To the clients and friends of ProfitScore:

Are We About to Get Paid Nothing to Own Stocks

In our research, we are always trying to identify market risks.  I present to you some analysis.

The S&P just closed April at an all-time high of 7,209.

Q1 earnings? Crushed it. 84% of companies beat estimates, the highest beat rate since Q2 2021. Net profit margins hit 13.4%, an all-time record.

Analysts are forecasting 21% earnings growth for 2026.
Everything looks fine. Right?

Not so fast, Skippy!

What if I told you: the equity risk premium is gone.

The Math

The forward 12-month earnings yield on the S&P 500 is roughly 4.8%. The 10-year Treasury is yielding roughly 4.5%.

The premium you’re being paid to take stock market risk, the entire reason equities exist as an asset class above bonds, is roughly thirty basis points.

Historical average ERP: 4-5%. We’re at a tenth of that. Oppenheimer called it “among the lowest on record.”

Equity-level drawdown risk. Treasury-level returns.

Kinda like Russian Roulette, right?

The story this chart tells in one glance: the blue area is what equity investors used to get paid. It’s almost gone. And the only other time it disappeared this completely was right before the dotcom crash, visible on the far left where the lines actually cross.

“But Earnings Are Great”

Yes. They are. That’s all bull’s right? That’s the trap.

The market isn’t expensive because the future looks dark. It’s expensive because it has already priced in the best earnings environment in modern history.

What happens if margins mean-revert? If the AI capex cycle slows? If Q2 surprises don’t keep coming in 20% above estimates?

You don’t need a recession to crack this market. You just need anything less than perfect.

The Concentration Trap

The S&P 500’s three-year total return is 86%. The equal-weight index? 43%.

That gap has happened exactly once before. The late-1990s tech bubble.

When a client says they own “the market,” what they actually own is a leveraged bet on a handful of names. The index is a momentum factor; what could go wrong? 
 

You are not factoring in AI productivity! Ahhh….. yes, good ol’AI.

I know what you’re thinking. This time is different. AI will deliver productivity gains that justify the multiple.

Maybe. The bull case has teeth. AI is a genuine game-changing technology, closer to electricity than to crypto.
But “the technology is real” and “today’s prices pencil out” are not the same statement.

Ask the 1999 internet bulls. They were 100% right about the technology. And Cisco still fell 90%, and didn’t recover its 2000 peak for over twenty years.

$2.1 trillion in AI capex is not a return. It’s a wager. Capex booms historically wipe out a lot of the people who fund them, even when the infrastructure they build matters. Ask Lucent. Or Nortel. Or WorldCom.

And margins at all-time highs aren’t a baseline. They’re a ceiling. You cannot pay a premium multiple on record margins by also assuming margins go higher from here.

Believe in AI. Just don’t confuse believing in AI with believing in today’s prices.

The Question Worth Thinking About

When the equity risk premium was 5%, everything made sense. The math paid you to ride out the volatility.

When the equity risk premium is 0.3%, I am not going to lie, I lose a little sleep.

Remember what I wrote in February about correlation breakdown?

Diversification fails right when you need it most. Valuation traps work the same way.  Also, last month, when we talked about everyone using AI to find trading systems and the potential problems that may (probably will) arise.

The Bottom Line

Stocks at all-time highs. Margins at all-time highs. ERP at all-time lows.

You can argue with any one of those in isolation. It is hard to argue with all three at once.

I know, I know, there are always naysayers, and most of them (not all) are broke.  I am just a curious researcher and would like to share some interesting thoughts with you all.