Ryan Miller
China holds the fusion record. America holds the only licensed fusion plant on planet Earth. Only one of those is something a fund can actually buy. So in this week's episode of Making Billions, you will get the six layers of the fusion stack and what each one actually sells. A scoreboard of who is ahead with the trigger that would flip each row. The arithmetic on what waiting costs a private equity or venture fund. And the one thing about China's fusion program that almost nobody in finance has said out loud. And I'm not handing you a vibe, I'm handing you the report number, the public law, the budget line, and the name of the agency that signed the license. Check everyone before your coffee gets cold. I've helped over 2,000 people to launch and scale investment funds, and I'm doing that for myself at all.
Ryan Miller
So I have the privilege of sitting across from investors interviewing some of the brightest minds in the business. And the question they ask is never about the technology. Here is the first plan. First, I'll kill the story you were told about in 2022. Then the two races and the scoreboard. Then the other half of this race you cannot place a bet in. And then a man who nearly went under learning the difference between owning a permission and owning a plant. All this and more coming right now. Here we go.
Ryan Miller
Before we dive in, just a quick disclaimer. Nothing in this episode is legal, financial, tax, or investment advice. This is just for entertainment purposes, and you should always check with accredited professionals before making any decision from this show or otherwise. Now, with that said, let's dive in.
Ryan Miller
You were told that America achieved net energy from fusion in December 2022. That's not exactly what happened. Here's what I think happened. The National Ignition Facility put 2.05 megajoules of laser energy into a fuel target. I got 3.15 megajoules back. That ratio is real. It is peer-reviewed and it is a genuine scientific first. Nobody's taking that away. Now, the number that was not on the screen, the government accountability office wrote it down in March 2023. Their words, operating the lasers to run the experiment required about 100 times the energy released by the fusion event. So the experiment did not produce more energy than it consumed. The same report puts down the power drawn to run those lasers is about 300 megajoules. 300 in, 3 out. You see the problem? So let me put that in plain English.
Ryan Miller
There are really two ways to score a fusion shot. One compares the energy out to the energy that hit the fuel. The other compares it to the electricity you bought from the utility. The lab reported both. The press reported one. And here's the line that is checkable and binary. As of today, no fusion machine anywhere on planet Earth has put a single net kilowatt hour into any power grid. Not America, not China, no one. And that matters more than it sounds. Because every valuation in this sector, every headline about a race, every deck that lands on your desk is priced off a thing that has not happened yet anywhere. And that's not a reason to stay out, but it is a reason to stay precise into what you're actually buying. So if America is ahead in fusion, that shot is not why.
Ryan Miller
Now, think about in your mind what you took from that. Because here is what I took. When the loudest fact in his sector turns out to be half a fact, go look at what moved while everyone else was staring at it. That's what I took from that. There are two fusion races. They have different finish lines and different winners. China is winning the science race, and it is not close. China's East Tokamak held a plasma for 1,066 seconds on January 20th, 2025. Its best tokamak had been under construction since June 2023 and reported $2.1 billion, targeting its first plasma in 2027. America is winning a different race. The race to be a plant, a bank will finance. Because a plant does not get built when the physics works. It gets built when four things exist at once. A license, a site with a grid hookup, a buyer under contract, and somebody willing to carry the risk of building it. Not one of those is a physics question. So here's the scoreboard. Eight rows, and every row has a flip trigger. That is the single event that would move it from one side to the other. That is what makes it a scoreboard instead of a watch list.
Ryan Miller
China leads five rows. Plasma performance. The flip is an American machine matching that thousand-second run. Machines under construction is the other one. The flip is best slipping past its 2027 target. Next, we got government funding. The flip is Congress funding to the level the Chips and Science Act already authorized. So we want to look for that when we're looking to possibly position or to profit from this. Workforce, where China graduates roughly 10 times the fusion PhDs. And supply chain control, which is the row that should worry you most. China put Yttrium and Gadolinium under export license in April 2025. Those are the rare earths in the tape that makes fusion magnets. And only Russia and China still make enrich lithium-6. That's the material you breed fuel with. So America has made none since I think 1960s. So the flip on that one is an American or Allied lithium-6 line getting built. We want to look for that. And America leads in three rows: licensed power plants, binding contracts, and, of course, private capital. Five to three, that looks like a loss. But look at which three, because that matters. See, on June 16th, 2026, Helion Energy got two licenses for a fusion power plant in Malaga, Washington. And according to the company and the American Nuclear Society, those are the first licenses in the world for a fusion power plant. They came from the Washington State Department of Health, not the Nuclear Regulatory Commission, a State Health Department.
Ryan Miller
There are four fusion power contracts on Earth. Only two are clearly binding with the size we've already discussed. All four are making American power plants. And private money in July 2026, the Fusion Industry Association counted 56 fusion companies worldwide. 28 of those 56 are American. That includes all five that have raised over a billion dollars. China has four companies on that list. Can you see the difference in scale? So license plans, binding contracts, private money. Those are the three questions a lender asks before it funds absolutely anything. So read it this way: China is winning the science. America is winning the part that gets built and paid for. Two races, two finish lines, and only one of them has a ticket that you and I can buy.
Ryan Miller
Now, the part almost nobody in finance has said out loud. And it is the whole reason I made this episode. You cannot buy China's lead, not because it's illegal, because there's nothing to buy. On July 22nd, 2025, China set up a national fusion company in Shanghai, China Fusion Energy Company Limited, registered capital of 15 billion yuan at 50.35%. Kunlun Capital, which is a part of PetroChina, at 20%, the Shanghai Future Industries Fund and the National Green Development Fund at 15% together, China National Nuclear Power Plant at 6.65%, and a provincial power company at about 5%. See, every single one of them is a state entity that matters. And here's where most of the commentary gets it backwards. So let me be precise, because this actually matters. There is no American law stopping you. The outbound investment security program is Treasury's final rule of October 28th and 2024. It covers semiconductors, artificial intelligence, and quantum computing. Fusion is not named. The America First Investment Policy Memo of February 21st, 2025 tells Treasury to consider adding more sectors. Nuclear and fusion are not named there either. So there's no ban to be angry about. There is simply no share in existence for a U.S. fund to buy.
Ryan Miller
So what this means to you is that China's progress is an input to your pricing. It is not an asset in your book. A record plasma in Hefey can move the value of what you hold in America through sentiment, through policy, through the price of the magnets. It can never be the thing that you hold, unfortunately. You watch China to time and seize your American exposure. You do not watch it to buy China necessarily. So what do you do actually with all of that? Well, three things. You stop treating Chinese milestones as a buy signal because there is nothing on the other end of that trade. You start treating them as a pricing input on the American names that you can actually hold. You watch the choke points instead of the headlines. If China restricts another input the way it restricted yttrium and gadolinium, that hits the cost of every American developer on that list. That is a real transmission channel. A record plasma is not. And you stop paying a premium for fear.
Ryan Miller
When a sector is priced off of national decline, the story itself becomes part of the price. That is a part you can be wrong about, and it is most expensive. And while we are being careful with the numbers, you will hear that China spends $1.5 billion a year on fusion against America's $790 million. So you can see, we know that in financial analysis, when you're overspending on RD, you can extrapolate that forward to say likely if they hit something and odds are higher that they do when they hit it, then that might result in future value. And so that is an early signal. Maybe not for you, maybe you have different ones. But for us, we like to look at that and say, that's an early signal that they are really pushing hard on this on the RD side. Maybe they'll write it off or capitalize it. Regardless, they're saying they are really investing in building something impressive, some deep tech with some very impressive things for planet Earth. That comparison traces to an analysis dated in January 2025. The US figure is a real appropriation. The Chinese figure is an estimate carried on secondary sourcing, and it is now about 20 months old. Use it as a direction, never as a fact. If you would not accept that sourcing in a diligence memo, then do not accept that headline.
Ryan Miller
So now I want to talk about six layers and six clocks. So if the American half is the investable half, where do you actually stand? See, investing in fusion is not one decision. The stack has six layers and they pay on completely different clocks. So layer one is the reactor developers. They sell future electricity and the patents behind it. First cash is the early 2030s at best. Layer two is magnets and superconductors. They sell the tape magnet assemblies to every developer in layer one and to MRI grid buyers who have nothing to do with fusion. That is shipping product. It bills today. Layer three is the fuel cycle, tritium, enriched lithium six, breeding blankets, thin and mostly not American. Layer four is sites, land, and grid rights, permitted, powered land, and a hookup right. And here's the thing about layer four that makes it different from every other row. It gets paid by any generator or any large power user, not only Fusion. A data center wants it. A gas peaker wants it. A battery park wants it. Berkeley Labs queued up a report found in the typical American power project that came online in 2025. And it waited more than five years just to get its grid connection. And of all of that capacity that got in that line between 2000 and 2020, only 13% ever reached operation. Three-quarters of it was withdrawn. That is what scarcity looks like. A grid done right is not a fusion asset. It's an energy asset that a fusion plant happens to need. And layer five is the engineering, parts and services, vacuum vessels, gyrotrons, power electronics, engineering hours, build work. And watch who moves into this layer on purpose. First Light Fusion shut down its own reactor program in March 2025 to sell its amplifier technology instead. In its own words, that shift would significantly reduce its funding requirements while accelerating the path to revenue. All magic words that investors and courage shareholders like to hear. So a company ran this exact calculation on itself and moved into different layers. Then there's layer six, which is public proxies. As of today, there is one listed pure play, General Fusion, which listed on July 13th, 2026. And one merger that has not closed, Trump Media and Tay Technologies, who announced December 2025 and targeted for the fourth quarter of 2026. So we're coming right up on that. I cannot verify any fusion-specific funds. So read those layers again and notice something for me. Most of the money being raised is going into layer one. Layer one is also the layer with the longest wait. Layers two, four, and five, they're billing today. And layers four gets paid, whether or not fusion even works. Can you see where we're going with this?
Ryan Miller
Hey, if you're finding value out of this discussion, could you do me a huge favor? Could you just take a second and hit that like or subscribe button? It costs you nothing, but it tells the algorithm that this is valuable information and it helps us to get it to more people. Thank you. You're incredible. Now let's get back to the show.
Ryan Miller
So let me show you what this looks like in real money because two fusion companies just ran the same 18 months and got completely opposite answers. So Helion Energy in January 2025, it raised at a $5.4 billion valuation. Sixteen months later, in June 2026, it closed at $15.5 billion. So it roughly tripled. Then there's general fusion. In May 2025, it laid off at least a quarter of its staff after a $125 million raise completely fell. In August, it did a pay-to-play round of $22 million. That is the round where existing investors just put in new money or watch their stick get crushed. Then in July 2026, it went public through a shell company at about a billion dollars on paper, with roughly a hundred and fifty million actually in the bank. Same sector, same 18 months. One tripled, one went public just to stay alive. So what would you do?
Ryan Miller
Look at what separated them because it was not the physics. Helion had a binding contract with Microsoft signed in 2023 with penalties on the record if it fails to deliver. And in June 2026, it got the first licenses in the world for a fusion power plant. A contract and a license, permission and a buyer. That is not the story about who had a better reactor. Nobody knows who has the better reactor. It's about the story of which one was fundable while the science was still unfinished.
Ryan Miller
So I want to tell you about somebody that you may know, but you may not know this story about him. So in 1976, a 29-year-old bought a bankrupt hotel next to Grand Central Station in New York for about 10 million bucks. Then he handed the building to a New York state agency for $1 and took back a 99-year lease. Why give away a building that you just bought? Because the state does not pay property tax. That hotel had been carrying a bill of about $1.4 million a year. Under the new deal, he paid rent of roughly $250,000. So in January 2020, ProPublica got the final tally from New York City's tax office. Over the life of that deal, the taxes never paid came out to $410 million. And that developer was Donald Trump. So whatever you think of the man, and I'm not here to tell you what to think, that deal is a clean lesson, and almost nobody in finance has really pulled it out. We're gonna do it now. He did not make that money on the hotel, he made it on the paperwork. And here's that same method that we can assume the man in charge of the US economy at the time we're recording might follow a similar pattern.
Ryan Miller
Here's the method. One, he bought the permission, not the asset. The hotel was the delivery mechanism. The tax deal was the actual product. Two, he put the asset on someone else's balance sheet. The title went to the state agency. He held a lease. He got the economics without the ownership or risk. Three, he underwrote the approver before the market. He went to the state agency before he went to a lender. He was not asking whether New York needed a hotel. He was asking who could sign. And four, he kept the permission sellable and sold his interest in 1996. Now, hold that against fusion. The grid agreement is the permission. A state, not the NRC, is the approver. And a signed contract with a buyer who can pay is what makes this whole thing fundable. Same structure, different century. Now, I want to correct something that I nearly got wrong. People will tell you that four nuclear executive orders of May 2025 were a fusion story. The word fusion appears in none of them. But that is not the whole truth. One of them borrows a legal definition that does include fusion. And fusion got named directly six months later. So I would rather talk about what's working than to try to sell you certainty I do not have.
Ryan Miller
And here's why this matters to you. In February 2026, Pitchbook quoted a venture investor on what's actually being underwritten here. Her words, it is not a bet on revenue, it is a bet on fusion euphoria. It is not like we're underwriting somebody delivering a power plant during our fund's lifetime. See, that is a professional allocator saying the quiet part out loud. So let's talk posture. Educational framing, scenarios, not predictions. And you take all of it to your own fund council. So for venture capital, you're hunting three to five times your money at 20% a year inside a fund that has to close in 10 or 12 years. A company whose first real cash is the early 2030s cannot return a 2022 fund. So your question stops being, will fusion work? It becomes two other questions. Who leads the next round and at what mark? And what do I own if it never works? The second one is where permission changes your downside. A pure research company that fails returns about zero. And a company holding a license, a site with a grid, right, and real gear has a floor under it. And then there's private equity. You're hunting double your money at 15 to 25% a year over, I don't know, five or seven year hold. So do not buy the reactor. That clock does not fit into your fund. Buy what sits around it powered land, grid rights, engineering, service contracts, component markers. And you hold the tool that venture does not, project finance. A binding contract with a buyer who can pay supports debt that does not chase your fund. Let me put that in plain English. It means that an asset can start paying while the science is still unfinished. So your trap is the expensive one. Do not sign a completion guarantee on a first of a kind plant. And then there's deal syndicators. This is the segment where layer four sits closest to work you already do. Site control. A grid request in an agreement state and a licensing path with a precedent is a package. And it has value to a buyer that does not care which reactor ends up sitting on it. You're not underwriting fusion, you're underwriting the land that it needs.
Ryan Miller
So there's really four things to watch. And I want you to know where as well. Number one is the NRC final fusion rule. It's due at the end of 2027. So watch for that on the federal docket. Two, state frameworks. Tennessee's went live in June 2026. Watch state radiological health divisions, not Washington. Three, milestone completions in the energy department's public-private program. That is verified technical progress. It's published for free. Their staff check the work before the press went out. And then four, binding contracts, four exist. When the fifth and sixth appear, and whether they bind tells you more than any plasma could.
Ryan Miller
So what's the starter move? If you like this and you really want to start digging in and figuring this out for yourself, in the first week, pull eight companies in the energy department's milestone fusion program. Build one page, four columns, which state would each actually build in? Does it hold a license today? Does it have a binding contract with a named buyer? And what was its last known raise? And win. Every input's public. And it takes maybe an afternoon. The pro move is the institutional build. Build a standing watch list and give it an owner. Then take the step almost nobody takes. For every deal, write down the exit that does not need the tech work and name the buyer. Land, grid rights, license, gear patents. If you cannot name that buyer, you do not have a deal. You have a hope with a wire transfer attached. Spectators react, but operators, we position.
Ryan Miller
Now let me argue against myself. I'd rather be right than be loud. One, somebody delivers early and I look timid. Helion's plant is under construction right now with a 2028 stated target and a binding contract behind it. If they put a net electron on that grid in 2028, the equity holders capture repricing that no supply chain can touch. Two, the tailwind came from Congress, not the request. The administration asked for less fusion money two years running. Congress gave more both times. So if you're underwriting a policy tailwind, underwrite the appropriation, not the press release. Then three, my access argument could be overtaken. If China ever lists a fusion vehicle a foreign investor can hold or licenses its technology abroad, the wall I just described just got a door put in it. I have seen no sign of that. But absence of a sign is not proof. And I want you to hold that loosely rather than treat it as permanent. Then four, layer four could get crowded. If everybody decides powered land and grid rights are the safe way to play this, the scarcity that makes layer four work gets bid away. Scarcity rent is not a law of nature, it is a temporary condition, and other people can read a scoreboard just as well as you can. And then five, none of this works if nobody ever makes net electricity. That has not happened anywhere. Every timeline in this episode is a stated target, not an achievement. I hold my view. I'm telling you exactly where I think it breaks, in my opinion.
Ryan Miller
So here's what I want you to remember. There are two fusion races. China is winning the science, and America is winning the part that actually gets financed. And China's half is not something you can own because every share of it belongs to the state. So you watch China to price your American exposure. You never watch it to buy China. The stack has six layers, and only one of them makes you wait until the 2030s. The other five, they bill sooner, and one of them gets paid whether fusion works or not. And the three companies I showed you were not separated by physics. They were separated by whether somebody had already agreed to buy the output and whether a regulator had already said yes, green light. Own the permission, don't own the plant.
Ryan Miller
So I built you the brief. It's called the Fusion Allocation Brief, and it's in the link in the notes. The six layers with each one that sells when it pays, the access map with the ownership of China's national fusion company written out, the eight-row scoreboard with every flip trigger, the money table, so you can see what weight and cost you before you agree to even wait. And a seven question scorecard that puts any one deal into one of four results and two hard rules that only ever move you down. The bottom tier is named for the mistake this whole episode is about. Bought the plant. It's free, and it's in the link in the notes. And when you're ready to stop doing this alone, come join me inside of my community of Fund Raise Capital. It's how we find research and approach big capital sources. It's not financial advice, it's just a network of people talking about deals, and I show you how to raise capital with the proper strategies. So you get playbooks, a global network of asset managers, and weekly live sessions where you bring me the deal that you're stuck on. The application link is in the notes. So run the scoreboard once a quarter. Score your next deal against the seven questions before the deck impresses you. Check the clock before you check the science. And name the buyer who does not need the technology to work. You do these things, and you too will be well on your way in your pursuit of Making Billions.
Connect with Ryan Miller: Website | LinkedIn | Instagram