco, Benchmark, and EQT. Planet Ventures' slice is roughly 0.04% of the round. If Starcloud succeeds spectacularly, that $100,000 could multiply. If Starcloud fails, Planet Ventures loses the $100,000 and nothing more — while the press release has already done whatever work it was meant to do.

Which raises the question: who does the press release serve?

Starcloud itself is an interesting company. Founded in 2024 by engineers from SpaceX, Microsoft, and Airbus, it has already launched Starcloud-1 — a refrigerator-sized satellite carrying an NVIDIA H100 GPU, the first data-center-class GPU placed in orbit. It trained an AI model in space. The FCC has accepted Starcloud's filing for a constellation of up to 88,000 satellites dedicated to orbital computing. The company has raised roughly $450 million in total.

The physics are real. Space offers continuous solar exposure in sun-synchronous orbit and passive radiative cooling in a vacuum. Earth-based data centers are hitting genuine constraints: grid interconnection queues of five to seven years, water scarcity, community opposition, and capital costs pushing past $20 million per megawatt of compute capacity. The argument for orbital data centers is not frivolous.

The economics are the harder problem. Right now, a GPU cluster in space costs about four times more to operate than an equivalent cluster on Earth. The levelized cost per GPU-hour is roughly $10.91 in orbit versus $2.49 on the ground. The single biggest cost driver is not the chips — it is the radiators. In a vacuum, heat can only escape through radiation. A modern GPU rack generates about 120 kilowatts of waste heat. Launching enough radiator mass to reject that heat costs between $6 million and $36 million — for one rack. That is before you add solar arrays, structure, shielding, communications, and redundancy. Radiation in orbit degrades hardware, forcing operators to build in roughly 20% spare capacity just to maintain uptime.

Industry analysis projects cost parity between space and terrestrial compute around 2040 in a base case. That could come sooner if launch costs drop dramatically — from roughly $1,400–$1,800 per kilogram today to about $250 per kilogram with Starship. But even that requires Starship to achieve high cadence, reliable operation, and the capability to deploy massive radiator arrays. In the most optimistic scenarios, space compute is only about 30% more expensive in the early 2030s.

The timeline matters because Starcloud is still a prototype company with 15 employees and one satellite in orbit. It has no revenue. Its business model depends on launch costs falling by a factor of six to eight, radiators getting cheaper and lighter, hardware surviving years of radiation, and someone choosing to pay a premium for compute in orbit rather than waiting for the terrestrial supply chain to clear. All of these are open questions.

And this is where we come back to Planet Ventures.

The company's stated thesis is simple: retail investors cannot access deals like Starcloud, so Planet Ventures will access them on your behalf. The SPV structure isolates risk — each investment is ring-fenced. The company takes positions in early-stage companies across sectors and hopes that one or more become large enough to lift the stock price. Alongside Starcloud, Planet Ventures has been accumulating Bitcoin, reporting holdings of nearly 26 BTC as of July.

The problem is not that the thesis is wrong. The problem is that the structure makes it impossible to tell whether the thesis is working.

Consider what an investor in Planet Ventures actually owns. They own shares in a company worth about $12 million that holds a tiny stake in a company worth $2.3 billion. The $100,000 investment is so small that even a tenfold return in Starcloud would add roughly $900,000 to Planet Ventures' portfolio — barely meaningful against a $12 million market cap. And the connection is indirect, through an SPV, through a Series A extension round led by institutions that had $250 million to deploy.

The more interesting question is what the company is actually good at. Is it sourcing deals that big investors miss? Is it adding value through operational support? Or is the business model mostly about creating public-market awareness around private companies, using press releases to drive retail buying pressure in a stock with very thin liquidity?

A stock that moved from $0.03 to $0.25 over its last 52-week range, with just 5.15 billion shares outstanding, is not priced by fundamentals. It is priced by attention. And attention is a scarce commodity. Planet Ventures' product may not be venture returns at all. It may be the feeling of being connected to something big.

That is not to say the company is a fraud. There is a genuine gap between private venture markets and retail investors. If Planet Ventures can source deals, deploy capital efficiently, and generate real returns across its portfolio over time, the stock could prove to be a bargain. The barrier is that you cannot evaluate that thesis from the outside. The company does not disclose enough detail about its SPV terms, its fee structure, its carry arrangements, or the aggregate performance of its investment portfolio. You are asked to trust the pipeline without seeing the plumbing.

What would change my mind? A few things. First, transparency: detailed disclosure of each SPV's size, Planet Ventures' ownership percentage, fee and carry terms, and the portfolio's aggregate unrealized value. Second, concentration: the company would need to deploy enough capital in each deal to make the outcome material — $100,000 in a $250 million round is not a position, it is a receipt. Third, track record: one or two investments that actually generated outsized returns, verified by third-party data, would prove the model works.

Until then, the Starcloud investment is a useful lens. It shows a public company writing a check it can afford to lose, generating a press release about a company most investors cannot invest in directly, and creating the impression of access without the substance of it. The startup may turn out to be brilliant. The space data center thesis may come to pass. But the relationship between the $100,000 check and the investment case for the $12 million stock is almost nothing at all.

The test for any investor is simple: if you strip away the story about what Planet Ventures might invest in next, what is left that justifies the price you pay? That is the question the company needs to answer.