*****************************************************************************

* FIEND'S SUPERBEAR MARKET REPORT *

* October 7, 2026 *

* *

* e-mail: fiendbear@fiendbear.com *

* web address: http://www.fiendbear.com *

*****************************************************************************

Fiend Commentary
================

The Boom Has a Power Bill

Wall Street found another reason to celebrate Tuesday.

The S&P 500 climbed 0.6% to a record 7,819, finally surpassing its August high. The Nasdaq added another 0.45% to close at 27,600, its second consecutive record. Even the Dow gained 253 points, although it remains more than 5% below its August peak.

A modest retreat in Treasury yields helped. The 10-year fell from 5.31% to approximately 5.27%, while oil stabilized around $100. That was apparently enough relief for investors to resume buying their favorite technology stocks.

After a September that featured one of the worst bond selloffs in years, investors seem eager to believe the worst is over.

Perhaps it is.

But the most interesting development Tuesday wasn't another Nasdaq record. It was an announcement that helps explain both the extraordinary technology rally and some of the economic pressures building underneath it.

Artificial intelligence has developed an enormous appetite for electricity.

The AI Boom Needs a Power Plant

Google announced a massive agreement with Constellation Energy to secure 3,590 megawatts of electricity for its expanding operations.

The agreement includes a 20-year contract for 890 megawatts of additional nuclear generation, along with a separate 15-year arrangement for another 2,700 megawatts.

Constellation plans to invest more than $4.3 billion in upgrades to its existing generating facilities.

Investors loved it.

Constellation shares jumped more than 12%, helping propel utilities to the best performance among S&P sectors Tuesday.

This illustrates something important about the AI boom.

The investment requirements are moving well beyond chips and data centers.

AI requires enormous amounts of electricity, transmission capacity, cooling infrastructure, construction materials and financing.

The largest technology companies are now effectively competing to secure enough energy to operate the computing infrastructure they are building.

And there is a problem.

Electricity isn't something that can be manufactured overnight simply because Google writes another check.

The PJM electricity market, which serves 13 states and Washington, D.C., has experienced an extraordinary increase in capacity prices. Reuters reports those prices have risen more than elevenfold since 2024, partly because of growing data-center demand.

That is an astonishing increase in the cost of securing reliable electricity supply.

The investment may eventually increase generating capacity and improve the situation. But the first upgraded nuclear facilities under Google's agreement aren't expected to begin delivering additional electricity until 2028.

The stock market prices the opportunity immediately.

The physical infrastructure takes years to build.

Someone Has to Pay for the Boom

This raises an interesting question about the relationship between AI and inflation.

Technology companies are spending unprecedented amounts on infrastructure. That spending creates demand for equipment, skilled labor, construction materials, electrical components and financing.

All of those resources have competing uses.

Power companies can benefit enormously from new demand, but electricity customers may face higher costs when available capacity becomes scarce.

The average business doesn't have Google's negotiating power or financial resources.

Neither does the average household.

That creates another divergence within the economy.

The AI companies benefit from extraordinary investment demand.

Energy companies benefit from selling the power.

Construction and equipment suppliers benefit from building the infrastructure.

But businesses and consumers outside that ecosystem may face higher costs for electricity and capital.

The same investment boom supporting record stock prices can also contribute to the cost pressures making life more expensive elsewhere.

None of this means AI investment is inherently inflationary over the long term. Greater productivity could eventually produce substantial economic benefits.

But building the infrastructure requires real resources today, while many of those productivity gains remain promises about tomorrow.

The distinction matters.

Thirty Percent Earnings Growth?

Wall Street is preparing for another extraordinary earnings season.

Analysts expect S&P 500 profits to increase approximately 30.6% year over year during the third quarter.

That would be a remarkable performance.

But look underneath the estimate.

Technology earnings are expected to rise around 66.5%.

Energy earnings are expected to increase an astonishing 114.7%.

Those two sectors help explain why the indexes remain so resilient despite high interest rates and slowing employment growth.

They also explain why the headline earnings forecast may not tell the entire story.

Energy profits can soar because oil prices rise dramatically.

That is wonderful for energy shareholders.

It is considerably less wonderful for consumers paying six dollars for gasoline or businesses buying eight-dollar diesel.

Technology earnings can surge because companies are spending hundreds of billions on AI infrastructure.

That is wonderful for semiconductor companies and suppliers.

Whether all that investment eventually generates sufficient economic returns is another question.

There is nothing wrong with strong corporate earnings.

But investors should distinguish between broad economic prosperity and enormous profits concentrated in a handful of sectors.

September demonstrated how different those two things can be.

The Breadth Improved—But Not Enough

Tuesday actually produced a welcome improvement in market breadth.

On the NYSE, 1,654 stocks advanced against 1,083 declining.

That is considerably better than the persistent negative breadth we've been documenting.

New lows also declined sharply.

However, InvesTech still recorded 134 NYSE new lows against only 50 new highs. On the Nasdaq, advances and declines were nearly equal, while new highs and lows were also almost evenly matched.

The Russell 2000 declined approximately 0.7% despite records in both the S&P and Nasdaq.

So there was some improvement.

Just not enough to repair weeks of internal deterioration.

InvesTech's A/D Divergence indicator actually slipped slightly further, to -5.59, its weakest reading of the current decline.

That is worth keeping in perspective.

The S&P has reached a record.

The typical stock has not.

Until the broader market begins participating more consistently, these records remain heavily dependent on continued strength from the largest companies.

Five Percent Is Still Expensive Money

Meanwhile, the bond market finally enjoyed a modest respite Tuesday.

The 10-year Treasury retreated to around 5.27%, while the 30-year eased to approximately 5.64%.

Those numbers hardly represent cheap financing.

The 10-year remains near its highest level in more than two decades.

The Fed has raised rates once this year, yet long-term borrowing costs have climbed dramatically without waiting for additional action.

This creates an interesting dilemma for the AI infrastructure boom.

The technology giants have enormous cash reserves and access to capital. They can finance projects that smaller companies cannot.

But utilities, construction firms, equipment manufacturers and other suppliers still have to justify investments at today's higher cost of capital.

A project that looked attractive when borrowing costs were 3% may look considerably less attractive at 6%.

The AI boom may generate enough profits to overcome those higher costs.

The rest of the economy isn't necessarily so fortunate.

And there is another complication.

The federal government is competing for the same capital while running enormous deficits and carrying more than $40 trillion in debt.

Everyone wants money.

Not everyone can afford it at the new price.

The Fed Is Becoming a Spectator

Tuesday's trading also reinforced the growing assumption that Kevin Warsh will skip another rate hike in October.

The probability of an increase has fallen to approximately 19%, down from more than 50% a week ago.

The weak September employment report provided the excuse.

Warsh will probably take it.

The minutes from September's Fed meeting arrive Wednesday and may provide additional insight into how determined policymakers really are to continue tightening.

But markets are already looking beyond October.

They are increasingly betting that the economy can remain strong enough to sustain earnings growth while inflation moderates sufficiently to keep the Fed from raising rates aggressively.

That is an attractive scenario.

It is also one that requires several things to go right simultaneously.

Energy prices must remain under control.

The labor market must stabilize.

Corporate earnings must justify elevated valuations.

And the bond market must stop demanding higher yields.

Perhaps Tuesday's modest bond rally was the beginning of that process.

Or perhaps it was simply another pause in a much larger repricing of the cost of money.

The Next Test Is Earnings

For now, Wall Street has an answer to nearly every concern.

High oil prices? Energy companies make more money.

AI infrastructure costs? Technology companies and utilities benefit.

Weak employment? The Fed stops hiking.

Higher Treasury yields? Strong earnings justify higher valuations.

The logic isn't necessarily wrong.

But it is becoming increasingly dependent on the assumption that corporate profits can overcome almost everything else.

That assumption gets tested when third-quarter earnings begin arriving next week.

A 30% increase in S&P profits would certainly help justify some of the optimism.

The more revealing question will be how widely those profits are distributed and what companies say about costs, demand and financing conditions heading into 2027.

So far, the winners have been spectacular.

The average company has been considerably less impressive.

Tuesday gave us another record in the indexes and a reminder that the AI boom is moving into a phase requiring enormous physical investment.

That investment may eventually transform productivity and economic growth.

It may also require years of expensive construction, scarce electricity and increasingly costly financing before those benefits fully arrive.

Wall Street is already celebrating the profits of tomorrow.

The bond market is charging for the money today.

And the rest of the economy is beginning to discover what that costs.

---

Weekly Market Summary Page
[Return to the Fiend's SuperBear Page]