During the third quarter of 2026, there seemed to be only three things people wanted to discuss about the economy: energy prices, the Federal Reserve and data centers—and not necessarily always in that order.
Energy Prices and the Consumer
Obviously, people want to know when energy prices are going to start falling. I can’t imagine anyone enjoys paying. AAA reported that the national average on Sept. 24 was the highest ever for that time of year. Through Sept. 24, the average price for gasoline across the United States was $4.48 per gallon. The previous high for September was back in 2023 at $3.83. (1)
But when are consumers going to start feeling some relief?
Historically, gasoline prices typically fall during the fall after heavy consumption during the summer. This makes perfect sense as people hit the nation’s highways during the warmer months. However, as the above paragraph suggests, this hasn’t necessarily been the case in 2026.
Filling your tank can be downright depressing, and I wish there were some end in sight. Unfortunately, the problem with crude oil prices seems to be more geopolitical than economic.
As most people probably know by now, the U.S. started bombing Iran at the end of February 2026. At the end of that month, the ‘generic’ crude oil future on the NYMEX closed at $67.02/barrel. It had closed 2025 at $57.42/barrel. (2)
On Sept. 25., this same future was trading at $92.41/barrel, and the United States is still embroiled in the Middle East with no apparent end in immediate sight. As a frontpage headline on wsj.com read on Saturday morning (Sept. 26: “Trump Rejects Iran Ceasefire, Expects Renewed Bombing After Midterms.” (3)
The first two paragraphs of the insightful article by Alexander Ward in Washington and Summer Said in Dubai painted a pretty bleak picture for the U.S. consumer, potentially through the end of the calendar year:
President Trump has rejected Iran’s proposal for a seven-day ceasefire and has told aides he expects to resume bombing Iran after the November midterms, U.S. officials said.
Tehran’s proposal would have reopened the Strait of Hormuz and resumed nuclear talks in return for the U.S. lifting its blockade of Iranian ports, which is causing deep damage to the country’s economy. Trump says openly that Tehran is begging for a deal after the midterms that dismantles its nuclear program.
Basically, I would not expect significant relief in crude oil and gasoline prices unless the “situation” in the Middle East and the Strait of Hormuz meaningfully improves. Of course, stranger things have and will happen. However, crude prices soared AFTER the conflict started, and have remained much higher than they were as the situation has snarled.
Obviously, these higher energy prices have contributed to upward pressure on inflation overall, and the Federal Reserve doesn’t like that sort of thing. (4)
A Shift in Interest-Rate Expectations
According to the FedWatch tool on cmegroup.com, at the start of 2026, the Fed Funds futures market didn’t think there was any chance the Fed would raise the target overnight lending rate this calendar year. By that, I mean there was 0% probability. In fact, the futures market thought rate cuts were far more likely, perhaps even as many as three or four 0.25% reductions. (5)
However, thanks in large part to higher energy prices, the Consumer Price Index (CPI), the most commonly used inflation gauge in the U.S., has moved away from the Fed’s preferred direction thus far in 2026. The preceding link will take the reader to a very good article at Schwab, which explains how crude oil prices impact inflation. To that end, at the end of 2025, the trailing 12-month U.S. CPI Urban Consumers was 2.7%. At the end of August, it was 3.4%.
Clearly, that is moving in the wrong direction for monetary policy decision makers. And, with the situation in Iran apparently far from over, there is little reason to believe the CPI will fall substantially, or at least to a level where the Fed feels comfortable.
As such, the Fed actually raised the target range by 25 basis points (3.75%-4.00%) at the Sept. 16, 2026, meeting. (6) As I have previously pointed out, futures markets entered the year assigning essentially no probability to a 2026 rate increase. Better put, no one who was willing to put their money where their mouth is believed they would.
Now, as I type on Sept. 26, 2026, the same FedWatch tool at the CME is predicting the Fed might have an additional three or four more 0.25-percentage-point rate hikes left in this ‘tightening cycle.’ That would take the overnight rate to at least 4.75%, which would likely cause some budgetary pain for borrowers who have balances on any revolving debt, whether they be credit cards, HELOCs, personal lines of credit or other.
You can sum up everything I have written thus far thusly:
The Fed could remain inclined to raise the target overnight rate as long as inflation remains uncomfortably above its long-term target. Inflation could remain elevated if energy prices stay near current levels. Energy prices, in turn, may remain elevated while the conflict with Iran persists. The conflict with Iran could last until the U.S. can extricate itself in a manner which allows it to claim some sort of victory, no matter how small.
You know, it might not be a bad idea for me to put that on laminated cards and pass them out when someone asks me about such things. Then again, that would be arrogant and nerdy at the same time. In other words, people repellent.
This, then, takes me to the last topic of primary interest for this past quarter. To be sure, there was, or is, a lot happening in our dynamic, massive economy. There is no possible way I can cover everything in the space allotted here.
The Scale of AI Investment
However, it seems everyone wants to know how long this “artificial intelligence (AI) buildout will continue.” Perhaps ‘everyone’ is a bit strong, but a lot of people seem to be interested in the subject.
Towards the end of the third quarter, some of the biggest firms in the finance industry published reports with eye-popping estimates regarding the “AI spending boom and infrastructure investment.”
- Goldman Sachs estimates that its baseline scenario could require approximately $7.6 trillion in AI infrastructure investment through 2031, with hyperscaler annual capex reaching $1.4 trillion by 2028.
- As AI clusters scale to hundreds of billions in annual spending, memory, networking, and power infrastructure have become the binding constraints, not GPUs.
- Constrained supply has already pushed memory producers’ gross margins to roughly 80%, more than double their historical average, signaling significant pricing power.
If those numbers aren’t enough, Bloomberg’s Marton Kasnyik had more thoughts about Goldman Sachs’ research on the matter. (8) Here are a couple of bullet points from the article which might pique your interest:
- A Goldman team led by Ryan Hammond estimates next year’s AI capex spending will exceed $1.1 trillion and forecast an additional $1.4 trillion in 2028.
- Hyperscaler capex spending now exceeds cash flow from operations, which means large further growth will increasingly need funding via debt and equity issuance according to Goldman’s Hammond.
- These same companies will need about $300 billion in annual AI revenue in coming years just to break even on their 2026-2027 investment, again, according to Goldman Sachs.
If history serves as a guide, and I am supposed to say it doesn’t necessarily, the rate of growth in this unprecedented infrastructure buildout may slow sooner than many currently believe. That does NOT mean it is going to collapse, at all. The economy will likely continue to invest enormous sums into these technologies.
However, the money has to come from somewhere, and it is much harder, in absolute terms, to grow, say, 30% on $1.4 trillion than it is on $1.4 million.
Compounding a $1.4 trillion annual investment base at 30% for five years would produce annual spending of roughly $5.2 trillion. That simple calculation illustrates why today’s growth rate becomes increasingly difficult to sustain as the investment base expands. It does not mean AI spending is poised to collapse, but it does suggest the pace of growth may eventually moderate.
Obviously, at a minimum, the industry will have to generate that amount of money in revenue in order to essentially break even, at a minimum. To turn a decent profit on the investment, what would they need? Let’s say $2 for every dollar in capex. That doesn’t seem too unreasonable when you consider all the expenses associated with running a company. Under that purely illustrative assumption, the resulting revenue figure would be approximately $10,396 billion in revenue.
Okay, now we are going to have some real fun with math.
At the end of the second quarter of 2026, the Bureau of Economic Analysis (BEA) estimated the nominal size of the U.S. economy was $32,486 billion. (9) IF we assume an inflation rate of 3% and a real growth rate of, say, 3%, the nominal economy will grow 6% per year, on average. Therefore, in Year 5, U.S. GDP could be roughly $43,474 billion.
As such, $10,396 billion divided by $43,474 billion equals 23.9%.
While nothing is impossible, it is difficult for me to imagine what we spend on AI being equal to 23.9% of our total economic output, all of it, in 5-years.
Therefore, this recent infrastructure investment could slow down somewhat when more people start putting pen to paper. While this proverbial “back of the envelope” is decidedly simplistic, and there is undoubtedly ‘far more to it’ than I allowed, it is intended only to illustrate the scale required to sustain that rate of growth.
Investors may begin to see this more clearly by the middle of 2027, but the pipeline in new construction/investment appears likely to remain substantial into 2028 based on currently announced investment plans. At least, that is what my crystal ball (and I do have one), suggests here at the end of the third quarter of 2026.
In the end, there is only so much one can do in less than 2,000 words. So, I tried to cover what I believe are the current topics of most interest. Should longer-term interest rates continue to climb as they did in September, I will write about how the global bond markets are flooding the world in debt. It likely won’t be a very fun read, and I apologize in advance.
Until next quarter, thank you for your continued interest and support.
Sources:
- AAA – National Average Climbs Nearly 5 Cents Since Last Week. September 24, 2026.
- CME Group – Crude Oil Futures and Options.
- The Wall Street Journal – Trump Rejects Iran Ceasefire, Expects Renewed Bombing After Midterms. September 25, 2026.
- Bureau of Labor Statistics – Consumer Price Index News Release. September 11, 2026.
- CME Group – FedWatch, Target Rate Probabilities for 28 Oct 2026 Fed Meeting. Accessed October 1, 2026.
- Gov – Press Release: Federal Reserve Issues FOMC Statement. September 16, 2026.
- 24/7 Wall Street – Goldman Sachs Says a $7.6 Trillion AI Spending Boom is Coming. Skip the GPUs, Follow the Money Here. September 26, 2026.
- Bloomberg – Goldman Sees Hyperscaler AI Capex Rising 50% to $1.2 Trillion.
- Bureau of Economic Analysis – GDP (Third Estimate), Industries, Corporate Profits, State GDP, and State Personal Income, 2nd Quarter 2026; State PCE, 2025. September 30, 2026.
Advisory services, including investment management and financial planning, are offered through Oakworth Asset Management LLC, an SEC-registered investment adviser and a wholly owned subsidiary of Oakworth Capital Bank, Member FDIC and Equal Housing Lender. Registration with the U.S. Securities and Exchange Commission does not imply a certain level of skill or training. Investment products and services offered through Oakworth Asset Management LLC are not deposits, are not FDIC insured, are not bank guaranteed, may lose value, and are not insured by any federal or state government agency. Due to the ownership relationship between Oakworth Capital Bank and Oakworth Asset Management LLC, conflicts of interest may exist to the extent that either entity recommends the services of the other. Additional information regarding Oakworth Asset Management LLC, including its services, fees, and conflicts of interest, is available in the firm’s Form ADV at www.adviserinfo.sec.gov.
This communication is provided for informational and educational purposes only and should not be construed as investment, legal, tax, accounting, or other professional advice. The views and opinions expressed are those of the author(s) as of the publication date and are subject to change without notice.
The information presented is general in nature and is not intended to address the specific objectives, financial situation, or particular needs of any individual investor, account, or client relationship.
Nothing contained herein constitutes a recommendation to buy, sell, or hold any security, investment product, or investment strategy, nor should any information presented be relied upon as the basis for making an investment decision. References to specific securities, companies, industries, sectors, market indices, economic conditions, political developments, or geopolitical events are provided solely for informational and illustrative purposes and should not be interpreted as investment recommendations, endorsements, or predictions of future investment results. Investment strategies discussed may not be suitable for all investors. Individuals should consult their own financial, legal, tax, accounting, and other professional advisers before making any financial decisions.
This communication contains general information that may not be suitable for every investor and may be based on assumptions, estimates, opinions, and forward-looking statements that are subject to change. Actual results may differ materially from those expressed or implied, and no assurance can be given that any expectations, forecasts, or opinions discussed herein will be realized.
Investing involves risk, including the possible loss of principal. Past performance is not indicative of future results and should not be relied upon as a predictor of future performance. There is no guarantee that any investment strategy will achieve its objectives or avoid losses.
Any references to market indices are provided for illustrative purposes only. It is not possible to invest directly in an index. Index performance reflects the reinvestment of dividends, where applicable, but does not reflect the deduction of fees, expenses, or transaction costs, which would reduce returns. It should not be assumed that any client account will achieve returns comparable to any index discussed.
Information contained herein may be obtained from third-party sources believed to be reliable; however, Oakworth Asset Management LLC and Oakworth Capital Bank do not warrant or guarantee the accuracy, completeness, or timeliness of such information. References or links to third-party content are provided solely as a convenience and do not constitute an endorsement of any third-party views, products, or services.
Oakworth Asset Management LLC does not provide legal or tax advice. Individuals should consult their attorney, accountant, or tax adviser regarding their specific circumstances.