To The Moon

The S&P 500 Is Calm. The Market Beneath It Is Not.

Why a 5% Treasury yield, collapsing breadth and AI concentration could be the real story of Q4

For investors looking only at the headline index, the U.S. stock market still looks remarkably healthy.

The S&P 500 remains within roughly 2% of its August record high. The Nasdaq continues to hold up. AI-related mega-cap technology stocks are still attracting capital, and fresh earnings from companies such as Micron have once again demonstrated that the AI investment cycle is producing very real revenue and profit growth.

But beneath that surface, something much less comfortable is happening.

The market is increasingly behaving like two completely different markets at the same time.

One market consists of the mega-cap technology and AI companies that continue to command enormous amounts of capital.

The other contains banks, utilities, small caps, mid-caps, rate-sensitive businesses and companies without a convincing AI growth story.

And that second market has already been taking significant damage.

The question heading into October is therefore not simply:

“Can the S&P 500 make a new high?”

It is:

“How long can a handful of companies keep the entire index looking healthy while the majority of stocks weaken underneath?”

---

The most important number may not be the S&P 500

On September 30, the S&P 500 closed at 7,651.54, still up 11.8% for the year. The Nasdaq was up 15.6%, while the Russell 2000 remained up 12.7% YTD.

Those numbers sound reassuring.

But September revealed a completely different picture.

Nasdaq's September market review showed that the median U.S. benchmark declined 2.7%, while the cap-weighted S&P 500 held up much better. The Russell 2000 fell 5.3%, while the Nasdaq-100 gained 3.3%, helped by large-cap technology and communications stocks.

Even more striking was the performance gap inside the S&P 500 itself.

The equal-weight S&P 500 fell about 4.4% in September, while the conventional market-cap-weighted index was comparatively resilient.

That distinction matters enormously.

A market-cap-weighted index gives the largest companies the greatest influence.

So when Nvidia, Microsoft, Apple, Amazon, Meta and other mega-cap companies rise, they can offset weakness across hundreds of smaller constituents.

The index can therefore say:

> “Everything is fine.”

While the average stock is saying:

> “Not really.”

---

This is a breadth problem, not necessarily an index problem

Market breadth measures how widely a rally is being shared.

A healthy bull market generally has increasing participation: more companies move above their moving averages, more stocks make new highs, and gains gradually spread from leaders into other sectors.

The current market is moving in the opposite direction.

MarketWatch reported that NYSE stocks hitting new 52-week lows have outnumbered new highs for 23 consecutive sessions, while the equal-weight S&P 500 is approaching a rare seventh consecutive weekly decline.

The equal-weight streak is particularly interesting because it has historically been unusual. A seven-week losing streak would be only the third such occurrence, with previous episodes occurring around the 2002 dot-com bust and the 2022 bear market.

That does not automatically mean another 2002 or 2022 is coming.

History rarely repeats that neatly.

But it tells us something important:

The current divergence between index performance and individual-stock performance is historically unusual.

And unusual markets deserve more attention than markets that simply go up or down.

---

Then came the bond market

If there is one macro variable capable of changing the entire equation, it is the Treasury market.

On October 1, the U.S. 10-year Treasury yield briefly reached approximately 5.34%, its highest level since 2002. Reuters reported that the yield subsequently eased toward 5.2%-plus levels, but the move itself was significant.

The Treasury selloff was even more remarkable because the 10-year yield had already experienced its biggest quarterly rise in decades.

Nasdaq's September review showed the 10-year yield rising about 54 basis points, ending around 5.29%, while the 30-year Treasury yield reached roughly 5.62%.

This creates a completely different environment for stocks.

For years, investors became accustomed to thinking about bonds as the boring alternative.

Now the question becomes:

If a relatively low-risk government bond can yield more than 5%, what return should investors demand from equities?

That is where valuation becomes much more important.

A company trading at 40x earnings is not valued in isolation.

It is competing with the entire capital market.

And when the risk-free rate rises, the valuation investors are willing to pay for distant future earnings can fall.

---

Why small caps are feeling the pain first

The Russell 2000 provides perhaps the clearest demonstration.

Small companies tend to be more sensitive to financing conditions because they generally have less access to cheap capital, smaller balance sheets and greater dependence on external financing.

The Russell 2000 fell approximately 8.9% from its August 14 record close by late September, approaching the conventional 10% correction threshold.

And during Q3, the Russell 2000 ultimately declined 7.2%, while the Russell Midcap Index fell 3.0%.

This creates a fascinating split.

Large technology companies can potentially finance enormous AI investments from operating cash flow.

A small industrial company with expensive debt doesn't have the same luxury.

When the cost of capital rises, the market starts asking very different questions:

Can this company refinance?

Can it survive another two years of high rates?

Is its cash flow sufficient?

Does it actually have pricing power?

Can it grow without continually issuing debt or equity?

That is why a high-rate environment can expose differences that were invisible when money was cheap.

---

The “zombie company” problem

This is where the situation becomes more serious.

According to the data cited in the original market analysis, more than one-third of Russell 2000 constituents fall into the category of companies whose operating profits struggle to cover their interest expenses.

These companies are particularly vulnerable if rates remain elevated.

Importantly, this does not mean one-third of small-cap companies are about to collapse.

It means their financial structure leaves them more exposed to a prolonged period of expensive capital.

That distinction matters.

A company can survive high rates.

It is much harder to survive high rates plus weak demand plus refinancing pressure plus falling margins.

And this is exactly why bond yields are more than a macroeconomic statistic.

They are becoming a test of corporate balance-sheet quality.

---

Banks are sending another warning

Banks should theoretically benefit from a strong economy and active lending.

Yet the banking sector has not participated in the market's headline strength in the same way as AI-related technology.

The KBW Nasdaq Bank Index ended Q3 with a 5.48% YTD gain, but its performance has deteriorated materially from its mid-August peak.

Financials were also among the weakest S&P 500 sectors during September, while the broader market remained supported by mega-cap technology.

Why does this matter?

Because banks are effectively sitting at the intersection of several risks:

higher funding costs + credit quality + loan demand + bond-market volatility + economic expectations.

A market dominated by technology can tolerate weakness in banks for a while.

But if financial weakness spreads into credit markets, the implications become much broader.

---

And then there is the strange case of utilities

Utilities used to be one of the classic defensive trades.

Predictable cash flow.

Stable demand.

High dividends.

Low growth.

But the problem with a high-yield environment is obvious.

If Treasury securities suddenly offer attractive yields, investors don't necessarily need to accept expensive valuations for slow-growing dividend stocks.

That helps explain why utilities have suffered.

According to a Q3 market review, the S&P 500 utilities sector fell 12.4% during Q3, its worst quarter since Q1 2020.

This is particularly interesting because utilities are now caught between two forces.

On one side:

higher interest rates hurt valuation and financing.

On the other:

AI is creating enormous demand for electricity and data-center infrastructure.

So the long-term AI electricity story can be extremely bullish for power demand while utility stocks themselves can simultaneously struggle under the weight of higher financing costs.

That is an important distinction investors often miss.

A good industry story does not automatically mean every stock in that industry is a good trade at every valuation.

---

The paradox: AI is both the strength and the risk

And this brings us to the most important part of the story.

AI is currently doing something extraordinary for the U.S. market.

It is generating enormous capital expenditure.

It is increasing demand for GPUs, networking, memory, data centers and electricity.

And increasingly, those investments are showing up in corporate earnings.

Goldman Sachs Asset Management noted that adjusted S&P 500 earnings grew 32% year over year during Q2 2026, with more than 85% of companies beating expectations.

So the AI story is not simply hype.

There is real economic activity underneath it.

But that is precisely why the next stage matters.

The market has moved from:

“Will companies spend money on AI?”

to:

“How much money will they spend?”

And eventually:

“What return will they earn on that spending?”

That is a much harder question.

Goldman Sachs strategist Ben Snider recently argued that nearly half of 2026 S&P 500 earnings growth has been linked to AI-related investment and warned that the current AI capex contribution may not remain at the same pace into 2027.

If AI investment keeps producing accelerating revenue and profits, the market can potentially absorb high valuations.

But if AI spending slows while expectations remain extremely high, investors could suddenly start questioning the assumptions embedded in today's prices.

---

This is where the market becomes asymmetric

Imagine two scenarios.

Scenario A: AI earnings continue to accelerate

AI infrastructure spending remains strong.

Semiconductor demand remains healthy.

Cloud providers continue increasing capital expenditure.

AI applications begin generating measurable revenue.

Corporate productivity improves.

In that environment, today's market concentration could continue for longer than skeptics expect.

The weak breadth could eventually reverse as earnings spread from the AI leaders into industrials, energy infrastructure, utilities and other beneficiaries.

Scenario B: AI expectations begin to crack

Suppose AI capex growth slows.

Or semiconductor margins compress.

Or data-center returns disappoint.

Or enterprises discover that AI monetization takes longer than expected.

Suddenly the market loses its primary growth engine.

At the same time, Treasury yields are above 5%.

Small caps are already weak.

Banks are under pressure.

Utilities have already experienced a significant drawdown.

Market breadth is deteriorating.

In that situation, investors may no longer have a broad group of alternative sectors ready to absorb capital.

And that is the real vulnerability.

---

The market isn't necessarily predicting a crash

This is where investors need to avoid the opposite mistake.

Weak breadth does not automatically equal a coming crash.

The S&P 500 remains profitable.

The U.S. economy remains relatively resilient.

Corporate earnings remain strong.

AI investment remains substantial.

And large technology companies have balance sheets that are fundamentally different from the speculative companies of the dot-com era.

Even the unusual breadth statistics should therefore be treated as warning signals, not crash predictions.

In fact, there is another interpretation.

Perhaps the market is simply repricing old-economy assets while capital concentrates in the companies experiencing the strongest structural growth.

That would make the current divergence less like a bubble and more like an aggressive economic transition.

Both interpretations remain possible.

---

The real question for October

This is why I think watching the S&P 500 alone is becoming increasingly misleading.

For October, I would watch five things:

1. The 10-year Treasury yield

If yields stabilize or retreat, pressure on rate-sensitive stocks could ease.

If yields continue pushing above 5.3%-5.4%, valuation pressure could intensify.

2. Equal-weight vs. market-cap-weight S&P 500

If equal-weight starts outperforming, that would indicate improving participation.

If mega-caps continue rising while equal-weight keeps falling, the concentration problem becomes more extreme.

3. Russell 2000

Small caps are effectively the market's stress test for financing conditions.

A sustained recovery would suggest that investors are becoming more comfortable with rates.

Continued weakness would suggest the opposite.

4. Corporate earnings

Not just revenue.

Not just AI announcements.

The critical question is whether AI investment is translating into sustainable earnings and free cash flow.

5. Market breadth

New highs versus new lows.

Stocks above their 50-day and 200-day moving averages.

Advance-decline trends.

These indicators tell us whether the rally is becoming healthier—or increasingly dependent on fewer companies.

---

The biggest lesson

The most dangerous market is not always the one that is falling.

Sometimes it is the market that looks completely fine from a distance.

Today, the S&P 500 still looks strong.

But underneath it, the equal-weight index is weak, small caps are struggling, banks have lost momentum, utilities have been hit hard, Treasury yields have surged and market breadth has deteriorated sharply.

At the same time, AI continues to provide a powerful earnings and growth engine—and that is precisely what is holding the headline index together.

So the investment debate for Q4 is becoming much more interesting than simply bull versus bear.

It is really a question of concentration versus participation.

If AI earnings continue to exceed expectations, the market can potentially remain surprisingly resilient despite historically high yields.

But if AI expectations weaken at the same time that Treasury yields remain elevated, investors may discover that the apparent strength of the index was hiding a much weaker market underneath.

The S&P 500 may be only a few percent away from another record.

But the average stock is telling a very different story.

And in the months ahead, that difference may matter more than the headline index itself.