Nuclear: cost is a policy problem
The last big gigawatt reactor built in the United States finished a year and a half to two years ago and cost about $29 billion once delays and custom design were added in. China builds the same size reactor for $6 billion each. The reason is simple: China builds ten of the same kind at once, and they do not have to redesign each one to fit local politics.
If the Nuclear Regulatory Commission approved a single American design that could be permitted on a cookie-cutter basis and assembled the same way each time, the front-end capital cost of nuclear power here would drop from $28 billion to $6 billion. American technology is as good as China's. The difference is that China uses it constantly, so its builders keep getting better.
Getting out of the way is the first fix. The second costs the government nothing: change the tax code. Build a nuclear plant in Denmark or China and you can write off the concrete shell for taxes in five years, and against other income too. In the United States the same shell is written off in a straight line over 30 years. The Nordic countries that the political class admires actually run more investment-friendly tax codes than America does. Make the US code as investor-friendly as China's, Sweden's, or Denmark's, and most of the disadvantage America faces disappears.
The full fix list: drop regulatory roadblocks, be clear about what's expected in resource development, stop the export of harmful materials, standardize safety designs across reactors and refineries, and keep the cost of mistakes private rather than passing it to taxpayers. One more that industry would hate: require environmental bonding up front for all resource and extraction projects. Do these, and we would not even be having this conversation.
Financing decides who wins
Cheap money wins commodity cycles before production numbers ever hit the news. China's government used Chinese banks to funnel below-market money to Chinese companies, and that let them grow world-scale businesses in 30 years.
The math is stark. A company below investment grade that wants to build a gold mine pays 13% to 15% a year on its construction loan. An investment-grade company borrowing on its own balance sheet pays around 6.75%. A Chinese firm borrowing through the Industrial and Commercial Bank of China pays 3.5%. So an independent miner like Aris Mining carries a 13% to 15% cost of capital, while Zijin's is 3.5%. A company borrowing at 3.5% plays by completely different economics than one paying four times that. Geology matters less than the loan.
Government money: horrifying as a taxpayer, delightful as a speculator
Government capital gets misapplied every time. It is not steered by risk-adjusted net present value or the odds of return on capital. It is steered by whatever wins the most votes for the people who hand it out. That makes it a non-economic transaction.
Mountain Pass in California got a huge government grant, and that deposit has gone bankrupt three times. An antimony mine in Idaho is now getting grant money thrown at it aggressively, with spending deadlines that force the operator to scramble to qualify. He says it makes his life hell, and he'll take it anyway. Seven years ago, under the Biden administration, people who pushed for US uranium were treated like criminals. Now the same officials want to subsidize them. Being vilified felt cleaner than being subsidized. But the mining industry loves no money more than dumb money, and it is cascading in now. The smart money that ran ahead of it is enjoying wonderful times.
That points to the real edge for an investor: figure out where the dumb money is going and get in front of it. Watch where subsidies land and where grant money flows urgently, not where politicians make speeches. Capital allocation shows priorities before the public story catches up.
There is a hard line. Cutting regulatory friction strengthens competition. The government taking equity stakes in major companies is state control, the seed of socialism and eventually communism. That view drew a standing ovation at the conference. The open question: where does an enabling loan or grant end and state control begin? The answer here: it is not the government's job to pick winners and losers. Remove the obstacles instead. Bring capital in from private sources and let the free market decide where it goes. I would rather allocate my own capital than have Donald Trump or a senator, who knows far less about resources than I do, allocate it for me.
The urgency behind all this is telling. When government moves fast, it stops dotting its i's and crossing its t's. This comes after 13 years of pointless permitting on the very projects now being subsidized. The permitting system had to become practical instead of political, focused on not doing harm rather than pleasing certain elites, and that shift has started. A long-time contact at the Bureau of Land Management put it plainly: if you want to get something done, you've got two years, before the next administration change.
Oil: today's shortage is fake, the next one is real
The oil shortage right now is artificial and temporary. It comes from war. Today's higher oil prices reflect the threat of a shortage, not an actual one. If clearer heads de-escalate within about a ten-day window, we dodge the consequences.
China, Japan, and the United States hold real strategic reserves. Places that could never afford reserves, like Sri Lanka and Pakistan, are already behind. If a genuine shortage hits and oil gets rationed by price, that price will be far higher than now, though the exact number is unknown. North America gets hurt less because it produces plenty of oil. Someone in North America looks at the pump price, curses, fills up, and drives off. A taxi driver in Colombo, Sri Lanka looks at that same price and parks his cab. If US reserves fell far enough, the country could stop exporting, and it would; sending crude abroad while Americans can't drive would carry a heavy political cost, and Congress would agree to halt it.
The bigger danger is structural. A structural shortage arrives in late 2029 or early 2030. It comes from 30 years of underinvestment, and especially from three years of the industry cutting about $1 billion a year in the sustaining investment needed just to hold output. That damage does not show up in year one, two, or three. It shows up in the later years. No ceasefire or armistice fixes it. Only massive capital spending over a long stretch does, and the odds of correcting it in a reasonable timeframe are much lower than for the artificial kind. The price jump in oil during 2026 is a preview of what's coming.
