home

Unconvention

Blog

About

Media

Reading List

Contact

Overpriced Compared to Which Normal?

The valuation alarms assume prices will go back to normal. The old normal is gone.

November 17, 2026

Overpriced Compared to Which Normal?
finance

Share

This piece first appeared as a guest post in Money Machine Newsletter.

Charles Schwab's asset management arm expects U.S. large-cap stocks to return 5.9% a year over the next decade. It expects the U.S. bond market to return 4.8%. That's roughly one percentage point a year for taking on stock market risk.

For anyone who learned the capital asset pricing model, that should be jarring. CAPM says the overall market should earn the risk-free rate plus a premium for bearing market risk. Using Schwab's own long-term averages, stocks have beaten cash by about 6.5 points a year since 1970 (11.1% vs 4.6%). Schwab's new forecast puts that gap at 2.6 points, and against the 10-year Treasury it pegs the equity risk premium at about 2%. Taken at face value, the market is offering investors barely any reward for owning stocks.

The other alarms are ringing too. The Shiller CAPE ratio is near 40, a level the market has reached in only about 3% of months since 1957. Only the dot-com peak was higher. Warren Buffett's favorite gauge, total market value relative to GDP, sits around 236% against a long-term average near 168%. Buffett once said that a reading approaching 200% meant you were "playing with fire."

I'm not going to argue with any of those numbers. They're real, and anyone who invested through 2000 or 2008 should take them seriously.

But most people read them as a forecast that things will go back to normal: multiples drift down toward their long-run averages, the reward for owning stocks recovers, and patient investors get to buy at sensible prices again. That reading assumes the market that produced those averages is the market we still have. It isn't.

Source: Schwab Asset Management, Schwab's 2026 Long-Term Capital Market Expectations, Jan 2, 2026 (data through Oct 31, 2025). Annualized returns; historical averages use monthly data from January 1970.

Normal Was a Market That Broke Up Its Winners

The long-run averages behind every one of those gauges were set over a century in which Washington regularly cut its biggest winners down to size. Standard Oil was broken up in 1911. AT&T was split apart in 1984. IBM spent 13 years fighting a federal antitrust suit. In 2000 a judge ordered Microsoft broken in two, and although an appeals court threw out the breakup, Microsoft spent the next decade operating under a consent decree.

That threat put a ceiling on how big a winner could get, and on what investors would pay for one. Part of what kept the average P/E ratio average was the expectation that no company would be allowed to stay dominant forever.

The New Normal Protects Them

That expectation is gone. If you're interested, I wrote a series on how monopolies form, from Ticketmaster and Amazon to the meatpackers squeezing farmers and the platforms small businesses rent their customers from. The pattern is the same every time. A company wins by being useful, grows until customers and suppliers have nowhere else to go, and then starts collecting. What's new is that Washington has stopped standing in the way.

A federal judge ruled in 2024 that Google illegally monopolized search. When it came time to fix it, the court declined to make Google sell Chrome and imposed rules instead. A year later the FTC lost its case to unwind Meta's purchases of Instagram and WhatsApp. A jury did find against Live Nation this year, but that's despite the fact that the Trump administration attempted to settle for a small fine. For the largest platforms, the answer is rules and settlements, not breakups.

And Washington isn't just looking the other way. It's giving permission. The government took an equity stake in Intel in 2025, and it agreed to let Nvidia and AMD sell certain chips to China in exchange for 15% of the revenue. A government that owns a piece of its national champions, or takes a cut of their sales, isn't going to break them up. It's a partner.

The winners are using every inch of that room. Microsoft, Alphabet, Amazon and Meta are on track to spend roughly $730 billion on capital expenditures this year, with more than $1 trillion expected next year.

Source: CNBC, Oct 29, 2026; ValueAdd VC compilation of company guidance. Combined capital expenditures of Microsoft, Alphabet, Amazon and Meta. 2026 is guided; 2027 is the expected level (Goldman Sachs projects about $1.14 trillion).

The cautious crowd has one question about all that spending: where's the profit? In mid-2024, Goldman Sachs published a research report titled "Gen AI: Too Much Spend, Too Little Benefit?" Two years later, Goldman projects the same four companies will spend about $1.14 trillion in 2027. The skeptics keep checking the income statement for the payoff. They're looking in the wrong place.

The spending isn't just there to deliver profits. It's there to create distance. Every data center, custom chip and power contract is something the next competitor has to match before it can even start, and almost nobody can write that check. A startup with a better AI model still has to rent computing power from the companies it's trying to beat. Each quarter of spending pushes the nearest competitor further back, and the ones that can't keep up get bought, starved or quietly shut down. That isn't reckless spending. It's the most expensive moat ever dug, and it's being dug with permission.

The market has noticed. At the end of the third quarter, the 10 largest companies made up 39.2% of the S&P 500's value, roughly double their typical share in earlier decades.

Source: J.P. Morgan Guide to the Markets via Landmark Wealth (typical share in earlier decades, approximate); Seeking Alpha / Political Calculations (Sept 30, 2026).

A monopoly nobody is going to break up deserves a higher multiple than one that might be broken up, because its profits last longer. If we've decided to tolerate global monopolies, and sometimes to back them, then today's multiples aren't a bubble waiting to revert to the old average. They're the price of the new normal. Waiting for P/E ratios to come back down means waiting for a market that no longer exists.

How the Winners Build Moats at This Scale

That doesn't make every stock a buy. It has always been true that a few companies do most of the work. Hendrik Bessembinder, a finance professor at Arizona State, found that just 46 companies produced half of the roughly $91 trillion in wealth the U.S. stock market created since 1926. What's changed is how much the few have to fear. So the question is which companies are building moats that will hold. The ones that stayed on top through more than one technology shift did four things.

1. They corner the control point

Nvidia was founded in 1993 as a graphics chip company. In 2006 it launched CUDA, a software platform that let developers use its chips for general computing. That software layer became the place where AI developers build, and the chips followed. Apple did something similar with the App Store in 2008. It didn't need to make every app. It needed to own the store.

Cisco did the opposite. Between 1993 and 2000 it bought 71 companies rather than owning the layer that would matter next. Yahoo bought 114 and passed on Google twice. In March 2000 the market priced Cisco at more than 200 times earnings, and it took 25 years, 8 months and 13 days for the stock to get back to that price.

2. They cannibalize their own core

The iPod once made up nearly half of Apple's revenue. In fiscal 2008 it brought in about $9 billion, while the brand-new iPhone brought in $1.8 billion. Apple pushed the iPhone anyway, and by fiscal 2011 iPhone revenue was $45.9 billion. Amazon did the same in retail, opening its store to outside sellers who compete with its own inventory. Third-party sellers went from 3% of its physical merchandise sales in 1999 to 58% in 2018, and today they account for about 61% of units sold.

Source: Apple 10-K filings via Quartz, Fortune and Business of Apps (fiscal years; 2007 and 2008 iPhone revenue as originally reported, when Apple spread iPhone sales over two years, 2009 onward as restated). Amazon 2018 letter to shareholders (share of physical gross merchandise sales, 1999 to 2018) and Q1 2024 earnings call (share of paid units, dashed).

Intel shows what happens when a company protects its core instead. It held on to its PC processor business, missed mobile, and fell behind in manufacturing. Even a government stake hasn't fixed that: last quarter Intel's foundry business booked just $411 million in outside revenue alongside a $1.9 billion operating loss. The new normal protects moats. It doesn't dig them.

3. They buy the challengers early

Google bought Android in 2005, YouTube in 2006 for $1.65 billion, and DeepMind in 2014. Microsoft bought LinkedIn for $26.2 billion and GitHub for $7.5 billion, then backed OpenAI early. After OpenAI's 2025 restructuring, Microsoft's roughly 27% stake was valued at about $135 billion. With the FTC's case against Meta's Instagram and WhatsApp deals lost, buying the company that could threaten you looks safer than it has in decades.

4. They fund the fight from cash flow

The big four's $730 billion capex bill this year is being paid mostly from operating cash, not borrowed hope. Meta and Alphabet both reported negative free cash flow last quarter as spending outran cash generation, and the skeptics took it as proof of excess. It's the opposite. A company that can push its free cash flow below zero on purpose, and fund it from its own operations, is choosing distance over this year's earnings. No challenger has that option.

Picking Winners Still Means Picking

Microsoft is the reminder that price still matters, even for a winner. Its stock peaked in December 1999 and didn't regain that level until October 2016, nearly 17 years later. It largely missed mobile and search in those years, and it wrote off about $6.2 billion on the aQuantive ad business and $7.5 billion on Nokia's phone business. The comeback started when Satya Nadella pushed the company toward the cloud. Azure grew 48% last quarter.

Source: GeekWire and Seattle Times (Microsoft record close, Oct 21, 2016); FT via Slashdot and Parameter (Cisco record close, Dec 10, 2025).

Microsoft came back because it went back to running the playbook. Cisco's shareholders waited a generation because it never had one.

Now Is the Time to Pick

Schwab may well be right about the average. The S&P 500 might return something like 5.9% a year over the next decade. But the average is the wrong thing to wait on. If the old normal isn't coming back, neither are the old prices for companies that own their markets. The winners of this era are being set now, while their moats are still being dug.

The cautious fallback is the index fund: don't pick, just own everything. Fine, but know what you're buying. The big winners are in there, and so are the losers that will drag down the average. In the third quarter, while the top 10 grew, the other 493 companies in the S&P 500 lost 2.4% of their combined value. In the new normal, an index fund is a bet on the monopolies, watered down by a bet on everyone they're putting out of business.

It won't feel obvious. Forbes estimated in 2025 that if Bill Gates had never sold or given away his original Microsoft shares, they would be worth about $1.4 trillion. Through Microsoft's lost decade, selling looked like the wise call. Holding a winner never feels obvious while you're living through it, which is why you need tests rather than stories.

So the next time someone tells you the market is overpriced and will go back to normal, ask which normal they mean. The one where Washington broke up its winners is over. The question now is which winners you want to own.

This article is for informational purposes only and is not investment advice or a recommendation to buy or sell any security. Do your own research and consider your own goals and risk tolerance before investing.

Unconvention book cover

Enjoyed this article?

These ideas come from Unconvention — Sri Kaza's guide to competing smart as a small business.

Learn more

Related Articles

More articles

Copyright 2026

Sri Kaza