The AI boom is bringing new risks to the financial markets as investors flock to tech stocks and executives pay steep premiums to buy AI technology they cannot build in-house, warned two top finance executives.
AI has become the “number one topic of conversation” for both investors and corporate executives, Matthew Danzig, managing director at investment bank Lazard, said on a panel discussion with Citadel Chief Risk Officer Joanna Welsh at Reuters Momentum AI 2025 conference in New York this week. Companies are scrambling to articulate an AI strategy, often acquiring capabilities or proprietary datasets to stay competitive, the tech banker said.
“Every company that’s a potential target is figuring out their AI angle,” he said. Valuations are being driven to historically high levels as investors bet on potential profits instead of current fundamentals, he said. “It’s markets willing to pay for the future.”
NVIDIA’S Earnings
The industry will need roughly $7 trillion in capital by 2030 to fund its growth – and that is just for data centers, according to McKinsey & Co. Investors, however, have largely brushed aside concerns about the increasing amount of leverage in the system and the lack of revenue to support all the debt needed to finance that growth.
Shares of chipmaker Nvidia, which rose sharply after the $4.5 trillion company on Wednesday reported record revenue and a 65% year-over-year surge in net income for its fiscal third quarter, took a negative turn on Thursday. The shares were down 2.2% at $182.46 in afternoon trading, dragging other tech stocks down as concerns over an AI bubble resurfaced.
Beneath the frenzy, structural vulnerabilities lurk and cracks are starting to show.
Welsh said Citadel, which has $71 billion in assets under management, is prepared for potential drawdowns at any given time. The hedge fund’s risk models show that modern markets amplify shocks.
“Markets are just faster,” she said. “These volatility spikes and pulses, they hit harder, they fade faster, they repeat more often.”
Risks Are “Starting to Converge”
Risks in credit markets are “starting to converge and stack” with the AI boom, Welsh said, pointing to a rise in high-quality corporate issuance of 30- and 40-year bonds on assets with around four-year depreciation cycles. That means companies are paying off debt for a long time after the asset has potentially become obsolete. A mismatch between the debt maturity and the depreciation cycle is also a marker of increased risk because it strains cash flow.
At the lower end of corporate credit, issuers and investors have “equal enthusiasm” for zero-coupon convertible bonds issued by less creditworthy tech firms, she said. Zero-coupon convertible bonds are considered higher risk investments that give investors stock if the company does well and priority payment with bondholders in bankruptcy, but they don’t pay a coupon.
“Zero-coupon converts are having a big issuance year, same as they did in 2001, the same as they did in 2021,” she said, citing previous market downturns. “And if you stack that on top of the amount of capital that has gone into illiquid things like private credit,” she continued, “you can see how there’s some portfolios where… a brush fire could be pretty healthy.”
(Reporting by Dawn Kopecki in New York; Editing by Franklin Paul and Matthew Lewis)
