Bank of America chief investment strategist Michael Hartnett said soaring global bond yields are now one of the biggest threats to the AI capital expenditure boom, and he warned that the upcoming U.S. midterm election could become the trigger that sets off a broader market move.
In the latest edition of his Flow Show weekly note, Hartnett wrote that if Democrats sweep both chambers of Congress, U.S. stocks could fall by more than 10%, the dollar could weaken, bond yields could move lower, and the AI bubble could face a rupture. He called a Democratic sweep one of the market’s biggest tail risks and said investors have barely priced it in.
Polymarket data cited in the report showed the probability of Democrats winning both chambers had risen to 50%, far above the 10% probability assigned to a Republican sweep. Trump’s approval rating was put in a 35% to 40% range, well below the historical average of 53% seen two months before midterm elections, a political backdrop Hartnett said reinforces his warning.
Bonds, not jobs data, drew the sharper warning
Hartnett argued that last week’s most important market development was not the stronger-than-expected jobs report but the broad breakdown across global bond markets.
The U.S. 10-year Treasury yield climbed to 4.81%, close to levels seen during the 2008 financial crisis. The 30-year Treasury yield rose to 5.31%, its highest since 2007. In Japan, the 10-year government bond yield moved above 3.0% for the first time since 1996, while the 30-year yield reached 4.2%, about four times the Bank of Japan policy rate.
Europe also showed stress. Germany’s 10-year bund yield rose to 3.38%, the highest since 2011. The France-Germany spread widened to 88 basis points, and the Italy-Germany spread reached 84 basis points, both back to levels associated with the 2012 euro-area debt crisis. The Bloomberg Global Bond Yield Index climbed to its highest reading since 2007 and sat just 1 percentage point below the highest level of this century.
Hartnett summed that up with a simple line: "bonds lead bubbles." In his view, long-end yields, not stock-market storytelling, are the real anchor for the AI trade. Until global 30-year yields move back below 5%, builders and financiers of AI infrastructure are likely to keep lagging the companies applying AI technology.
The midterm election as an underpriced market variable
Hartnett said the midterm election is not a regime-change event on the scale of Thatcher or Reagan in 1980, or Brexit and Trump’s 2016 victory. It is unlikely, he argued, to alter the upward path of U.S. government spending in any fundamental way. Even so, he said the structure of the result still matters for asset prices.
Bank of America’s August fund manager survey showed 47% of respondents expected a split result with Republicans controlling the Senate and Democrats controlling the House. Another 23% expected a Democratic sweep, while only 9% expected Republicans to retain control of both chambers. Republicans currently lead 53 to 47 in the Senate and 218 to 212 in the House.
For Democrats to complete a sweep, Hartnett said they would need to win at least four of six vulnerable Republican-held Senate seats: North Carolina, with a 92% flip probability; Maine, 69%; Alaska, 64%; Ohio, 55%; Texas, 51%; and Iowa, 37%. They would also need to defend their own vulnerable seats in Georgia, New Hampshire and Michigan, with hold probabilities of 94%, 84% and 65%, respectively. Hartnett singled out Ohio, Texas, Iowa and Michigan as the battleground states investors should watch most closely.
Wall Street is also watching the Texas governor’s race. Incumbent Republican Governor Abbott was shown with 49% support, against 45% for Democratic challenger Hinojosa. Hartnett’s note said the race is being treated as an important signal for the policy direction of AI data-center expansion. Abbott has recently announced a pause order on data-center construction in an effort to stem polling slippage, according to the report.
How Hartnett maps the market impact of a Democratic sweep
Hartnett laid out a direct transmission mechanism for markets if Democrats take both chambers.
He said a shift from what he called "populist capitalism" to "populist socialism" would mean taxes and regulation move higher rather than lower, hurting corporate earnings. He also tied that scenario to policies aimed at lowering inflation, improving access to healthcare and easing K-shaped wealth inequality, which he said would hit the AI capital-spending boom and the "too big to fail" Wall Street structure directly. He added that a loss of Trump’s political capital would weaken his ability to act on priority issues including AI, resource concentration and diplomatic pressure.
Based on that framework, Hartnett’s preferred hedge in a Democratic sweep scenario is to short financial stocks and the dollar. He said U.S. equities would fall more than 10%, the dollar would weaken and bond yields would decline, while international equities would outperform, with Europe ahead of Asia.
If Republicans unexpectedly hold both chambers, Hartnett said that would signal a broad revival in risk appetite, give the AI bubble a green light and revive the narrative of U.S. dollar exceptionalism. The highest-probability outcome, in his view, remains a split government with Republicans controlling the Senate and Democrats the House, a setup he described as a moderate risk-on environment where "gridlock is Goldilocks."
Commodities, gold and a warning on crowded AI positions
Within that macro framework, Hartnett kept his longer-cycle core allocation unchanged: long commodities and gold as hedges against inflation and geopolitical risk. He said governments’ willingness to do "whatever it takes" on the fiscal side is suppressing long-end yields and supporting nominal GDP strength, leaving intact his strategic view that "any asset is better than bonds."
His warning on AI was even more pointed. Hartnett said hyperscale cloud providers have already seen free cash flow turn negative under the weight of capital spending commitments, and he wrote that the AI bubble could burst "at any time." His post-bubble framework is "long humiliation, short arrogance," which in practical terms means rotating toward long-duration bonds and defensive sectors such as consumer staples, mining and materials, and healthcare, while avoiding crowded positions tied to AI infrastructure buildout.
Hartnett also pointed to a longer-term reversal signal. Over the past 10 years, rolling returns for U.S. stocks were 15%, commodities returned 11%, and U.S. Treasuries posted -2%, the worst record in nearly a century. Historical data, he wrote, show that when long-run bond returns turn negative, it has often marked favorable buying points for stocks in 1939, 1974 and 2009, and for commodities in 1933 and 2018. That history underpins his tactical bullish view on bonds in the fourth quarter.