Growth, Rates & the AI Economy
We entered the year embarking on a broadening growth phase. Growth was firming domestically and internationally, earnings were strong and broadening, inflation was cooling, and market leadership was expanding beyond a narrow group of names.
The conflict with Iran, disruption surrounding the Strait of Hormuz, and subsequent spike in Oil introduced a reignition of the nagging inflation impulse that markets can’t ignore. Interest rates have marched steadily higher, valuation multiples have compressed, and what had been a relatively clean transition toward stronger global growth has been thrown some pressures to contend with.
As the third quarter closed, the market remains near all-time highs, it’s been more of a sideways consolidation and rotation of markets since Summer. The underlying economy has shown surprisingly little evidence that the geopolitical shock materially altered its trajectory thus far.
The economy remains resilient, but increasingly uneven. Earnings remain quite strong, but interest rates are moving higher. The AI capital expenditure cycle continues to accelerate, while some of the more rate-sensitive areas of the economy are beginning to see the bite that comes with higher rates.
Beneath the headline indexes, market breadth has deteriorated considerably. That leaves us with perhaps the most important question heading toward 2027:
How long can the incredibly strong AI and capital expenditure cycle overpower everything else?
The Economy: Resilient, But Uneven
Perhaps the biggest surprise over the past several months has been what hasn’t happened. Despite higher Oil prices, higher interest rates, and considerable geopolitical uncertainty, the hard economic data has yet to show a meaningful broad-based slowdown.
Consumer spending remains particularly resilient. Recent research suggests real consumer spending is tracking at roughly 3.5%-4.0% annualized growth in the third quarter, supported in part by rising household net worth even as real income growth has begun to cool. There are signs that consumers have had to use part of their savings to push through and keep spending.
Manufacturing has also continued to improve and has firmly been in expansion all calendar year after years of weakness. At the same time, actual factory production has been less impressive than the surveys, reminding us that improving sentiment still needs to translate into hard economic activity.
And then there is capital spending. The AI buildout has become one of, if not THE defining economic force of this cycle. Hyperscaler investment is estimated at approximately $800 billion for 2026, up from just $71 billion in 2019, with estimates approaching $1.1+ trillion for 2027. Staggering.
That spending is increasingly extending beyond semiconductors and software into data centers, power generation, electrical equipment, infrastructure, industrials, energy, and materials.
It is an extraordinary investment cycle to say the least. It may also be masking weakness elsewhere.
The Consumer & Housing: Where Rates Are Biting
The consumer has been remarkably resilient throughout this cycle, but the picture beneath the surface remains bifurcated. Higher-income households continue to benefit from elevated financial assets and home equity, while lower-income consumers are considerably more exposed to higher prices for gasoline, food, rent, credit, and overall inflationary pressures.
Importantly, Americans today spend a much smaller portion of their budgets on energy than they did 15 or 20 years ago. That helps explain why the recent Oil rise has not created the degree of demand destruction we might have expected in prior cycles.
Housing, however, is telling a different story. Construction remains soft, inventories are climbing, purchase applications have fallen toward multi-year lows, and home-builder sentiment has declined to a four-year low. Housing turnover is approaching stall speed.
This is exactly where higher rates should bite first, and we have already had a weak few years of housing. Historical analysis suggests that each 1% increase in the two-year Treasury yield cools real residential investment a few percentage points, while automobile and household-durable spending also tend to weaken.
Yet the impact of higher rates on the likes of Technology equipment investment or the area most closely associated with today’s AI investment boom is considerably smaller.
That distinction brings us to the Fed.
The Fed’s Predicament
Inflation has moved back firmly above 3% with recent headline CPI at 3.4%, driven in meaningful part by higher energy & service prices. If you strip out food and energy it sits at 2.4%. Remember the Fed’s mandate of price stability targets inflation of 2%. At the same time, economic growth remains resilient, consumption is firm, and the AI investment cycle continues at an extraordinary pace. Services and materials prices have remained firm as well, contributing to the stubbornly sticky inflation.
Normally, the prescription is pretty straightforward: Raise interest rates → increase the cost of capital → Slow demand → which helps bring inflation back toward target.
But there is very little that is normal about this cycle.
A meaningful portion of today’s growth is being powered by an AI capital investment cycle that may be considerably less sensitive to interest rates than the rest of the economy.
The world’s largest technology companies are competing for computing capacity, data centers, power, semiconductors, and ultimately leadership in what they believe could become one of the defining technologies of the next several decades.
Does another half-point interest rate hike cool that spending? Probably not.
But that same half-point matters considerably more to a prospective homebuyer, an automobile purchase, a small business seeking financing, or a leveraged company refinancing its debt.
And therein lies the Fed’s predicament. We said the Fed was in a pickle a few years ago, It just became even more precarious. Not all of today’s inflation is demand-driven, there lies the difficulty. A portion of the recent increase is supply-side related, particularly through energy, geopolitical disruption, and commodity prices.
Monetary policy is a more of a blunt instrument. The Federal Reserve can make mortgages more expensive through increasing the Cost of Capital. So, it can quasi-slow automobile purchases and can increase corporate financing costs.
But it cannot produce another barrel of Oil. And it cannot easily cool the strategic urgency behind this historic AI buildout. Remember, we as consumers and many small businesses tend to make economic decisions. The AI race has partial decisions being made to win at all costs that are frankly not economic. And the biggest players have the deepest resources.
The risk is that higher rates increasingly pressure the portions of the economy already not benefiting from the AI CapEx boom, while having considerably less influence over the investment cycle keeping aggregate growth elevated.
Raise rates too little and inflation may remain sticky. Raise them too aggressively and the Fed risks weakening housing, consumer durables, smaller businesses, and leveraged companies without meaningfully slowing the source of the very growth it is attempting to restrain to help cool prices.
The pickle is real. And the pressure is building.
Earnings vs. Interest Rates
Fortunately, corporate earnings continue to provide an important offset.
Recently, 80% of S&P 1500 earnings estimates have been rising, a five-year high spanning large, mid, and small-cap companies. The breadth in earnings thus far is undeniable.
Consensus expectations remain strong, with approximately $361 of S&P 500 earnings in 2026 and $417 in 2027, implying another roughly 15% increase next year if those estimates prove accurate. That’s after growing 30+% this year! To put that in perspective, annual earnings growth has averaged approximately 7.5% over the past 25 years.
Today’s rate of change is nothing short of extraordinary. Estimates continue to rise, margins remain strong, and the earnings broadening we have discussed throughout the year is real.
So why has an extraordinary earnings year produced a much more ordinary year for stock prices? Interest rates.
The earnings story this year has been remarkable, with earnings growth approaching 30% by some measures. But much of that fundamental improvement has been offset by valuation compression. Put simply, companies are earning considerably more, but investors are willing to pay less for each Dollar of those earnings.
That helps explain why a roughly 30% earnings growth story has translated into low-double-digit equity returns on the indexes rather than anything approaching the growth in profits. The timing is notable as well. Equity multiples peaked around the same time interest rates bottomed last Fall. Since then, better earnings have been fighting against a steadily higher discount rate. And that inverse relationship has become extreme.
Going back 60 years, the relationship between the average stock and interest rate is now among the most negatively correlated in the available historical data. That’s saying something.
Or another words: Rates up, average stock down.
As we end the quarter, the market increasingly appears to need a reprieve in rates before broader equity participation can sustainably improve again. There is one notable exception: Energy. Energy is the only major sector positively correlated with bond yields, making it one of the few areas that may help mitigate a persistent higher-inflation, higher-rate backdrop.
That doesn’t mean earnings no longer matter. Quite the opposite. The earnings have been there all year and are expected to continue to be there. The question is whether the interest-rate environment will allow investors to pay more for them.
The Question Beneath the Surface
Market breadth has weakened drastically as the quarter has come to a close. The percentage of stocks trading above their 50-day moving average has fallen sharply as the 10-year Treasury yield has moved higher. That deterioration is particularly interesting because it is occurring alongside strong earnings.
The market may simply be reflecting the same bifurcation appearing in the economy. On one side sits an extraordinary AI investment cycle supporting Technology, infrastructure, power demand, industrial investment, earnings, and overall GDP. On the other side are housing, consumer durables, leveraged companies, smaller businesses, and other rate-sensitive portions of the economy beginning to feel the cumulative effect of higher borrowing costs.
So the difficult question becomes:
How long can the AI and CapEx cycle overpower everything else if rates stay higher for longer?
If AI investment remains strong while inflation and rates eventually moderate, the combination of strong earnings, better valuations, and improving breadth could create a powerful feedback loop higher.
But if rates remain elevated or continue moving higher, the pressure on portions of the economy outside the AI ecosystem is likely to build.
Eventually, something has to give.
Either inflation and interest rates retreat, allowing the broader economy and market breadth to improve, or higher rates persist long enough that weakness outside the AI ecosystem becomes more difficult for the CapEx boom to offset.
The Forest View
There is plenty to remain constructive about. Economic growth has proven remarkably resilient. The consumer continues to spend. Earnings and margins remain the strongest in years. The AI investment cycle remains powerful, and global markets continue to offer opportunities at generally lower valuations.
But the margin for error has narrowed. We are increasingly watching an economy operating at two different speeds: an extraordinary investment cycle powering one side while higher interest rates gradually apply pressure to another.
The question is no longer simply whether the economy is growing. It is whether one historic capital investment cycle can continue to carry an increasingly uneven economy and whether inflation and interest rates can moderate before that divergence becomes a larger problem. Nobody knows exactly where that line is.
Our objective is not to predict every turn. It is to follow the evidence, recognize when conditions change, and position portfolios accordingly.
For now, earnings remain strong and the expansion remains intact. But the market is making something increasingly clear: better earnings alone may no longer be enough.
As we head into the final quarter of 2026, a turn lower in inflation, Oil, and interest rates could give strong fundamentals room to breathe again. If that reprieve doesn’t come, the pressure beneath the surface is likely to continue building.
Something has to give.
Michael Harris
