🏆 BASTION DEEP VALUE AWARDS
🇺🇸 USA · Top 25 by market cap
23 of the 25 biggest US companies score below 50 on value. Two are actually cheap. Handing out awards across 6 categories 👇
1️⃣ Forward P/E (cheapest on next-year earnings)
Winners:
🥇 💾 Micron 11x $MU
🥈 🛢️ Exxon 12x $XOM
🥉 🏦 JPMorgan 14x $JPM
Outsiders:
🚗 Tesla 202x $TSLA
🔷 Intel 101x $INTC
🖥️ AMD 58x $AMD
2️⃣ P/E 2028 (cheapest on earnings 3 years out)
Winners:
🥇 🏦 JPMorgan 12x $JPM
🥈 💾 Micron 13x $MU
🥉 📱 Meta 14x $META
Outsiders:
🚗 Tesla 116x $TSLA
🔷 Intel 45x $INTC
🏬 Costco 38x $COST
3️⃣ EV/EBITDA (cheapest on enterprise value)
Winners:
🥇 🎩 Berkshire 6x $BRK.A 🥈
🛢️ Exxon 7x $XOM 🥉
📱 Meta 10x $META
Outsiders:
🚗 Tesla 99x $TSLA
🖥️ AMD 64x $AMD
⚙️ Lam Research 46x $LRCX
4️⃣ Free Cash Flow Yield (most cash per share)
Winners:
🥇 💊 J&J 3.3% $JNJ
🥈 📱 Meta 3.2% $META
🥉 🛢️ Exxon 3.0% $XOM
Outsiders:
🗄️ Oracle −3.5% $ORCL
🔷 Intel −0.6% $INTC
📦 Amazon −0.1% $AMZN
5️⃣ Dividend Yield (most generous to shareholders)
Winners:
🥇 🛢️ Exxon 2.8% $XOM
🥈 💊 J&J 2.4% $JNJ
🥉 🏦 JPMorgan 2.0% $JPM
🏅 TOP AWARD: Bastion Deep Value Score
Winners (most undervalued):
🥇 🏦 JPMorgan 68 $JPM
🥈 🛢️ Exxon 65 $XOM
🥉 📱 Meta 48 $META
Outsiders (most expensive):
🔷 Intel 10 $INTC
🚗 Tesla 11 $TSLA
🖥️ AMD 13 $AMD
How the Bastion Deep Value Score works
Score from 0 to 100. We rank each stock against the entire global market across every value metric: Forward P/E, P/E 2028, EV/EBITDA, EV/EBIT, FCF Yield, Dividend Yield, P/BV. Then we combine these into a single composite rating and convert it into a percentile. A score of 100 means the stock lands in the top 1% globally on value, 90 means top 10%, and so on down the scale.
Every hedge fund I respect is suddenly talking about the same thing, and... it is not the chips.
It is the one bottleneck that breaks the entire AI story if it is not solved. Around 20 public companies sit on it. I put them all in one map across 5 layers.
Let's dive into it 🧵
Here is the thing nobody priced in two years ago. We spent a decade with flat electricity demand in this country. Utilities planned around it. Then AI showed up asking for gigawatts at a time.
The Electric Power Research Institute now thinks data centers could eat 9% to 17% of all US electricity by 2030, up from roughly 4% in 2023. Former Google CEO Eric Schmidt told Congress the sector may need 67 more gigawatts by the end of the decade. That is not a tweak to the demand curve. That is a new industrial revolution landing on a grid built for a different century. Every company below sits somewhere between a power plant and a server rack. This is the map.
🔌 POWER GENERATION & UTILITIES
Start at the source. These are the companies that actually make the electrons. For years this was the most boring corner of the market: regulated returns, slow growth, dividend investors only. Then the hyperscalers started signing power contracts directly with generators, and the whole category repriced.
$VST Vistra
This is the one I watch most closely in the group. Vistra signed Meta to a power purchase agreement for roughly 2,600 megawatts at its PJM nuclear sites, which tells you everything about where this is going: tech giants are now buying nuclear output directly. Q1 2026 adjusted EBITDA hit a record for a first quarter at $1.494 billion. They have hedged almost all of their 2026 generation, and they have bought back about 30% of the company since late 2021. A generator that trades like a buyback machine with an AI tailwind bolted on.
$CEG Constellation Energy
The largest nuclear fleet in the country, and the company that put nuclear back on the front page when it agreed to restart Three Mile Island for Microsoft. In January it closed the $21.8 billion Calpine acquisition, adding around 23 gigawatts of mostly gas and renewable capacity, and Q1 2026 revenue more than doubled the year before to $11.1 billion. The thesis is simple: when an AI company wants carbon free baseload power tomorrow, there are very few phone numbers to call, and this is one of them.
$GEV GE Vernova
If you only own one name in this entire map, my honest take is that it should probably be this one. GE Vernova makes the gas turbines and the grid equipment, the literal picks and shovels of the buildout. In a single quarter its Electrification segment booked $2.4 billion in data center equipment orders, more than it booked in all of 2025. Total backlog sits around $163 billion and management pulled forward its $200 billion target to 2027. The gas turbine backlog jumped from 83 to 100 gigawatts in one quarter, and they are raising prices into that demand. This is the cleanest expression of the trade.
$BEPC Brookfield Renewable
Note the ticker: this is Brookfield Renewable, $BEPC, not the $BE on most charts (that is Bloom Energy). Brookfield operates about 47 gigawatts and is developing a pipeline north of 200. It signed a framework with Microsoft to deliver over 10 gigawatts, roughly eight times the size of the largest single corporate power deal ever signed before it, plus a multi gigawatt hydro deal with Google. It also owns about half of Westinghouse alongside Cameco. The patient, contracted, dividend paying way to play the same wave.
⚛️ SMALL MODULAR REACTORS
Now the speculative end. The promise here is clean, firm baseload power in a compact box you can site right beside a data center. The catch: almost none of these are producing commercial power at scale yet, so you are buying a timeline as much as a company. Price that carefully.
$OKLO Oklo
The most exciting and the most expensive name in the room. In May the NRC approved the principal design criteria for Oklo's Aurora powerhouse in under half the usual review time, a real regulatory step forward. The customer pipeline is around 14 gigawatts, anchored by a 12 gigawatt agreement with Switch and a 500 megawatt deal with Equinix, and it added a research partnership with NVIDIA and Los Alamos. Just remember Oklo plans to build, own and operate its reactors and has essentially no revenue yet. This is a call option on a 2028 plus story.
$SMR NuScale Power
The one with the regulatory lead. NuScale has NRC design approval for both its 50 and 77 megawatt modules, which genuinely derisks deployment. It is sitting on about $1.2 billion in liquidity and is working toward a definitive power agreement with TVA through its ENTRA1 partner, with its first project tied to RoPower in Romania. Revenue was a rounding error last quarter because the licensing work wrapped up, so this is still a story about getting the first units in the ground.
$BWXT BWX Technologies
The adult in the room, and the name I would own if I wanted nuclear exposure without buying a lottery ticket. BWXT actually makes money: Q1 2026 revenue of $860 million and net income of $91 million, and it raised full year guidance. It builds reactors for the US Navy, produces medical isotopes, and just acquired Precision Components Group to push into commercial nuclear manufacturing. While the SMR startups sell the future, this one sells into it today.
$XE X-energy
Brand new to the public market. X-energy IPO'd on April 24 at $23 a share, raised about $1.02 billion, and came out around a $12 billion valuation with Amazon as its anchor backer holding nearly a third of the company before the listing. It pairs an 80 megawatt reactor design with its own proprietary TRISO fuel, and its order book already tops 11 gigawatts including Amazon's commitment to as much as 5 gigawatts by 2039, plus Dow and Centrica. Reality check: it lost about $390 million on $109 million of revenue in 2025, and first deployments are not expected until the early 2030s.
⛏️ CRITICAL MINERALS
You can build every reactor on the list above and they are paperweights without fuel. This is the front end of the cycle: mining, enrichment, conversion, and the magnet metals the whole grid runs on. Quick note: I swapped the misfiled Northland slot for Energy Fuels here, which is a genuine US critical minerals producer.
$CCJ Cameco
The blue chip of the uranium world. Q1 2026 net earnings jumped 87% and adjusted EBITDA rose 44% to $509 million on stronger prices and volumes. The kicker is Westinghouse: Cameco owns roughly half of it alongside Brookfield, so it captures both the fuel and the reactor technology side of the renaissance. When people want uranium exposure without a science project, they buy this.
$LEU Centrus Energy
The reshoring play, and a fascinating one. Centrus is the only production ready uranium enricher in America, sitting on a $2.3 billion enrichment backlog, a $900 million HALEU award from the Department of Energy, and a notice from the NNSA that it intends to sole source enrichment work to them. It is pouring over $560 million into its Oak Ridge centrifuge factory and is even exploring a fuel joint venture with Oklo. This is a national security story wearing a stock ticker.
$UUUU Energy Fuels
This is what $UUUU actually is. Energy Fuels runs White Mesa, the only conventional uranium mill operating in the United States, and it is the rare company licensed to produce both uranium and separated rare earth oxides under one roof. Its 2026 uranium guidance implies growth of 50% to 150%, and it is now turning out the dysprosium, terbium and magnet metals that everything from EV motors to grid hardware depends on. Uranium and rare earths, the two supply chains Washington is most desperate to pull back from China, in one company.
$NLR VanEck Uranium and Nuclear ETF
If you would rather own the whole theme in one line instead of picking a winner, this is the basket. $NLR holds the nuclear value chain end to end: reactors, enrichers, miners and the utilities running the plants. A lot of this very map sits inside it, with Constellation, Cameco, Centrus, BWXT and Energy Fuels all among its largest positions. The lazy way to be right about the sector even if you pick the wrong individual stock.
🔧 POWER INFRA & GRID
Between the power plant and the server rack is the least glamorous and maybe most investable layer of all. Transformers, switchgear, cooling, and the crews who build it. The dirty secret of the AI buildout is that the grid itself is the bottleneck. Interconnection queues run years, and the equipment to connect anything is on backorder.
$VRT Vertiv
The purest grid adjacent winner so far. Q1 2026 sales rose 30% to $2.65 billion, with the Americas up 44% on data center demand, earnings per share up triple digits, and guidance raised twice in two quarters. Vertiv makes the power and thermal systems that keep a data center alive, and it just joined the S&P 500. When the chip names sneeze, this one catches it, but the order book keeps validating the story.
$HUBB Hubbell
Boring on purpose, and that is the point. Hubbell makes the electrical and utility hardware, the transformers, metering and grid components, that every new data center and every grid upgrade quietly requires. It will never 10x in a year, but it sells into both the AI buildout and the broader grid replacement cycle at the same time. This is the ballast in the basket.
$POWL Powell Industries
My favorite quiet story in this section. Powell makes custom electrical equipment for utilities, energy and now data centers, and the demand signal is screaming: orders up 97% last quarter, a record $1.8 billion backlog, and right after the quarter closed it landed a single data center order worth more than $400 million, the largest in its history. It did a three for one split this spring and carries no debt. A small cap industrial running into a structural tailwind.
$PWR Quanta Services
The labor. Quanta physically builds and upgrades the grid, the part of this problem that no software fixes. Q1 2026 revenue rose 26% to $7.87 billion and its backlog hit a record $48.5 billion. If all of the generation and transmission above actually gets built, a meaningful slice of it gets built by crews like these. The pick and shovel play on the wires themselves.
🖥️ DATA CENTER POWER
The wild card, and the highest beta corner of the map. These started as bitcoin miners, which means they already owned the one thing everyone now wants: large blocks of interconnected power and the land around it. They pivoted to hosting AI compute, signing leases with the hyperscalers and the neoclouds. Enormous growth, real execution, and serious single customer risk. Size accordingly.
$IREN IREN
The furthest along. Formerly Iris Energy, IREN has a Microsoft AI cloud partnership worth billions, a power pipeline around 4.5 gigawatts, and high performance computing on track to make up the majority of its revenue by the end of the year. It already trades like an infrastructure company rather than a miner, because increasingly that is what it is.
$WULF TeraWulf
TeraWulf describes itself as a power company that happens to build digital infrastructure, which I think is exactly the right framing for this whole row. It has locked in over $12.8 billion of contracted compute revenue through long term leases with the Google backed Fluidstack and Core42, anchored by its Lake Mariner site and scaling toward a gigawatt of power. Its leasing revenue more than doubled year over year. Controlled power, leased to AI, on a multiyear contract.
$CORZ Core Scientific
The contrarian one. CoreWeave tried to buy Core Scientific in an all stock deal, and in a rare moment of shareholder backbone, the holders voted it down in late 2025. So it stays public, and it kept the prize: roughly $10 billion or more of contracted revenue with CoreWeave across about 590 megawatts, while converting its old mining sites into AI colocation. You are betting the company creates more value alone than the buyout offered.
$CIFR Cipher Mining
The earliest stage of the pivot, rebranding toward AI as it goes. Cipher signed a hosting deal backed by Google's Fluidstack, with Google taking around a 5% stake, plus a 300 megawatt arrangement tied to AWS, building toward a contracted compute backlog around $9 billion. Highest risk, least proven, most torque if the leases convert to cash on schedule.
⚡️FINAL THOUGHTS
Step back from the tickers and a pattern jumps out. The market is paying up for the same insight at five different points on the same wire.
The stability lives at the bottom and the middle. Cameco, Hubbell, Quanta Services and BWX Technologies make money today and sell into a buildout that is contracted for years. They will not triple overnight, but they do not need a single thing to go right that has not already happened.
The growth lives at the edges. GE Vernova is the rare name that has both, scale and acceleration, which is why I keep coming back to it. The reactor startups and the former miners are where the imagination is, and also where the disappointment will be when timelines slip, because timelines always slip in nuclear and in construction.
The clearest read of all is that the AI story quietly handed the baton from the chip layer to the power layer, and most people are still watching the wrong race. You cannot run the model without the electrons, and the electrons are the scarce thing now.
I will say the obvious part out loud: this is a map, not advice. I am pointing at where the money is moving, not telling you what to buy. Do your own work on every one of these, especially the speculative names where a single contract or a single regulator can move the whole thesis.
If this saved you a week of research, do me a favor and bookmark it, then send it to the person in your group chat who only owns Nvidia. The power bottleneck is the second half of that trade.
@SeriousVerySam Meine Frau kommt aus dem Pott.
Der Kunde in der Apotheke möchte no a Päckle Bäbber und a Guck derzua.
der Bäp = Kleber, etwas ist bäbbig = klebrig (lang mi mit doine bäbbige Hend et ah)
der Bäbber = Kleber, Aufkleber
die Bäbber = hier Pflaster, bäbbad ja au
Goldman Sachs on memory:
- DRAM to remain in undersupply until atleast 2028 (undersupply of 5.0%, 5.9% and 3.9% in the years 2026, 2027 and 2028 respectively).
- DRAM global demand revised upwards, now expected to grow by 28%, 20% and 19% y/y in 2026, 2027 and 2028 respectively.
- NAND to also remain in undersupply until atleast 2028 (undersupply of 4.4%, 4.6% and 3.0% in 2026, 2027 and 2028 respectively).
- NAND global demand revised upwards, now expected to grow by 20%, 23% and 19% y/y in 2026, 2027 and 2028 respectively.
- HBM to remain in undersupply until atleast 2028 (undersupply of 5.4%, 6.0% and 4.3% in 2026, 2027 and 2028 respectively).
- HBM TAM expected to be $56 billion, $116 billion and $168 billion in 2026, 2027 and 2028 respectively.
Slowly, the overall consensus is shifting towards memory remaining in shortage until atleast 2028, while it was 2027 previously.
$MU $SNDK $DRAM $EWY
Nvidia $NVDA CEO Jensen Huang just said we are at the beginning of a new market
CPUs FOR AGENTS
Here's who is already signed up to use Nvidia's new Vera CPU:
EARLY ADOPTERS:
- OpenAI
- Anthropic
- SpaceX $SPCX
CLOUD PARTNERS:
- Nebius $NBIS
- Oracle Cloud $ORCL
- CoreWeave $CRWV
- Nscale
- Crusoe
- Firmus
- Lambda
- Together AI
ECOSYSTEM PARTNERS:
- Dell Technologies $DELL
- HPE $HPE
- Lenovo
- Super Micro Computer $SMCI
- Aivres
- ASRock Rack
- Asus
- Compal
- Foxconn
- Gigabyte
- Hyve Solutions
- Inventec
- Mitac Computing
- MSI
- Pegatron
- QCT
- Wistron
- Wiwynn
This is an excellent interview btw
Nicolai (Norwegian Sovereign Wealth Fund CEO) asks the IBM CEO if AI a bubble
Listen very very carefully to his answer
Top 50 US Tech Stocks: Forward P/E vs. Expected Earnings Growth
The regression line shows what P/E each growth rate "should" earn.
Most Undervalued
💾 SanDisk $SNDK
🗄️ Western Digital $WDC
📡 Broadcom $AVGO
🎮 NVIDIA $NVDA
🔌 Marvell $MRVL
Most Overvalued
🖥️ Intel $INTC
❄️ Snowflake $SNOW
🛡️ CrowdStrike $CRWD
☁️ Cloudflare $NET
🚗 Tesla $TSLA
Robert Greene: Learn by Doing.
“The brain is designed to learn through constant repetition and active, hands-on involvement. Through such practice and persistence, any skill can be mastered.”
A British biologist looked at 200,000 years of human history and found that the entire reason humans broke out of poverty was not intelligence, not language, not even agriculture, but one mechanism so simple a 6-year-old could explain it.
His name is Matt Ridley.
He is a zoologist by training, an evolutionary biologist by career, and in 2010 he wrote a book called The Rational Optimist that quietly argued the most important fact about human progress had been hiding in plain sight for the entire history of economics.
Naval Ravikant has been telling people to read everything Ridley has ever written for the last 15 years. The reason is the argument inside this one book.
For 200,000 years, anatomically modern humans walked around with the same brain you have right now. Same skull size. Same neural architecture. Same raw capacity for language, planning, and abstract thought.
For roughly 190,000 of those years, almost nothing happened. Generation after generation lived and died inside the same Stone Age toolkit their great-great-grandparents had used. Then somewhere around 50,000 years ago, the line on the chart of human progress started to tick upward. Then it bent. Then it exploded.
The question Ridley spent years on was the only question that mattered. What changed.
It was not the brain. The brain had been the same for 190,000 years. It was not language, which had existed long before the takeoff. It was not even agriculture, which arrived only 10,000 years ago and was actually preceded by the upward bend, not the cause of it.
What changed was that humans started trading with strangers.
This sounds too small to be the answer. Ridley argues that it is the answer to almost everything. The moment one human exchanged a useful object with another human from a different group, something happened that no other species on earth had ever done.
Two ideas that had developed in isolation came into contact. The flint knapper learned what the spear maker had figured out. The fisherman from the coast learned what the hunter from the forest had figured out. The two pieces of knowledge fused into something neither side could have produced alone.
Ridley calls this ideas having sex. The phrase sounds frivolous and it is meant to. The point is that ideas, like genes, get better when they combine with other ideas from different lineages.
An idea sitting inside one head, no matter how brilliant the head, eventually hits a ceiling. The same idea exposed to ten thousand other ideas does something genes do under sexual reproduction. It mixes. It recombines. It produces offspring nobody planned.
The cleanest proof of this argument is the most uncomfortable case study in the book. Tasmania.
Around 10,000 years ago, rising sea levels cut Tasmania off from mainland Australia. A population of roughly 4,000 humans was now isolated on an island, with no possibility of contact with the rest of humanity. They had the same brains. The same language. The same starting toolkit as their cousins 150 kilometers north. The natural experiment was now running.
What happened next is something no economist or geneticist had ever predicted.
The mainland Australians kept inventing. Boomerangs. Spear-throwers. Fishing nets. Bone needles for sewing fitted clothes. Watercraft with paddles. Their technology compounded slowly across the centuries.
The Tasmanians went the other way. They did not just fail to invent the new tools their cousins were developing. They started losing the tools they already had. Fishing was abandoned within a few thousand years. Bone tools disappeared. Fitted clothing disappeared. They forgot how to make fire from scratch and started carrying lit firebrands from camp to camp instead, relighting their fires from a neighbor's whenever their own went out.
By the time European explorers arrived in the 17th century, the Tasmanians had the simplest toolkit of any human society ever recorded. Their material culture had gone backward for 8,000 years.
The archaeologist Rhys Jones called it a slow strangulation of the mind.
Joseph Henrich at Harvard later proved with formal mathematical models that there was nothing wrong with Tasmanian brains. There was something wrong with their network. A toolkit requires a critical mass of people exchanging skills to maintain itself.
The act of teaching a skill is imperfect. Every generation loses a small percentage of what the last generation knew. If your population is large enough and trading widely enough, those losses get caught and corrected by someone else who still remembers.
If your population shrinks below a certain threshold and stops mixing with outsiders, the small losses compound until entire technologies disappear.
This is the part that should haunt anyone reading this in 2026.
Intelligence is not a property of the individual brain. Intelligence is a property of the network the brain is connected to. A genius in isolation will produce less than a mediocre thinker inside a dense exchange of other mediocre thinkers.
The thing your ancestors needed in order to break out of 190,000 years of stagnation was not better brains. It was better connections between brains they already had.
The implication for any individual is direct and uncomfortable. If you are smart and isolated, you will be outproduced by people half as smart who are connected.
The most successful people in any field are almost never the smartest people in it. They are the ones positioned at the intersection of the most idea flows. They are reading more authors than their competitors. They are talking to more people from more disciplines. They are in the rooms where ideas from different lineages bump into each other.
Ridley ends the book on the line that sounds optimistic but is actually a warning its this "The future will be invented by people who connect ideas, not by people who guard them."
3 of the 10 best-performing large caps over the past month make memory chips. The HBM supercycle continues.
1. Memory is carrying the whole top.
Four of the top twelve are memory and storage makers: 🇰🇷 Samsung +157%, 🇺🇸 Micron $MU +88%, 🇰🇷 SK hynix +82%, 🇯🇵 Kioxia +75%. They move in lockstep because the driver is the same: HBM/DRAM/NAND pricing supercycle feeding AI datacenters.
2. Asia is beating the US in hardware.
The top of the board belongs to Korea, Taiwan, and Japan: 🇰🇷 Samsung +157%, 🇰🇷 SK hynix +82%, 🇯🇵 Murata +86%, 🇹🇼 UMC $UMC +88%, 🇯🇵 Kioxia +75%, 🇹🇼 MediaTek +66%. The US mega caps look quiet next to them, 🇺🇸 NVIDIA $NVDA around +6%, 🇺🇸 Apple $AAPL +15%. The gains moved up the supply chain, toward the companies making the chips and components rather than the end-brands.
3. The rally is narrow.
Leaders ran up tens of percent, but the median sits near zero. Banks, staples, healthcare, and telecom are flat or down. A handful of semis and AI software names are doing the heavy lifting, with no broad participation behind them.
4. Money left defense, oil, and China.
The laggards cluster tightly: oil (🇧🇷 Petrobras $PBR −15%, 🇬🇧 BP $BP −11%, 🇳🇴 Equinor $EQNR −11%, 🇨🇳 CNOOC −10%), Chinese tech (🇨🇳 PDD $PDD −15%, 🇨🇳 Meituan −12%, 🇨🇳 BYD −11%, 🇨🇳 Alibaba $BABA −9%, 🇨🇳 Tencent −7%), staples (🇺🇸 Walmart $WMT −12%, 🇺🇸 PepsiCo $PEP −9%, 🇺🇸 Costco $COST −6%), and healthcare (🇺🇸 Regeneron $REGN −13%, 🇺🇸 HCA $HCA −13%). A textbook risk-on month, capital rotating out of defensives into AI beta.
Buffett and Terry Smith agree on one thing:
FCF/share is king.
It measures how much cash a business generates for each shareholder.
Here are 6 modern-day FCF/share compounding machines:
1) ServiceNow – $NOW: