American Electric Power

$Amazon(AMZN.US) is funding a nearly 1 gigawatt nuclear power plant in Indiana to help feed its data centers

$Amazon(AMZN.US), $American Electric Power(AEP.US) and X-energy $X-Energy(XE.US) announced a joint development deal Monday to study building 12 of X-energy's Xe-100 small modular reactors at the site of a retired coal plant in Sullivan County, Indiana

Each reactor makes 80 MW, so the full plant would add 960 MW of carbon-free, around the clock power for the regional grid and $Amazon(AMZN.US)'s data centers

$Amazon(AMZN.US) is putting up milestone-based development capital and intends to buy the power under a long-term contract that covers the plant's costs, so $American Electric Power(AEP.US)'s existing customers are not stuck with the bill

It is the second project in the $Amazon(AMZN.US) and X-energy push to add more than 5 GW of new US nuclear capacity by 2039, following the Cascade facility with Energy Northwest in Washington state. $Amazon(AMZN.US) led X-energy's $700M funding round in 2024

The catch is that this is still feasibility work, with no final investment decision. No Xe-100 is running yet, and X-energy's first plant, a four reactor project at a $Dow(DOW.US) site in Texas, is still in the final stage of NRC review for a construction permit

Last week the market punished AI stocks over how the buildout gets financed. This is the other half of that bill. Power is turning into the real bottleneck, and the hyperscalers are now underwriting their own power plants more than a decade out. $Amazon(AMZN.US) is already backing over 40 GW of carbon-free energy across 700+ projects, but the payoff here runs on nuclear timelines, not earnings calls

Source: StockMarket.News

New: Thiel Macro (Peter Thiel's fund) just dropped their Q2 2026 13F

Thiel showed zero holdings for two straight quarters and now Thiel is back with a reported ~$419M portfolio

Top holdings:

• Amazon $Amazon(AMZN.US): 28.2%

• Vista Energy $Vista Energy(VIST.US): 18.1%

• Vistra $Vistra(VST.US): 14.1%

• American Electric Power $American Electric Power(AEP.US): 10.1%

• DTE Energy $DTE Energy(DTE.US): 9.6%

• FirstEnergy $Firstenergy(FE.US): 9.5%

• CMS Energy $CMS Energy(CMS.US): 9.4%

• X-Energy $X-Energy(XE.US): 0.9%

For context: Thiel co-founded PayPal and Palantir, was the first outside investor in Facebook, and runs Founders Fund, which was the first institutional money into SpaceX

Some thoughts on 13F season disclosures that we got…

First, the top 10 names that were bought in Q2 across the largest funds:

$Microsoft(MSFT.US), $Meta Platforms(META.US), $Visa(V.US), $Amazon(AMZN.US), $BRK.B, $S&P Global(SPGI.US), $Alphabet - C(GOOG.US), $Disney(DIS.US), $Capital One Financial(COF.US), $Thermo Fisher Scentific(TMO.US)

It is obvious that the blue chip companies and Mag 7 names continue to be preferred by larger investors over the semiconductor names. Four of the top 10 stocks bought were Mag 7s. While the semiconductor names are getting a bid, they are getting it from the David Teppers, Brad Gerstners, and Chase Colemans of the world. The larger funds that aren’t as aggressive are just buying discounts on some of the highest quality businesses like Visa, Amazon, Meta, and Microsoft.

$Uber Tech(UBER.US) got some big buys. Ackman added to his position by 14%, Tepper added to his by 21%, and Terry Smith bought 8M shares.

$Alphabet - C(GOOG.US)L was the most heavily sold Mag 7. While it was in the top 10 most bought stocks, that was because of concentration vs number of funds. 26 larger funds reduced their position while only 4 added to it, but the ones who added did so heavily. Berkshire increased their stake by 45%.

$SpaceX(SPCX.US) had some heavy buyers but it seems like many of these funds had existing xAI positions, like Nvidia and AMD, and they didn’t sell in Q2 but they didn’t freshly buy on the open market.

$Micron Tech(MU.US) and $Sandisk(SNDK.US) were not super present across disclosures. Tepper sold 40% of his position. Brad Gerstner added 209K shares.

$Taiwan Semiconductor(TSM.US), $Amazon(AMZN.US), and $NVIDIA(NVDA.US) continued to be very strong large caps that either got a decent amount of buying or holds across portfolios with few sells.

Peter Thiel disclosed a $418M portfolio with 72% of the names allocated to energy which might end up being the next bottleneck that gets significant buyers. His positions: $Amazon(AMZN.US), $Visa(V.US)IST, $Visa(V.US)ST, $American Electric Power(AEP.US), $DTE Energy(DTE.US), $Firstenergy(FE.US), $CMS Energy(CMS.US), and $X-Energy(XE.US).

Overall, it seems like the bias has been around buying large caps with sustainable earnings growth over smaller growth stocks. Financials are not really showing up either, neither is healthcare, but the market is broadening to them. Tech continues to dominate. The S&P is at all time highs so obviously there is significant buying over selling but the question now is, at what point do many of these funds allocate outside of tech or do they continue to double down on the discount that blue chips and Mag 7s are seeing vs the semi names that had a hard July but still trade at much higher multiples vs the rest of the market.

To be honest, $Uber Tech(UBER.US) is the most compelling to me from these disclosures. It has been stuck for literally 2 years, the FCF is growing massively, and the big guys are buying in heavily.

Are there any names disclosed that you are starting to get interested in?

Source: amit

⚡ 𝐔𝐏𝐃𝐀𝐓𝐄: $Bloom Energy(BE.US) Bloom Energy Expands Partnership With MiTAC to Power AI Server Manufacturing

👉 𝗞𝗲𝘆 𝗛𝗶𝗴𝗵𝗹𝗶𝗴𝗵𝘁𝘀:

➤ 𝗕𝗹𝗼𝗼𝗺 𝗘𝗻𝗲𝗿𝗴𝘆 will deploy a fuel cell microgrid at 𝗠𝗶𝗧𝗔𝗖'𝘀 Fremont campus.

➤ Microgrid will power MiTAC's 𝗔𝗜 𝘀𝗲𝗿𝘃𝗲𝗿 manufacturing operations.

➤ Project expands Bloom's contracted onsite power across 𝘁𝘄𝗼 California sites.

➤ Bloom now serves nearly 𝘁𝘄𝗼 𝗱𝗼𝘇𝗲𝗻 AI infrastructure customers.

➤ Bloom's AI infrastructure portfolio totals approximately 𝟮𝟱𝟬 𝗠𝗪 of capacity.

➤ Fuel cell systems will operate as an 𝗶𝘀𝗹𝗮𝗻𝗱𝗲𝗱 𝗺𝗶𝗰𝗿𝗼𝗴𝗿𝗶𝗱.

➤ Project supports expanding AI server production as demand accelerates.

➤ Bloom says fuel cells provide fast-deploying, reliable onsite power.

➤ Bloom has also deployed fuel cells for customers including 𝗔𝗘𝗣, 𝗕𝗿𝗼𝗼𝗸𝗳𝗶𝗲𝗹𝗱, 𝗘𝗾𝘂𝗶𝗻𝗶𝘅, 𝗡𝗲𝗯𝗶𝘂𝘀, and 𝗢𝗿𝗮𝗰𝗹𝗲.

👉 𝗪𝗵𝘆 𝗧𝗵𝗶𝘀 𝗠𝗮𝘁𝘁𝗲𝗿𝘀:

➤ Highlights rising power demand across the AI infrastructure supply chain.

➤ Reinforces Bloom's growing position in onsite power for AI infrastructure.

➤ Demonstrates manufacturers are adopting onsite energy to avoid grid constraints.

➤ Expands Bloom's exposure to the fast-growing AI infrastructure market.

👉 𝗘𝘅𝗽𝗲𝗿𝘁 𝗦𝘁𝗮𝘁𝗲𝗺𝗲𝗻𝘁:

𝗔𝗺𝗮𝗻 𝗝𝗼𝘀𝗵𝗶, Chief Commercial Officer at Bloom Energy:

“As the buildout of AI infrastructure accelerates, demand for power is growing at every layer of the ecosystem. We're seeing that demand translate into rapid customer adoption. Today, Bloom serves nearly two dozen AI infrastructure customers with approximately 250 MW of capacity, from a running start of nearly zero in this segment just two years ago. Projects like MiTAC demonstrate that Bloom is the standard for onsite power, helping customers – from data centers to advanced manufacturers – move faster and build the infrastructure powering the AI economy.”

Data center developers figured out that utilities will build whatever infrastructure you ask for so they're submitting the same projects to multiple utilities, bid shopping for the best rates and fastest connections. AEP Ohio fielded requests for 190 gigawatts when their entire system is only 37 gigawatts

The real issue is that utilities actually build this stuff. Billions in substations transmission upgrades, the whole infrastructure stack gets deployed based on inflated promises. When those projects never materialize or get scaled down, the costs don't disappear. They get passed to ratepayers through higher electricity bills.

The bigger picture is messed up though. Tech companies are increasingly building their own on site power with fuel cells and generators anyway so ratepayers are funding grid infrastructure to compete against backup power systems the hyperscalers were going to use anyway.

So basically, we're all subsidizing Meta and Googles infrastructure bets while they hedge by building their own power anyway. And when it doesn't work out we're stuck holding the bag with stranded assets. How is that not a regulatory failure?

Source: StockMarket.News

The government may be shut down, but politicians keep trading.

Rick Larsen just disclosed a buy of American Electric Power he made Oct 6.

AEP is a major U.S. utility operating power transmission and distribution across several states.

Larsen sits on the House Transportation & Infrastructure Committee - which oversees laws and funding for energy infrastructure.

We’ll keep watching this one.