According to a Bank of America survey, investors' allocation exposure to global stocks reached the highest level in nearly five years, and most respondents predicted that the economy would continue to grow. The survey was conducted from August 7 to August 13. A total of 180 respondents were interviewed, and the total assets under management were US$525 billion. Net 56% of respondents are currently overvalued by global stocks, the highest since November 2021. Strategist Michael Hartnett wrote in the research report that a record 56% of respondents predicted that the economy would “not land”; 72% believed that the Federal Reserve would not raise interest rates before the US midterm elections. The most crowded deals: going long on global semiconductors, shorting yen, and going long on “magnificentseven.” Major tail risks: AI bubble, disorderly rise in bond yields, second wave of inflation, geopolitical conflict. About 38% of respondents believe that AI hyperscale enterprise capital expenditure is most likely to trigger a systemic credit incident. Even so, 71% of respondents believe AI capital spending will not be reduced in 2026; 58% say AI will not impact the labor market until 2028 at the earliest.

According to a Bank of America survey, investors' allocation exposure to global stocks reached the highest level in nearly five years, and most respondents predicted that the economy would continue to grow. The survey was conducted from August 7 to August 13. A total of 180 respondents were interviewed, and the total assets under management were US$525 billion. Net 56% of respondents are currently overvalued by global stocks, the highest since November 2021. Strategist Michael Hartnett wrote in the research report that a record 56% of respondents predicted that the economy would “not land”; 72% believed that the Federal Reserve would not raise interest rates before the US midterm elections. The most crowded deals: going long on global semiconductors, shorting yen, and going long on “magnificentseven.” Major tail risks: AI bubble, disorderly rise in bond yields, second wave of inflation, geopolitical conflict. About 38% of respondents believe that AI hyperscale enterprise capital expenditure is most likely to trigger a systemic credit incident. Even so, 71% of respondents believe AI capital spending will not be reduced in 2026; 58% say AI will not impact the labor market until 2028 at the earliest.