Hard to Handle

The biggest technological changes have rarely arrived smoothly for investors. Railroads, electricity and the internet each reshaped the economy, and each came with stretches of overbuilding, sharp swings in prices and genuine fear about who would win and who would be left behind. AI appears to be the latest example, and the volatility of the past few weeks fits that pattern.

We would resist reacting to each headline and focus instead on where the risks are building, keeping a sound investment strategy in place so short-term swings do not force long-term mistakes. The next week will test that, with the Fed on Wednesday, most of the largest technology companies reporting earnings and the midterms next Tuesday.

The AI Scare
Much of the selling over the past three weeks traces back to the question of which AI business models might make money. The technology itself is not in doubt, and we think it will change how nearly every company works. But the largest companies are spending at a pace that assumes they will be the winners, and none of them has yet shown a clear path from that spending to an adequate return.

The stock market damage so far is modest. The Nasdaq is about 5% below its October 6 record and still up about 13% for the year¹. It feels worse than that because the selling has been concentrated in the names that many investors own.

Companies are funding more of the build-out with debt and new equity rather than cash flow, and that adds volatility. Lenders are underwriting that debt without a proven economic model, and new shares need existing capital to buy them. That supply is about to grow, with Anthropic filing for its IPO this week and OpenAI reportedly considering one as soon as next year.

The stress has shown up more clearly in credit. US corporate bond issuance reached $2.195 trillion through September, up 24.5% from a year earlier² and on the way to a record, with outsized investment-grade borrowing by the hyperscalers. Heavy supply and growing uncertainty about the payoff have pushed up the extra yield investors demand over Treasuries. High-yield spreads have widened to 338 basis points from 260 in late August and investment-grade spreads to 91 from 79, a bigger relative move than many US stock indexes¹.

And this reaches well past the biggest tech companies. As of Q2 2026, business investment in computers and software was about 5% of GDP and added roughly 0.75 percentage points to GDP growth from Q3 2025 through Q2 2026, about a third of the total¹. If that spending keeps growing next year, we estimate it could add a similar amount to GDP growth in 2027, depending on the pace. If it stalls, the economy loses that support. Add the wealth tied to these stocks and the power, construction and credit markets that serve them, and the whole economy would feel an AI slowdown.

Some investors think this is a bubble, and they may be right. But for now, the evidence points to a repricing of risk and financing rather than a collapse in demand.

Bonds Act Like a Hedge Again
The same doubts that widened credit spreads have recently helped Treasuries.

For most of the year, heavy government and corporate issuance and the inflation risk from the war with Iran helped push the 10-year yield from 4.19% in early January to 5.31% on October 5¹. A $1.99 trillion deficit in fiscal 2026, roughly 6% of GDP, added to the pressure³.

But the AI uncertainty described above has weighed on both growth expectations and credit, and investors have recently moved toward safety. On October 22 the 10-year yield fell 17 basis points to 4.96%, its largest one-day drop since August 2024, even with oil well above $100¹. Five-year breakeven inflation expectations, the market’s read on future inflation, did not move much¹, so the rally came from worries about growth rather than an improved outlook for prices.

For portfolios with heavy AI exposure through the index, longer-term Treasuries have acted like a hedge again. But that protection has limits, because there would be fiscal issues with any slowdown in growth.

A Meeting Pause
Whether that bond rally holds depends partly on the Fed on Wednesday. After September’s hike to a range of 3.75% to 4.00%, fed funds futures put the odds of a hike this week at only about 11% and of a 25-basis-point hike in December at about 82%⁴. Chairman Warsh and other officials have signaled no urgency to move at back-to-back meetings, so the base case is a pause now and a hike in December.

We think that is sensible. Pausing between hikes gives the Fed time to see whether prices outside energy keep rising, and a balanced labor market, with payrolls up 29,000 in September and unemployment at 4.2%¹, supports patience. It also protects the credibility the Fed earned in September without adding to the tightening markets are already doing.

Warsh said in September that he wanted tighter financial conditions, and he has them, led by wider credit spreads, higher Treasury rates and soft stock prices. Those do part of the Fed’s work for it, which means the funds rate may not need to rise as far as it otherwise would.

The risk is in the communication. In July we noted his line about investors learning to “play the ball, not the referee.”⁵ With no forward guidance, the market will read Wednesday’s statement and press conference as the message, and a hawkish tone could add to volatility in the days after the meeting. That could happen even if the policy itself is reasonable.

Inflation is Broadening Some
How far the Fed goes will come down to inflation, and the September numbers gave it reason to stay alert.

Energy is the obvious problem. Brent crude is up 88% this year to about $115, with gasoline back above $4 per gallon¹. The energy component of CPI is up 20.4% from a year ago, which explains much of the 3.7% headline rate¹.

Prices outside energy are also firming. Some of that is energy leaking through, with airfares up 22.4% from a year ago¹. But core CPI, which excludes food and energy, has risen at a 3.4% annualized pace over the past three months against 2.6% over the past year¹, so the trend is moving in the wrong direction. Goods prices have turned higher, helped by a 9.7% annual rise in computer prices¹ as AI demand pushes up the cost of memory. Services and the stickier parts of the index are firming as well. Food is the exception.

The slower-moving measures are calmer. Core PCE was up 3.0% in August and the Dallas Fed’s trimmed mean, which Warsh has said he prefers, was up 2.2%¹. That gap might be driving some of the debate inside the Fed. A central bank can usually wait out a one-time oil shock, but that is harder to justify when prices outside energy are rising too. And the longer oil stays near these levels, the more it feeds into the cost of everything else.

Still, we expect some technical relief next year. Even flat oil prices would pull energy out of the year-over-year numbers, memory and computer prices should rise more slowly as new supply arrives, and tariff increases are dropping out of the comparisons. As those roll over, inflation should ease into 2027 and give the Fed room to stop.

Housing Suffers
Higher inflation and interest rates have clearly hurt housing. The average 30-year mortgage rate hit 7.40% this month, up more than a full point from a year ago, and outside a short stretch in late 2023 it is the highest since 2000¹. Mortgage rates sit 2.36 percentage points above the 10-year Treasury against a 20-year pre-pandemic average of 1.76¹, so either that spread narrows and mortgage rates ease, or lenders are charging more for consumer risk.

Activity has gone from flat to falling. Existing home sales are down 2.2% from a year ago, while active listings are the highest for a September since 2019, so supply is back near pre-pandemic levels¹. Prices have stalled, with the national median price up just 1.3% versus last year¹. Many homeowners hold mortgages far below today’s rates, which keeps them from selling and limits the damage, but also keeps the market frozen.

Housing and stock portfolios are the two largest pieces of household wealth. If both soften at once, consumer spending will likely feel it.

That importance also means a weak housing market gets attention quickly from the Fed, the mortgage agencies and Washington. The unusually wide gap between mortgage rates and Treasuries gives policymakers room to lower borrowing costs through a form of quantitative easing without a Fed rate cut, and we expect affordability to get more focus once the midterms are behind us.

Earnings Still Positive
Against higher rates, weaker housing and nervous credit markets, earnings remain a positive offset. Third-quarter earnings for the S&P 500 are on track to grow 29.5%, the third straight quarter above 25%, and profit margins are near their highest since at least 2009⁶.

The outlook is solid as well. Consensus estimates call for earnings growth of 33.1% this year and another 16.0% in 2027, and companies are backing that up, with nearly twice as many issuing positive guidance as negative⁶.

Valuation is less stretched than the headlines suggest. The forward price-to-earnings ratio is 18.7, below both its five-year average of 19.8 and its ten-year average of 19.1⁶. Most of the hyperscalers report this week, and their spending plans matter as much for the AI debate as overall index earnings.

Brazil and US Elections
Outside the US, Brazil was a reminder of how quickly political expectations can change. After conservative candidate Flávio Bolsonaro led the first round earlier this month, markets priced a market-friendly win, and Brazilian stocks and the real rallied. Then left-leaning incumbent Lula won Sunday’s runoff for a fourth term. The real fell sharply, and Brazilian stocks fell about 20% in dollar terms the day after the vote⁶. There are still procedural questions around the vote, though the electoral court has certified the result.

Recent wins on the right elsewhere in the world made a conservative turn across emerging markets look more likely, and Brazil may still turn out to be the exception. But if more emerging economies stay with left-leaning governments, we think the policies that tend to come with it, like looser budgets and more state direction of the economy, would likely weigh on those markets’ returns.

As for the US, prediction markets continue to lean toward a Democratic sweep in next Tuesday’s elections. Even with a change in control, the policies that matter most for markets right now, the war with Iran and the regulation of AI, may not move much. Inflation, earnings and oil are likely to matter more than the result. And after Brazil, we would hold those odds loosely.

Sources
Fred 1
SIFMA 2
treasury.gov 3
CME Fed Watch 4
federalreserve.gov 5
Factset 6