Markets adjust to a new normal

Rob Fox

December 10, 2025

Key Points

• The degree of economic uncertainty related to trade policy is lower than it was a year ago.
• Fears of an AI-driven stock market bubble bursting and damaging the broader economy are overblown.
• We expect that reduced uncertainty, lower interest rates and the OBBBA’s stimulative effects will help support the economy in 2026.

Entering 2026, economic uncertainties are much reduced from a year ago. Although a more pugilistic U.S. trade policy was expected from the new administration, the size and scope of the April “Liberation Day” tariffs sent stocks plummeting—the S&P 500 Index bottomed out 27% below its February high. But after the initial market mayhem, the tariffs were first delayed, partially walked back, then various trade deals were cut, and by late June, the stock market recovered its losses.

The effective across-the-board tariff rate is now about 17% but based on tax collections, the actual average import tax paid is only about 10% (versus 3% in 2024). We expect that to drop even further as the reduced tariffs on China and imported food products take effect, and more bilateral agreements are finalized. Although tariffs have not completely gone away, the uncertainty factor surrounding them has faded into the background, at least for now.

The reduced market anxiety can be seen in historically low volatility metrics for equity, bond and currency markets, as well as in historically tight corporate credit spreads (suggesting easier credit conditions). At the same time, gold and cryptocurrencies, which were viewed as havens from U.S. policy risks, have pulled back from record highs. After plummeting in the first quarter, the dollar stabilized and has gained 4% in the fourth quarter. U.S. treasury yields have edged lower given strong investor demand. The much ballyhooed “sell America” trade appears to have fizzled out.

With tariffs fading from the fore, artificial intelligence has become the main focus of financial market prognostication. While AI was already very much on everyone’s radar a year ago, the massive capital investments in AI and related segments have exploded to unforeseen levels—almost $400 billion this year alone from just the top five AI players. The direct investments in AI and its related infrastructure, combined with the wealth effects from the surging stock market, have conservatively added 1% to GDP this year. We expect those structural forces to continue through at least the next 12 to 24 months.

Is the stock market in a bubble?

The fear is that the AI boom has lifted the entire stock market to unsustainable heights and that a big market correction would lead to a sharp pullback in consumer spending, maybe even causing a recession. That’s one possibility but not the most likely scenario. Corporate earnings remain extremely strong and aggregate corporate debt levels are historically low—most of the stock market gains in the second half of 2025 were attributable to continually improving earnings expectations, not irrational exuberance.

Today’s AI boom is often likened to the great railroad boom of the late 1800s that led to a few different economic downturns. But recent work done by Sankey Research points to a more apt comparison: the oil and gas shale boom between 2010 and 2015 with its roughly $1 trillion investment. In relative terms, both booms represented roughly 0.3% of total GDP, but the aggregate capital expenditure in the oil and gas sector ran well over 100% of free cash flows versus an average of about 60% for hyperscalers today.

Over the longer term of the next three to five years, AI will likely play out similarly to the shale boom—overproduction of a commoditized product, lower than expected earnings and disappointing industry profit margins. After peaking in 2014, oil stocks languished for years before prices and profitability returned. However, just as the overall economy benefited from the lower oil and gas prices, likewise we will see productivity gains from AI across the entire economy. (As we noted last quarter, there is extremely scant evidence that AI is taking any type of toll on the labor market yet). As long as most of the AI capital expenditures come from existing cash flow rather than debt, structural risk to the economy should be limited. But keep an eye on increasing debt levels, lengthening depreciation schedules and off-balance sheet commitments.

There are other reasons to believe the economy will continue to remain steady in 2026. With the one-time tariff effect fading by end of the first quarter, core inflation is likely to resume its downward trend and that will give cover to the Fed for additional rate cuts—current odds favor three 25-basis-point cuts by year-end 2026. Fed Chairman Powell’s term ends in May, and his replacement will almost certainly lobby for lower rates. Moreover, the Congressional Budget Office estimates the One Big Beautiful Bill Act, primarily via the accelerated depreciation provisions, will boost GDP growth by almost a full percentage point in 2026. We acknowledge the labor market has cooled from the post-COVID cycle and is now more in line with historic norms. But near 4% wage growth and sub 5% unemployment are well within the margin of safety for a growing economy in 2026.