The market has spent the past month doing something it probably should not have been able to do this comfortably. Bond yields have risen, oil has stayed expensive, mortgage rates have pushed higher and the Fed has kept another hike in play, yet equities have continued to hover around record territory because earnings have been carrying enough weight to keep the whole structure upright.
That makes next week less an earnings season than a stress test.
Takeaways by Dark Side of the Boom™
-
Earnings have been the market’s shock absorber. Bond yields rose, oil stayed expensive and the Fed kept the tightening threat alive, yet equities held up because profits kept arriving with enough force to offset the macro drag.
-
The problem is that expectations are no longer modest. Deutsche Bank’s Bankim Chadha sees Q3 earnings growth around 34%, above consensus near 27%, which means the market is already leaning into another very strong quarter.
-
AI is doing an enormous amount of the lifting. Goldman Sachs’ Ben Snider sees AI infrastructure beneficiaries driving more than half of S&P 500 EPS growth alongside 116% hyperscaler capex growth.
-
The headline index may be telling a cleaner story than the average company. Societe Generale warns that concentration, base effects and elevated long-term growth assumptions can flatter the aggregate picture.
-
The consumer is still the macro contradiction. Confidence is poor and affordability pressure is real, but spending has held up because the households doing the most spending are not necessarily the households feeling the most pain.
The Last Shock Absorber Gets Tested
The market has spent the past month doing something it probably should not have been able to do this comfortably. Bond yields have risen, oil has stayed expensive, mortgage rates have pushed higher, and the Fed has kept another hike in play, yet equities have continued to hover around record territory because earnings have been carrying enough weight to keep the whole structure upright.
The question is not whether companies can beat estimates. They usually do. The real question is whether the earnings machine can keep running hot enough to absorb everything else the macro backdrop is throwing at it.
Second-quarter S&P 500 earnings grew around 33% year on year, a pace that normally belongs to the rebound phase after an outright earnings recession rather than a mature expansion. The rolling 12-month picture makes the same point: profit growth is not merely healthy, it is running at a speed that historically does not last forever.
The quarter-on-quarter comparison is even more striking because it shows just how unusual the current burst has become.
Deutsche Bank’s Bankim Chadha thinks the momentum can persist, with Q3 earnings growth around 34% versus consensus closer to 27%. That leaves room for another above-average beat season, but it also raises the hurdle. When investors are already expecting a very strong quarter, good numbers become maintenance rather than surprise.
Goldman Sachs’ Ben Snider expects AI infrastructure beneficiaries to account for more than half of S&P 500 EPS growth this quarter, alongside hyperscaler capex growth of 116%. That is an extraordinary amount of index earnings power being generated by one investment theme, and it changes how the headline number should be read.
The S&P can look like the whole corporate sector is booming when in reality a relatively small group of companies is doing much of the heavy lifting.
BCA Research’s sector work makes that concentration visible.
That does not make the earnings boom fake. It makes it narrower than the aggregate number suggests.
There is also a second tailwind inside the profit story that deserves more attention: energy. Higher oil prices lift producer earnings quickly, while the drag on everyone else arrives more slowly through transport costs, input prices, household budgets and margins.
The oil company gets the windfall now. The rest of the economy pays the invoice later.
So the aggregate earnings number can look stronger even while the underlying economy is beginning to absorb a higher energy tax.
AI has been even more important in delaying the kind of profits recession that might otherwise have followed the Fed tightening cycle. The arrival of ChatGPT and the capex boom that followed gave corporate earnings another engine just as monetary policy was trying to take one away.
That is why this cycle has looked different.
But the danger is that the market has started treating exceptional growth as normal.
Societe Generale’s equity strategy team is cautious on that point. Market-cap weighting can distort the earnings signal because the fastest-growing companies become larger pieces of the index, which means their earnings growth gets a bigger weight in the total. Strong performers therefore make the aggregate look stronger simply by becoming more important.
Base effects can also exaggerate the picture after periods of weakness.
And then there is the longer-term problem. Societe Generale argues that high long-term EPS growth expectations have historically been a warning sign rather than a comfort, with stronger long-term assumptions often followed by weaker subsequent returns.
Markets are happiest when expectations are low enough to beat.
They become much less forgiving when perfection is already in the price.
For much of the twentieth century, profit margins behaved like a classic mean-reverting series. Labour took some, capital took some, competition did its work and margins oscillated.
Margins have kept grinding higher, and on a broader national-income basis corporate profits are now taking an historically large share of the economy.
That matters because the equity market is effectively asking investors to believe not only that earnings stay strong, but that historically elevated margins can remain elevated while wages, energy, interest costs and politics all push from the other side.
But that is a much harder bet than simply saying earnings are good.
The consumer makes the picture even more complicated.
Affordability is becoming one of the loudest political and economic issues heading into the midterms. Diesel and gasoline prices are sharply higher, the 30-year mortgage rate has climbed to 7.38% and one-year inflation expectations in the New York Fed survey have risen to 3.9%.
Confidence is ugly too, with the Conference Board measure down at a 12-year low.
Retail sales recently posted their strongest gain since March and remain above the level at the start of the year.
The consumer says things feel terrible, but the cash register is still ringing.
The explanation is that the household sector is not one consumer. Oxford Economics’ Michael Pearce points out that around 40% of spending is done by the top 20% of households, while confidence surveys are population-weighted. Higher-income households can therefore keep aggregate spending strong even while a much larger share of people feels squeezed.
Bank of America’s Shruti Mishra adds another layer. The bank’s proprietary data suggest spending growth has become less K-shaped over recent months, with the gap between higher- and lower-income households narrowing.
That narrowing matters, but it may prove fragile if energy stays expensive.
Consumer stocks are telling a similarly mixed story. Discretionary names have recovered much of their early-year weakness, while staples are still up for the year.
The market is not pricing a consumer collapse.
But neither is the household picture clean enough to dismiss the affordability problem.
Atlanta Fed data show wage growth for the bottom quartile running below the top quartile, an unusual pattern that has persisted for close to two years.
That is where the consumer paradox becomes more fragile.
Spending can keep growing while confidence stays weak, but if lower-income wage growth remains soft and energy costs continue to rise, the cushion starts getting thinner precisely where households have the least room to absorb another hit.
That is why this earnings season matters more than the usual beat-rate scoreboard.
The market is not simply asking whether companies can clear estimates. It is asking whether the profit machine is strong enough to keep carrying high rates, expensive energy, stretched margins and a consumer that is still spending despite telling anyone who asks that life feels worse.
If Deutsche Bank is right and earnings growth holds around 34%, the market gets another extension on the current bargain.
If Goldman is right about the concentration, the headline number may still look spectacular while the underlying dependence on AI becomes even more obvious.
And if Societe Generale is right that long-term expectations have become too optimistic, then strong earnings could still arrive without producing the kind of equity response investors have grown used to.
The market does not need bad earnings to struggle.
It may only need earnings that are less extraordinary than the price already assumes.