AI risk is dominating conference calls as investors dump stocks

Core Insights - The current quarter is witnessing significant corporate earnings growth, yet the focus is shifting towards the potential threat posed by artificial intelligence (AI) [1][3] - Mentions of AI disruption in management calls have nearly doubled compared to the previous quarter, indicating rising investor concern [1] - Despite strong earnings growth, the S&P 500 has remained stagnant due to fears surrounding AI's impact on future earnings [4] Earnings Performance - Fourth-quarter earnings for S&P 500 companies are up 12% year-over-year, surpassing the initial expectation of 8.4% [3] - Over 75% of companies have reported positive earnings surprises, which is above average [3] Market Reactions - CBRE Group Inc. experienced a 20% stock selloff after its CEO suggested AI could reduce long-term demand for office space [2] - Stocks perceived to be at risk from AI have seen significant declines, with UBS Group AG reporting a 40% to 50% drop in affected stock baskets over the past year [7] Sector Impact - Media, software, and staffing sectors are identified as the most vulnerable to AI disruption, with financial and professional services also being affected recently [5] - In contrast, companies like Taiwan Semiconductor Manufacturing Co. and SK Hynix Inc. are benefiting from AI-related demand, contributing to record highs in Asian markets [6] Short Selling Trends - Short interest in stocks at risk from AI has increased, with the percentage of shares out on loan rising from about 2% to over 5% in the UBS basket [11] - Stocks such as Randstad NV and Ubisoft Entertainment SA are among those with heightened short interest [11] Capital Expenditure Trends - Despite concerns about AI disruption, capital spending by major tech companies (Amazon, Alphabet, Meta, Microsoft, Oracle) surged by 72% in 2025 and is projected to increase by another 63% this year [12] - A potential catalyst for easing market fears would be a reduction in capital spending announcements from these hyperscalers [13]