Investors and companies are increasingly seeking ways to hedge risks associated with the rapidly expanding artificial intelligence (AI) sector, driving strong demand for derivatives linked to AI infrastructure and computing power. Data from Cboe Global Markets shows that put options tied to CoreWeave, a key player in AI-focused cloud computing, surged to record levels in October and have remained elevated. Some of the most active contracts involve put options expiring in 2028, which would pay out if CoreWeave shares fall to $35, compared with their recent trading range near $55.
At the same time, credit-default swap (CDS) spreads—the cost to insure against debt default—widened markedly for major hyperscalers including Amazon, Alphabet, Microsoft, Meta, and Oracle. Bond data provider Solve reported that these spreads reached their highest levels since 2021 over the past month, signaling broad market concern about credit risks in the sector rather than isolated company-specific issues. Eugene Grinberg, co-founder of Solve, noted the consistency of the widening across these firms, highlighting especially sharp increases for Oracle, where spreads rose by more than one percentage point.
The growth of AI infrastructure has also prompted firms to devise new hedging mechanisms. Quantitative trading firm Jump Trading recently established a dedicated desk for trading computing power, capitalizing on persistent price inefficiencies among providers. The firm develops over-the-counter derivatives to help market participants hedge against significant price declines in computing resources.
Dave Olsen, president of Jump Trading, described how company users of AI infrastructure might purchase put options to guard against drops in rental rates for Nvidia chips, which are widely used in AI workloads. For example, an AI service provider renting Nvidia chips could buy a put option that grants the right to sell computing power at a floor price by a certain date, insulating the business from sharp revenue declines if demand softens. Jump Trading typically takes the other side of these trades and may also arrange cash-settled contracts or sublease computing power to third parties.
Rental prices for Nvidia B300 chips have surged about 56% this year to $7.56 per hour, according to the Silicon Data index. Jump Trading might agree to purchase these chips if the price dropped to levels near $4 per hour. The firm is carefully assessing counterparty creditworthiness and default risks before entering into contracts.
In addition to Jump, other established trading firms are exploring new derivatives linked to computing power. Susquehanna International Group has received inquiries from data centers and privately held “neocloud” operators seeking hedges against semiconductor price volatility, according to Nick Volbi, head of institutional prediction markets at Susquehanna. Put options on indexes representing computing power costs can provide protection against price declines, possibly offering collateral advantages for borrowers seeking financing. “If I buy this form of insurance, can I secure better financing terms?” Volbi said.
There is also nascent activity in prediction markets focused on potential regulatory risks to data-center growth. Kalshi offers contracts on whether states such as New Hampshire, Michigan, or Maine will impose moratoriums on new data-center developments before 2028, although trading volume remains limited.
Despite the rise in insurance-oriented trades, many participants remain bullish on AI’s prospects. Olsen noted that hedging is viewed as prudent risk management rather than an indication of bearish sentiment. “You don’t need to be bearish to want some insurance,” he said.
Market conditions reflect some of this caution. The Nasdaq Composite index fell 2.9% on Friday, marking its worst monthly performance since March 2025, though the Nasdaq-100 showed slight recovery Sunday evening. Investors enter a new month with U.S. elections approaching and upcoming earnings reports from major tech companies including Apple, AMD, Qualcomm, and Super Micro Computer.
Portfolio manager Solaka, who oversees accounts concentrated in large technology stocks, observed a growing interest in protective strategies from investors—including many former industry employees—seeking to hedge their substantial gains. Some face trading restrictions tied to their companies, increasing demand for options and credit-default swaps as a form of risk mitigation. “These people have made so much money, they’re quitting to hedge their stock,” Solaka said, underscoring the evolving strategy among tech insiders and broader market participants amid AI’s rapid expansion.