💰 Don’t Let Doomerism Doom Your Portfolio

THE MAIN EVENTDon’t Let Doomerism Doom Your Portfolio 🧨

On Wednesday, Nvidia reported the kind of quarter that breaks your sense of scale.

Revenue hit $96.2 billion, up 106% from a year ago. Its data center business alone generated $89 billion, more than double last year’s total. Management expects the current quarter to come in at $108 billion, even while assuming zero data center compute revenue from China.

And this is not supposed to be the finale. Nvidia sees revenue growing roughly 70% next fiscal year, well above Wall Street’s previous estimate of 44%.

Yet Nvidia trades at roughly 18.5 times expected earnings over the next 12 months. That is a lower valuation than Coca-Cola, Microsoft, and even Cisco.

The market is pricing Nvidia as if its hypergrowth is about to fade. I think that skepticism has gone too far.

What Nvidia Actually Costs 🧾

A forward price-to-earnings ratio compares a company’s stock price with the profit analysts expect it to generate over the next 12 months. At 20 times forward earnings, investors are paying $20 for every $1 of expected annual profit.

For perspective, here are a few forward P/E ratios, alongside each company’s latest reported revenue growth:

• Apple: 33.2x. Revenue grew 16%.

• Coca-Cola: 25.8x. Revenue grew 7%.

• Microsoft: 24.8x. Revenue grew 18%.

• Cisco: 21.9x. Revenue grew 18%.

• Nvidia: 18.5x. Revenue grew 106%.

You tell me which one stands out.

In fairness, these are very different businesses. Coca-Cola’s stability, for example, may deserve a premium.

Meanwhile, Nvidia faces customer concentration, a potentially cyclical spending boom, and competition from custom chips developed by Google, Amazon, Microsoft, and others.

But those risks do not fully explain why a company growing revenue by 106% trades below the average S&P 500 stock.

Nvidia has spent the past three years repeatedly clearing Wall Street’s forecasts, often by wide margins. For today’s valuation to make sense, that streak does not merely need to slow. It needs to end soon, followed by a sharp drop in growth.

That is what the market is pricing in. The adoption data point in the opposite direction.

Why I Think the Market Is Wrong 🤖

In May 2024, Google processed 9.7 trillion AI tokens each month across its products and APIs. (Think of tokens as the small units of information processed by an AI model.)

By May 2026, that figure had reached more than 3.2 quadrillion. That is an increase of roughly 330 times in two years.

Even after that explosion, adoption remains… surprisingly early.

Only one in five US businesses currently uses AI in any business function. Among those that do, 57% use it in just one to three of the 15 functions tracked by the Census Bureau. Meanwhile, about half of American adults still say they never use an AI chatbot.

I know I am on the early end. I use AI every day to write this newsletter and run my businesses, and it returns far more value than it costs.

Most people I know are nowhere near that point. Some use it as a slightly better Google search. Some use it to decorate a room. Plenty avoid it altogether.

That is not evidence that AI has peaked. It shows how much unused runway there still is.

I expect AI to become as fundamental to work as search engines, spreadsheets, and email. Competing without it will eventually feel like showing up with an encyclopedia while everyone else has the internet.

The Risk That Gives Me Pause ⚠️

Competition is the real risk.

Google, Amazon, Microsoft, and other Nvidia customers are developing their own chips. If those chips become good enough, they will take market share from Nvidia.

But Nvidia can lose market share and still grow revenue, as long as the overall AI chip market expands faster than its share shrinks.

If the market grows 20%, Nvidia could lose roughly one-sixth of its current share and still sell as much as it did last year. If the market grows 30%, it could lose nearly one-quarter.

For Nvidia’s revenue to fall, its market share must decline faster than the entire market grows.

That makes the real bear case a double hit: custom chips take substantial share at the same moment AI infrastructure spending slows.

To be fair, regulation could slow the AI buildout. Governments may restrict chip exports, model development, or data-center construction. But I have a much harder time imagining demand collapsing while AI usage is multiplying and most businesses have barely begun adopting it.

Competition may divide the pie. But it does not mean the pie will stop growing.

The Bottom Line 🧭

I understand the unease. Nvidia is enormous, AI is disruptive, change can be scary, and the amount of money pouring into this is difficult to comprehend.

But I believe AI is more likely to become omnipresent than to flatline.

There is even an economic idea for this: Jevons paradox. When technology makes a useful resource cheaper and more efficient, we often consume more of it, not less.

As AI models become cheaper to run, companies will find more tasks worth handing to them. Lower costs can create more demand for computing power, not less.

That does not make Nvidia invincible. Competition, margins, and infrastructure spending still matter. But skepticism about AI and protecting your portfolio are not the same thing.

You can be an AI doomer. Just don’t let doomerism doom your portfolio.

FAST FACTSHomes Down, Premiums Up, Fees Capped 😵‍💫

🏠 Home Prices Are Falling In 64 Metros: One in five of the 300 biggest US housing markets saw prices drop year over year in July. Check yours. [Read]

🏗️ Builders Begin Bidding For Buyers: 63% offered incentives in August and 35% cut prices by an average of 6%, the 16th straight month above 30%. [Read]

💸 Your Health Premium Jumps Again: Big employers project a 9.2% rise for 2027, on top of the $6,850 the average worker already pays for family coverage. [Read]

👶 Trump Accounts Get A Fee Cap: Funds holding the one-time $1,000 for kids born 2025 to 2028 can charge just 0.1% a year under proposed IRS rules. [Read]

🥇 Even Gold Has A Limit: Morningstar says an individual should hold no more than 15%, with bullion near $4,700 and up just 8% this year. [Read]

💊 The Drug Nobody Approved: Eli Lilly found 14,000 listings across 100 countries selling retatrutide, a weight-loss shot no regulator has approved. [Read]

📱 Meta Settles For $12.1 Billion: The proposed deal with 51 state attorneys general limits under-18 users to two hours a day on its apps. [Read]

🇨🇦 The Trade War Restarts: New 50% duties hit $20 billion of goods with no USMCA exemption, and Canada retaliates on September 8. [Read]

CHART OF THE WEEKWhat Sobriety Did To Alcohol Stocks 🍷

America just told Gallup it drinks less than at any point since 1939. But the market already knew that.

If you put $10,000 into these five big alcohol stocks three years ago, with dividends reinvested, you would have just $7,653 today. The same $10,000 invested in the S&P 500 is $18,077.

The three-year damage, by name:

• Boston Beer: -49%

• Constellation: -46%

• Diageo: -40%

• Molson Coors: -27%

• AB InBev: +45%

Their own filings are blunter than any survey:

• Molson Coors and Constellation each took multi-billion-dollar write-downs over the past two years as U.S. beer and wine demand softened.

• Diageo cut its dividend in half, from about $1.03 a share to $0.50.

• AB InBev's no-alcohol beer grew 27% last quarter, while its US beer volumes fell.

There is a lesson in that last one. In a shrinking category, you do not want the biggest brand. You want to own whoever sells the replacement.

MARKET MOODTariff Refunds Lift Abercrombie, Sneaker Slump Sinks Dick's 👟

Winners

Abercrombie & Fitch ($ANF) - Market Cap: $6.5B (Week-to-Date: +33.7%)

Getting money back from the government isn’t normally an earnings driver, but Abercrombie booked a $100 million tariff refund this quarter. That helped it more than double the profit Wall Street expected. Net sales also rose 5% to $1.3 billion, and management raised its full-year profit outlook by about 25%.

CrowdStrike ($CRWD) - Market Cap: $232.1B (Week-to-Date: +18.8%)

You might remember CrowdStrike from July 2024, when a bad software update bricked 8.5 million Windows machines and canceled 3,000 US flights in a single day. Well, nobody left. Revenue beat expectations at $1.5 billion, the company swung from a loss to a profit, and it raised its growth outlook for the year. The stock now trades at more than double its pre-outage peak.

Synopsys ($SNPS) - Market Cap: $89.0B (Week-to-Date: +16.8%)

Two companies make the software used to design almost every chip on Earth, and Synopsys is the larger one with about 46% of the market. Nvidia, AMD, Broadcom and Apple processors all get drawn in it before a foundry ever builds them. Profits grew 15% this quarter, beating expectations, and management raised its full-year outlook for the second straight quarter. Almost as if chip design is in demand. 😉

Losers

Dick's Sporting Goods ($DKS) - Market Cap: $11.8B (Week-to-Date: -28.1%)

Dick's just had its worst day in years. Its own stores saw a healthy 4.9% comparable sales gain. But Foot Locker, which Dick's bought last year, saw comparable sales fall 3.6%. Management cut its full-year profit forecast by about 16%. CEO Lauren Hobart called the market "increasingly promotional," which is a polite way of saying everything is on sale.

Boston Scientific ($BSX) - Market Cap: $67.6B (Week-to-Date: -7.3%)

Two bad headlines landed on Boston Scientific inside the same 24 hours. A cyberattack on August 25 knocked systems offline, leaving hospitals unable to order or receive its devices. It is also pulling 165,000 units of an artery device off shelves over a tip that can break off mid-procedure. Both events hurt sales this quarter, sending the stock lower.

Sandisk ($SNDK) - Market Cap: $217.4B (Week-to-Date: -7.0%)

Sandisk makes flash memory, and lately its stock trades on headlines as much as on demand. Reports Monday said the Trump administration may let Apple, one of the biggest memory buyers on the planet, buy from Chinese suppliers. That would pull a huge customer away from every memory maker outside China, and cost them the power to charge a premium. Sandisk, Micron and Western Digital all fell.

WORDS TO REMEMBERSkepticism Has A Price Tag 🧠