Beige Book, Portfolio Review, & Top Trades (6/3/2026)
Timeline
0:00 — Market overview
1:40 — Housing and mortgage stress
5:29 — AI power demand
8:44 — Rising utility costs from AI
10:34 — AI pricing and subscriptions
12:42 — AI infrastructure overload
15:40 — AI monopolies and chip shortages
17:50 — Housing affordability crisis
21:23 — Consumer debt and car payments
21:56 — Jobs, inflation, and the economy
24:32 — Oil inventories and energy markets
26:48 — U.S. oil exports
29:30 — U.S. oil reserve depletion
34:19 — Global oil geopolitics
37:48 — World oil supply outlook
39:49 — Renewable energy transition
42:19 — Bezos, billionaires, and AI
49:33 — Webinar agenda and transition
50:02 — Recent top trades
50:53 — SoFi (SOFI)
51:34 — Medtronic (MDT)
52:32 — Macy’s (M)
54:20 — Alcoa (AA) and aluminum demand
56:06 — Trade review and performance
56:52 — Nike (NKE) discussion
59:30 — 2026 trade ideas
1:00:36 — Generac (GNRC)
1:01:36 — $700/month portfolio review
1:03:16 — Portfolio returns and options strategy
1:04:51 — Beige Book discussion
1:07:18 — Economic cycle and consumer fatigue
1:29:25 — Macy’s real estate value
1:30:51 — “Chipless” portfolio strategy
1:31:14 — Long-term portfolio holdings
1:40:08 — Diversification versus Magnificent 7 stocks
1:42:14 — Nike options question
1:50:00 — Portfolio allocation and risk management
1:52:00 — Hedging discussion
1:54:37 — Elon Musk IPO and market bubble concerns
Phil’s Stock World Webinar Transcript
Market Overview
Meanwhile, there’s a lot of cool stuff going on in the AI field. Let’s look at what the indexes are doing first. We’re down half a point on the S&P, half a point on the NASDAQ, and the Russell is down 1.5%. Ouch.
This is what we thought was going to happen — we just thought it would happen sooner. Consumers are completely screwed. It’s just horrifying.
Economic Data — Wednesday
Today is Wednesday and we have some data to go through.
Mortgage Applications: Down 2.5%. They just keep going down — last week was down 8.5%, and it keeps accumulating. This is happening because rates are going up, so houses get more expensive and more people get priced out of the market every time rates tick up. Another section of the population simply can’t afford homes anymore.
Also, how are you going to save up for a house when you’re spending $100 to fill up your gas tank?
Here’s the core problem: most analysts and investment bankers are upper-middle-class people who never knew what it was like to be poor. They never had to live paycheck to paycheck. More than half — probably 70% — of the people in this country are living paycheck to paycheck. They’re juggling credit cards, stretching payments, getting payday loans. There’s no budget cushion. And the analysts running commentary on TV don’t think that way because to get on TV as an analyst, you basically need an Ivy League degree, which means either you’re on scholarship and a genius, or your family paid the $80K a year tuition without blinking.
AI & Power Grid
I wanted to make an important point about power plants. It takes longer to build a power plant than a data center. So demand will outpace power supply for quite a long time as data centers multiply. Energy is going to get expensive.
There’s a finite amount of power. Everyone knows what a brownout is — when it’s too hot in the summer, the electric company throttles electricity to various areas so heavy devices may not work. But you can’t tell a data center to shut down between 11am and 2pm on a 100-degree day. Data centers run at full capacity 24/7, serving customers in America, Japan, Europe, everywhere. They don’t sleep. They don’t take holidays.
A single large data center can use as much energy as 500,000 homes — and the largest electric company in America serves 10 million customers. But it’s even more than that, because 500,000 homes don’t run at peak energy constantly. A data center does. It’s very understated how much power these things are going to draw off the grid.
Here’s the real problem for consumers. Say electricity is 5% of a data center’s total operating cost. If it doubles to 10%, that’s only a 5% increase in their costs — they raise their price from $19 to $20 a month for your subscription. But for the homeowner who lives near the data center and faces the same electricity cost increase, their bill doubles. So the data center’s customer pays $1 more per month, while the person living where the data center is built pays $200 more per month.
They are building hundreds of data centers around this country. Almost everyone in America will be affected. The true cost of your $20/month ChatGPT subscription could effectively be $220/month once you factor in your electric bill.
And on top of that, Anthropic has already announced they’re moving away from flat monthly plans to metered usage — so you’ll pay for X amount of computing time and then pay more beyond that. The “all you can eat for $20” model is going away. AI is becoming a resource that companies with deep pockets will be able to afford far more than individuals.
It’s like a drug dealer. They don’t hand you a gram of cocaine at full price right away. They give it to you cheap, get you hooked, and then start charging. Everyone’s been hooked on AI — and now they’re going to start charging.
You’re also going to pay with a bigger water bill. Water consumption from data centers is horrific. Electric costs go up, water costs go up, water treatment costs go up. It’s a massive tax on consumers and nobody is asking these companies to pump the brakes.
Does every American need access to OpenAI, Anthropic, Google, Facebook, and Microsoft’s AI — all simultaneously? And do all of those companies need to build their own redundant data centers everywhere? This is like the early telephone era, when every phone company was stringing their own wires everywhere and cities looked like a tangled mess of cables. Eventually the industry had to rationalize. We’re doing the same thing now — putting tremendous strain on the power grid and water systems with five-times-redundant infrastructure everywhere, with no urban planning.
The AI companies spent a lot of money putting the Republicans back in power to deregulate so they could do whatever they want. But they’re also crowding out competition — all chips are bought up for the next two years. You can’t start a competing AI company without billions of dollars, just like you can’t start an airline when Boeing is seven years backlogged. The incumbents have essentially locked up the market.
Back to Housing
The problem with housing isn’t just about interest rates. The difference between a 5% and a 6% rate on a $300K mortgage is maybe $50/month. That’s not the issue. The issue is people don’t have $100,000 for a down payment. Movers got expensive. Furniture got expensive. Appliances got expensive. The entire cost of moving has exploded. If you can barely pay rent, how do you get out of paying rent to go buy a house?
Healthcare is up, food is up, everything is up. 70% of the population simply does not have the money to buy a house.
The standard housing calculation assumes 110 million homeowners, with roughly 1 in 20 flipping their home every year — that’s theoretically 5 million homes turning over annually. We’re nowhere near that. We’re at about 3 million. Why? Because most existing homeowners can’t afford to leave the home they’re in. They can’t afford a new mortgage, a new down payment, or moving costs. So only about 30% of homeowners are actually able to transact — and they move once every 20 years — which means you effectively get down to maybe 1.5 million people who can buy per year, and you need a matching seller. The numbers get unworkable fast.
Oh, and the average car payment is now $800/month. I was floored by that.
ADP Employment
ADP employment came in strong. People are employed, but they’re still living paycheck to paycheck. Employment is what’s keeping things from completely collapsing. But it doesn’t take much — if you’re making $10K/month and living paycheck to paycheck, unemployment benefits of $1K/month won’t cover anything. That’s a brutal situation that’s more common than people realize.
Services PMI / Factory Orders
Services PMI came in at 50.7, down from 50.9. Not healthy — going in the wrong direction.
Factory orders were up 4.8%, but much of that is defense-related. ISM manufacturing is also getting stronger — but again, it’s being driven by the war. We’re bombing people and that’s showing up as economic strength.
Oil Inventories
Oil inventories decreased by 8 million barrels, but gasoline inventories increased by 3.4 million barrels. Distillates increased by 1.5 million, propane up 2 million.
8 million barrels came out of the Strategic Petroleum Reserve (SPR), from 365 down to 357 million barrels.
Compared to last year: commercial inventory is barely changed. Gasoline was 228 million barrels last year, now 215. Distillates went from 107 to 102. We’ve used 43-44 million barrels from the SPR, so about 10% has been withdrawn.
Here’s the sick part: last year we exported 5 million barrels of oil per day. This year it’s 6 million. We’re a country using 14 million barrels per day and exporting 6 million. We’ve increased export output by 7% — about 150 million extra barrels shipped out of the country since the start of the year.
During a war, when prices are going up and there are shortages, we’re exporting more and more oil. What kind of policy is that?
Now here’s the scary long-term picture. The US has 46 billion barrels of proven oil reserves. We’re extracting 20 million barrels per day — that’s 7.3 billion barrels per year. Divide 46 billion by 7.3 billion, and we have roughly 6 years of US oil production left.
That’s two years of Trump, then four years for whoever is next. Unless the next president takes drastic action, we will go from being an oil exporter to a 100% oil importer in six years.
Biden understood this. The logic of transitioning off oil was that if you have 10 years of reserves left, “drill baby drill” is suicidal. You should cut usage in half to extend your runway to 20 years and invest in alternatives. Instead, we’ve killed the clean energy programs, abandoned alternative energy infrastructure, and accelerated extraction. We will have burned through our reserves and have no alternative ready.
Exxon and Chevron know this — that’s why they’re heading back into Venezuela. Venezuela has 300 billion barrels. After being kicked out years ago, we’ve essentially forced our way back in.
At the global scale, the world uses about 100 million barrels per day — roughly 36 billion per year. Total world reserves are around 1,500 billion barrels. Divide and you get about 40 years of global oil supply remaining.
We are 5% of the world’s population consuming 20% of the world’s oil. Of course we’re running out first. And nobody has priced this in.
Beige Book Summary (via AI Analysis)
The Beige Book is not recessionary, but it’s clearly more uncomfortable than the last one.
Overall: The economy is still growing — 10 of 12 Fed districts reported slight-to-moderate growth. But the quality of that growth is deteriorating.
What Changed:
- Manufacturing improved in 9 districts, but that strength is being driven by a narrow band — defense, data centers, and industrial demand — and it’s not flowing through to the broader consumer economy.
- Consumer spending is increasingly bifurcated: higher-income households are resilient, middle-income households are stretching dollars, and lower-income consumers are visibly under strain.
Inflation: Prices increased at a moderate-to-strong pace, and most districts reported higher inflation than the prior report. It’s no longer just energy — it’s broadened to shipping, packaging, groceries, and fertilizer. Widening, worsening inflation is a big deal for the Fed.
Labor: Low hiring, low firing — a classic late-cycle condition. Employment showed little or no change in 11 districts. Workers are reluctant to switch jobs. The labor market apparatus is effectively broken: there are 7.6 million open jobs, but only 110,000 people were hired last month. At that pace, it would take 76 months to fill the open positions.
War Effects: Energy inflation is being directly attributed to the Middle East conflict. But the key point is: when the war ends, it won’t immediately fix things. The disruption to fertilizer supply during planting season is already locked in. If farmers couldn’t get fertilizer in March and April, there’s nothing to be done now in June. We’re going to see weaker harvests in the fall. That’s essentially guaranteed at this point and will show up in food prices.
Deeper Concerns:
- Consumer bifurcation is structural, not just war-driven.
- Rising delinquencies in residential mortgages, consumer loans, and agricultural loans.
- Margin compression: non-labor input costs are rising faster than selling prices.
- Commercial and residential real estate softness.
Bottom line: The Beige Book is hawkish not because growth is too hot, but because inflation is too sticky. The Fed can’t easily cut — the war gives them a dilemma because an energy shock can both hurt growth AND push inflation higher. Cutting into that risks validating another inflation wave.
Portfolio implication: This supports a barbell approach. On one side, companies with real demand exposure — defense, power, data centers, industrial. On the other side, avoid consumer discretionary, low-end retail, auto financing, and restaurants. This is a cost shock economy with pockets of industrial strength — and even those pockets are being driven by building data centers and military spending. That’s not a healthy economic foundation.
Top Trades — Past Month
We’ve had four top trade alerts since May 1st:
1. SoFi Financial (SOFI) We picked it up on the early May dip. It pulled back and we still like it. Phenomenal growth, great bank. Still completely playable at current levels.
2. Medtronic (MDT) Picked up May 18th. Medical devices — aging population, people need spare parts. They drifted into earnings and then shot back up. Great timing, everything filled nicely.
3. Macy’s (M) Good earnings — EPS beat 324%, revenue up 1.5%, raised full-year guidance. Q2 guidance was a little light, but if they’re raising full-year estimates, why are we upset about Q2? Doesn’t make sense. Still very playable. We like Macy’s primarily as a real estate play — their flagship Manhattan building alone on 34th Street may be worth what the whole company is being priced at. They own prime real estate across the country, much of it paid off decades ago, and they’re often the landlord for surrounding properties too.
4. Alcoa (AA) Aluminum is critically important for cooling data centers — it’s the primary material used to conduct heat away from equipment. You can’t just double aluminum production overnight; it takes years to bring a new mine online. Supply will remain heavily constrained. Plus aerospace uses it. Long-term story is excellent. The stock popped after we entered, pulled back a bit, but we still love it.
Portfolio Review
$700/Month Portfolio Up 289% over the past 12 months of reviews. Positions include: SoFi, Medtronic, Macy’s, Conagra, Helen of Troy, iPath, SQ (hedge), Owl, Cleveland Cliffs, Natural Gas (UNNG — confident it holds $11), Barrick Gold, Energy Transfer, Pfizer, Permian Resources, ULCC, Uranium.
Notable: this portfolio has no chip stocks. Deliberately chipless.
Long-Term Portfolio Started June of last year with $500,000. Now at $2 million — up roughly 300% in one year, or about 75% annualized. Diversified across many positions without chasing the Magnificent 7.
Key holdings include: AGNC (REIT), Barrick Gold, Best Buy, Barclays, Conagra, Constellation Energy (CEG), Cleveland Cliffs, Comcast, Coinbase, Crocs, Cisco, Enterprise Products Partners (EPD), First Solar, Gap, GEO Group, Google (one of the only Mag-7 names we hold), HP, H&R Block, IBM, Levi’s, Lockheed Martin, McDonald’s, Micron, Nike (struggling — probably shouldn’t still own it), Novo Nordisk, ON Semiconductor, Oracle, Owl Rock (Blue Owl), Occidental, Pfizer (stock of the year), PulteGroup, Pinterest, Qualcomm, Schlumberger, SoFi, Stellantis, Signature Bank, AT&T, Target, Toyota, Toll Brothers (taking a bath — need to review), Uber, UPS, Whirlpool.
Notable observation: Very few Magnificent 7 names. Deliberately. You don’t have to chase popular stocks to make great returns.
Quick note on Microsoft’s quantum chip: Microsoft just announced a 1,000x improvement in their quantum chip. This is a major problem for cryptocurrency — quantum computers will likely be able to crack blockchain encryption within 3 years. Every Bitcoin transaction relies on encryption that a quantum computer can break in seconds. This extends to bank accounts and credit cards. The industry has 2-3 years to implement quantum-resistant encryption. This is a ticking clock.
Q&A Highlights
On Jeff Bezos: I only vaguely remember seeing the interview mentioned. In general, even well-intentioned people who become extraordinarily wealthy very quickly lose touch with how normal people live. They’re surrounded by layers of staff and security — those same layers prevent them from having real conversations with ordinary people. They get out of touch fast.
On the Alcoa options trade (from Brad): The reason I sold 20 short puts versus 30 on the bull call spread is simple: I never want to be fully committed unless I absolutely love a name. I always want room to roll and double down if it moves against me.
The way I think about short put selling: Alcoa is at $82. The $55 puts are trading for about $9.25. If I sell those for $9.25, my net cost if assigned is $46 — that’s 44% below where the stock is trading. If it never gets there, I keep $18,000 in premium. If I do get assigned, I own a great stock at a massive discount. That’s essentially free money.
Think of it like a car dealer offering to pay you $10,000 today if you promise to buy a $100,000 car for $55,000 if it ever goes on sale. If it doesn’t go on sale, you keep the $10K. If it does, you buy a car you wanted at nearly half price. Why would you turn that down?
If things go wrong, I can roll my 20 short $55 puts to 40 short $40 puts — doubling down at a lower strike. The way I decide my position size is by imagining the worst case: if I end up owning 4,000 shares at $35, is that acceptable? At $140K with $70K of margin required, that fits comfortably within our allocation block. So yes, I’m fine with it.
On Alcoa in the $700/month portfolio (from Gabbor): I wouldn’t buy it there. It’s too expensive for that portfolio’s allocation model. If you have under $100K, you should divide your portfolio into 10 blocks of $10K. With Alcoa at $82, one contract could get you assigned on 100 shares at $8,200 — that’s nearly your entire block. You can’t double down, you can’t roll, you’re completely straitjacketed. Don’t play stocks you can’t maneuver around. For smaller portfolios, stick to stocks in the $20-$40 range.
On portfolio hedging with SQ and TNA (from Larry): The portfolio is broadly S&P-correlated. But my bigger concern is the NASDAQ. I think the NASDAQ is going to take a massive hit — consumers collapsing will kill the Russell, and the NASDAQ’s insane concentration (70%+ in the Magnificent 7 and their orbit) makes it extremely vulnerable. Nvidia can’t sustain 200% markups on chips forever. When competition arrives and margins compress, the multiple collapses.
The bigger potential trigger: SpaceX, Anthropic, and OpenAI combined represent nearly $4 trillion in expected market cap. Their IPOs will demand roughly $200 billion in cash from the market. That money has to come from somewhere. If SpaceX’s IPO prices at $1.75 trillion and subsequently underperforms, it could trigger a broad reassessment of all tech valuations — the kind of moment that causes cascading failures, similar to what ended the dot-com era.
On AT&T downgrade (from JC): Of course I still like AT&T. It’s the phone company. They trade at 9x forward earnings and just keep cranking out money. Their earnings consistently beat. The channel is roughly $25-$30. At current levels, I’d look to buy back the short calls we sold and wait for the next bounce to sell them again. They could go down to $18 before I’d get really aggressive, but I have very high confidence they hold the $18-$20 floor.
In fact, my daughter owns AT&T as one of her two stocks (along with Ford). She bought at around $18, sold $20 calls for a couple bucks, and collects the 4.5% dividend. With her cost basis around $16, she’s effectively collecting 7-8% yield. Simple trade, but very effective.
Thanks everyone for coming. See you next week.
