Everything About AI
NextFin News - Two investment firms have filed with the US Securities and Exchange Commission to sell exchange-traded funds that would let retail investors bet on the risk that artificial-intelligence companies default on their debt, a product category that until now lived almost exclusively on institutional credit desks. The filings arrive as the AI trade endures one of its sharpest selloffs in years, with credit-default swaps on Oracle, CoreWeave and Meta at or near record spreads.
Tuttle Capital Management and Defiance ETFs have each lodged paperwork with the SEC for funds tied to credit-default swaps, or CDS, on individual AI-linked borrowers. Defiance's suite references CoreWeave, Oracle and Meta, alongside a broader AI Infrastructure Credit Hedge ETF. The instruments, normally used by banks and hedge funds to insure bond holdings against default, would also be wrapped into highly leveraged ETFs linked to baskets of companies exposed to the AI buildout, including chipmakers.
The move turns a distress signal into a tradable product. When the cost of insuring a company's debt spikes, that is usually a warning flashed inside the institutional credit market. Now the warning itself is being packaged for the brokerage-app crowd - and it is arriving precisely when the warning lights are brightest.
The Filing Wave: What Is Actually Being Sold
Tuttle Capital has filed for at least five credit-default swap ETFs targeting the technology and AI ecosystem: an Oracle fund, a Magnificent 7 fund, a Semiconductor fund, a Hyperscaler fund and a CoreWeave fund. The filings were made through ETF Opportunities Trust and were flagged on November 3, 2026. Defiance's suite, filed through Tidal Trust V, names the Defiance AI Infrastructure Credit Hedge ETF, a Defiance CRWV Credit Hedge ETF, a Defiance META Credit Hedge ETF and a Defiance ORCL Credit Hedge ETF, with an effective date of January 19, 2027.
The mechanics matter. A credit-default swap is a contract in which the protection buyer pays a periodic fee - quoted in basis points, or hundredths of a percentage point, of the notional amount - to the protection seller in exchange for compensation if the referenced borrower defaults. Under the standardized conventions adopted after the 2008 crisis, investment-grade names carry a fixed coupon of 100 basis points and high-yield names 500 basis points, with an upfront payment bridging the gap between the fixed coupon and the market's actual risk price. For a retail investor, that plumbing is invisible: the ETF share price is what moves.
Defiance's filing contemplates CDS positions with a notional amount of five to ten times the fund's net assets. In other words, a 1% move in the underlying credit spreads can translate into a 5% to 10% move in the fund. That is the same leverage architecture that has made single-stock leveraged ETFs a retail phenomenon - and the same architecture that regulators have repeatedly warned is unsuitable for investors who cannot monitor positions intraday. In its own SEC filing, Defiance said the proposed products are suitable only for knowledgeable investors who understand the risks and can watch their portfolios throughout the day.
The timing is not accidental. The filings land days after the worst equity session for the AI trade in years. On November 5, 2026, the tech-heavy Nasdaq Composite fell 5.4%, entering a correction - a 10% decline from a recent peak - and wiping out roughly $2.5 trillion in market value in a single day, a record. The Dow Jones Industrial Average dropped more than 1,800 points. Nvidia, the year's bellwether, fell 9.3% and finished the session down for the year. Heading into the final hour of trading, the Magnificent Seven had shed close to $900 billion in aggregate market value; Nvidia alone was worth about $374 billion less than the prior close, with Tesla down about $161 billion and Microsoft about $117 billion. The iShares Semiconductor ETF fell close to 11%, its worst day since March 2020, with every component down more than 6%.
"A spread level of 1,000 basis points or above is generally a sign of credit distress," said Gavan Nolan, executive director of fixed income pricing at S&P Global Market Intelligence.
The Credit Stress Behind the Product
The equity rout has a credit mirror image, and it is uglier. CoreWeave, the AI cloud-computing company whose debt-funded expansion made it a symbol of the buildout, has seen the cost of insuring its bonds surge into territory that credit markets treat as distress. Its five-year CDS spread has traded above 1,000 basis points - intraday readings reached as high as 1,194.5 basis points, according to ICE Data Services, pricing in more than a 60% probability of default within five years. Its 9% notes due 2031 have traded near 80 cents on the dollar, pushing the yield to maturity above 15%.
For a retail investor trying to grasp what those numbers mean: a five-year CDS at 1,194.5 basis points implies an annual protection cost of roughly 11.9% of the notional. To buy $10 million of five-year protection on CoreWeave, an investor would need to pay approximately $1.82 million upfront, followed by quarterly payments of $125,000, under the quoted contract structure. That is the institutional price of fear. The ETF version strips out the collateral calls, the upfront wire and the counterparty due diligence - and leaves the direction.
The stress has spread up the capital structure to borrowers with investment-grade ratings. Oracle's five-year CDS widened to a record 288 basis points in mid-October, up from 260 basis points a week earlier and from 144 basis points at the start of 2026, according to market data compiled by S&P Global Market Intelligence and LSEG. Meta's five-year CDS reached a record 108 basis points this month, and Nvidia's hit a record 101 basis points. SpaceX's five-year CDS has also climbed to record territory near 203 basis points. Single-name credit-default swaps on high-grade big-tech issuers were barely active a year ago but are now among the most-traded contracts outside the financial sector, according to the Depository Trust & Clearing Corp., with roughly $895 million in Alphabet-related CDS outstanding.
The financing math explains the anxiety. Hyperscaler borrowing could reach $400 billion in 2026, up from $165 billion in 2025, according to Morgan Stanley estimates. Industrywide AI-related spending is expected to exceed $3 trillion, much of it funded with debt. One analysis of the buildout puts AI-related borrowing at $250 billion in 2026, rising to $420 billion in 2027, with 35% of next year's capital spending expected to be debt-funded. Oracle's fiscal 2027 first-quarter operating cash flow of $23.1 billion fell short of capital expenditure of $28.5 billion, leaving negative free cash flow of $5.4 billion - the arithmetic of a company borrowing today to build capacity it must monetize tomorrow.
"You know there is going to be a lot of misallocation of capital," said Robert Dafforn, chief investment officer of distressed-debt specialist Polus Capital, pointing to the "very highly levered" borrowing inside the industry.
Why This Time the Retail Investor Can Short the Warning
The deeper story is not that AI debt is risky - speculative borrowers have always been risky. It is that the instrument for expressing that view has migrated from the over-the-counter derivatives desk to the ETF wrapper, and with it the audience has changed.
Credit-default swaps were at the center of the 2008 financial crisis. The ABX.HE indices, which tracked baskets of subprime mortgage securities via CDS, became the market's distress gauge: by March 2008, the BBB and BBB- sub-indices of every vintage were quoted at roughly 10% to 15% of par, meaning protection buyers were being paid 85% to 90% of notional upfront against the chance the underlying mortgages defaulted. Those contracts were traded by dealers, hedge funds and insurers - not by households. The 2008 crisis also produced the standardization that makes today's retail packaging possible: fixed coupons, upfront payments, auction settlement and formal determinations committees to decide whether a credit event occurred.
That standardization is what lets an ETF hold CDS exposure without turning every shareholder into a derivatives counterparty. But it also severs the retail investor from the mechanics that made CDS a hedge rather than a bet. A bank buying protection on a corporate bond it already owns is hedging. A retail investor buying a 2x long CDS ETF on a single name is taking a directional view on spreads - and on the daily reset that governs leveraged ETFs.
Here is the second-order problem most investors will not think about until it hits them. A CDS position profits when spreads widen. But leveraged ETFs reset their exposure daily to maintain a constant multiple of the day's return. In a market where spreads gap - as they did when CoreWeave's CDS jumped 16% in a week and 164% over a year - the daily reset compounds against the holder during volatile, mean-reverting stretches. The investor can be right about the destination, credit stress, and still lose money on the path. That is the familiar trap of single-stock leveraged ETFs, now applied to credit derivatives on companies whose debt is funded by the same capex cycle the investor is betting against.
There is a second asymmetry worth naming. The buyers of these products are betting that credit spreads will widen from already-distressed levels. But the sellers on the other side - the counterparties writing the protection inside the fund - are being paid record premiums to take that risk. When the entire market has repriced, the marginal trade is no longer "smart money versus the crowd." It is the crowd paying a premium to be told what the crowd already knows.
The counter-thesis is real, and it deserves weight. Higher CDS prices reflect the cost of obtaining credit protection under prevailing market conditions; they do not, by themselves, prove that defaults are imminent. No one questions whether demand for CoreWeave's services is real - the company has contracts with all the major hyperscalers. The question is whether the cash flows will ultimately arrive. And for Oracle, an investment-grade issuer, the bear case requires a downgrade from BBB status, which would force sales by investors restricted to investment-grade holdings - a technical trigger, not a fundamental collapse. A BBB downgrade would also risk pressing IG-only investors into forced sales, a dynamic flagged by strategists tracking the name.
Even so, the appearance of retail-facing CDS products at the moment credit spreads hit records is itself a signal. Product wrappers tend to arrive after the smart money has already positioned. By the time a trade is simple enough to fit in a brokerage app, the asymmetry that made it attractive has usually narrowed.
What Comes Next: Three Horizons
In the short term, the products are filings, not funds. The Defiance series carries a January 19, 2027 effective date, and SEC approval is a separate question. The Tuttle filings alone, however, signal that issuer appetite for AI-themed credit instruments has moved beyond equities and options.
Over the medium term, the base case is that these funds launch into a market where the credit repricing is already partly done. CoreWeave's spreads above 1,000 basis points and Oracle's near-record 288 basis points mean the distress trade is crowded; a stabilization in AI capex plans or a pause in rate pressures could compress spreads and squeeze the very investors who bought the warning. The upside case is a genuine credit event - a missed interest payment or a downgrade - that sends spreads wider and validates the products. The downside case is that the equity correction proves cyclical, the debt markets calm, and the CDS ETFs decay on daily resets while the underlying AI buildout continues.
The falsifying signal is concrete: if Oracle is downgraded from BBB and its five-year CDS holds above 300 basis points while CoreWeave's spreads remain elevated for two consecutive months, the view that this is a crowded, late-arriving trade is wrong - the credit deterioration is structural and the retail products arrived on time. If instead CoreWeave's five-year CDS falls back below 600 basis points and Oracle's retreats toward 150 basis points as AI capex guidance stabilizes, the structural-repricing thesis fails and the products become a lesson in timing.
Longer term, the structural shift is harder to reverse. Hyperscalers have moved from equity-funded growth to debt-funded infrastructure, and the CDS market on big-tech names has gone from dormant to liquid. That plumbing is not going back. What may revert is the price of the risk - and the investors who buy the packaged version of it will learn, as leveraged-ETF holders have before, that owning the gauge is not the same as understanding the engine.
The irony is sharp enough to state plainly: the products that let retail investors bet against AI debt arrive precisely because the market has already decided that debt is where the AI trade breaks. The warning became the product - and by the time it was on the shelf, the warning had already been delivered.
Explore more exclusive insights at nextfin.ai.
