The US is preparing $4.2 billion for Vistra. The AI boom is starting to change American energy
Artificial intelligence has long ceased to be just a story about Nvidia, chips, or data centers. The larger the AI infrastructure that technology companies build, the more electricity they need. And energy, in my opinion, is becoming one of the most interesting parts of the whole AI trend.
One of the most recent examples is Vistra $VST. According to Reuters, the US government plans to provide the company with approximately $4.2 billion in financing to help increase the output of its nuclear power plants.
And the most interesting part of the whole news is that Vistra does not have to build completely new nuclear units.
More electricity from existing plants
The money is to be used primarily for so-called power uprates – that is, modernizations of existing nuclear power plants, which allow higher output from the same facility.
This can include turbine upgrades, equipment modifications, or the use of more efficient fuel.
According to Reuters, the increase in capacity should affect at least three of the four nuclear power plants operated by Vistra.
In total, the company has:
6 nuclear reactors, 4 nuclear power plants, and more than 6.5 GW of nuclear capacity.
That corresponds to electricity for approximately 3.25 million households.
And the possibility of increasing production from existing assets is, in my opinion, very interesting. Building a completely new nuclear unit is extremely expensive and can take many years. Increasing the output of an already operating plant can be a significantly faster path.
Why does the US need so much new electricity?
American electricity consumption is entering a growth phase after a long period of relative stagnation.
One of the main reasons is AI data centers.
Training and operating large AI models require enormous amounts of computing power, and data centers need stable energy supplies practically 24 hours a day, 7 days a week.
On top of that, there is further electrification of the economy, electric vehicles, and, for example, crypto mining.
This creates a very interesting situation: technology companies may have enough chips and capital to build data centers, but without sufficient electricity, their entire infrastructure will be of little use.
And that is where companies like Vistra start to become interesting.
Nuclear is coming back into play
The US government is also significantly changing its approach to nuclear energy.
The Department of Energy this year launched the UPRISE program, whose goal is to increase the output of existing reactors, extend their lifespan, and return some shut-down nuclear sources back into operation.
The goal is to add approximately 2.5 GW of nuclear capacity by 2027 and 5 GW by 2029.
That is, in my opinion, a fairly fundamental change. Just a few years ago, many investors viewed traditional energy companies as a relatively boring sector. But AI can change the whole dynamic.
Suddenly, high-quality energy infrastructure becomes one of the conditions for further development of the technology sector.
Why am I interested in Vistra?
I have already bought Vistra for my portfolio in the past, and members of my community know about these purchases.
But my thesis is not based solely on one news story about $4.2 billion in financing.
I especially like the combination of several factors:
stable electricity production, exposure to nuclear energy, growing demand for electricity, and potential long-term growth in consumption driven by data centers.
If AI infrastructure develops at the pace currently planned by companies like Microsoft, Meta, Amazon, or Google, someone will have to supply all these data centers with electricity.
And that is why, in my opinion, the winners of the AI boom may not be only chip manufacturers.
They may also be companies that own power plants, transmission infrastructure, and other energy assets.
...
The $4.2 billion financing alone does not mean that Vistra automatically represents a cheap or risk-free investment. Projects must be realized, investments can become more expensive, and after strong stock growth, the valuation also needs to be monitored.
But something else is more important to me.
The AI boom is gradually revealing another bottleneck in the entire infrastructure – electricity.
And if technology companies really build data centers on the scale they currently plan in the coming years, energy could be one of the most interesting secondary ways to bet on AI growth.
I have been following Vistra for some time and I also hold it in my portfolio - I share all my purchases transparently here on my Bulios membership!
If you are interested in my specific purchases, company valuations, and more detailed investment theses, you can join the Ordinary Investors community, where I publicly share all my purchases. Membership costs about one lunch a month and it will be worth it!