The free lunch of AI is over?

Source: Wall Street Insight

Over the past two years, the market has assumed that someone will always foot the bill for AI infrastructure. That assumption was rejected by the credit market this week.

On July 28, South Korea's KOSPI fell 10.84%, triggering its 8th circuit breaker of the year. Nvidia fell 5%, SK Hynix fell 7.47%, and Japan's Kioxia fell 18%.

On the same day, Apple rose 1.2%, reclaiming the title of the world's most valuable company. The Dow Jones rose 0.51%, and all three major European indices closed higher.

On the same day, within the same news cycle, gains and losses were completely opposite. The market wasn't panicking—it was voting. The assets being sold off were companies deeply tied to AI infrastructure, while the assets being bought were those not reliant on the AI infrastructure narrative. The trigger for this vote wasn't in the stock market; it was in the bond market.

Who Is Saying No

On Monday, Nvidia's five-year CDS surged 14 basis points to 82 basis points in a single day, the largest increase for the contract. Oracle's CDS rose to 215 basis points, Alphabet's CDS rose to 67 basis points, and Broadcom, Amazon, and Meta all hit new records. A new bond issued by SpaceX in June fell below its issue price on the secondary market.

The pricing logic for CDS is completely different from stocks. Stock pricing looks at growth potential, narrative, and the future. CDS buyers are pension funds, insurance companies, and hedge funds. They don't read stories; they only read cash flow—whether the cash you are generating now can cover the principal and interest when they come due.

For the past two years, the pricing power for AI assets has been entirely in the hands of the equity market. The assumption of equity investors has been: Capital expenditure equals a moat. Whoever has the largest capex wins. Every GPU is a down payment on future market cap. Under this narrative, the combined capital expenditure of the five big tech giants for 2026 is projected to be $737 billion, compared to $261 billion two years ago.

Buyers in the credit market have remained silent because they don't trade on "stories." They were just waiting.

The direct trigger for Monday's trading action was Nvidia's July 25 announcement of a $500 billion AI infrastructure partnership with SK Group, along with a $250 billion financing guarantee for OpenAI and a $350 billion chip procurement financing arrangement. The potential exposure from OpenAI alone exceeds $600 billion, more than Thailand's entire annual GDP. The market spent the weekend digesting the implications of this number—come Monday's open, CDS spiked directly.

Prior to this, signals had been accumulating for three weeks. On July 9, S&P downgraded Oracle from BBB to BBB-, just one notch above junk status. Last week, Alphabet's free cash flow turned negative by $5.855 billion in its earnings report, the first time this has happened in its twenty years as a public company. When these events occurred in isolation, the market chose to digest rather than panic—Oracle's issue could be attributed to overly concentrated clients, Alphabet's could be seen as still being in an investment phase. Each was treated as its own separate problem.

Nvidia's $600 billion revolving financing scheme connected the dots into a line. Oracle's biggest client is OpenAI, OpenAI is the core beneficiary of Nvidia's revolving financing scheme, and Alphabet, Meta, and Amazon all operate on the same "borrow money to build infrastructure" financing model. These are not a collection of independent events. This is the same balance sheet, being pledged as collateral repeatedly under different borrowers' names.

The credit market took three weeks to digest the signals, establish the connections, and make a judgment. On Monday, the judgment was delivered.

How Credit Entered the Loop

The traditional model for the semiconductor industry is linear: Nvidia sells chips to cloud providers, cloud providers rent computing power to AI companies, capital flows from end-users layer by layer upstream, and each layer bears its own credit risk.

Nvidia's current model transforms this into a circular one. Nvidia provides credit to clients → Clients use this credit to purchase chips → Chip sales confirm Nvidia's revenue growth → Revenue growth supports its rating and valuation → A stronger balance sheet supports larger guarantees → The cycle continues.

Within this closed loop, part of Nvidia's revenue growth does not originate from end-user demand but from the credit it injects itself. Sal Naro, Chief Investment Officer at Coherence Credit Strategies, describes this type of arrangement as "financial alchemy." John Flood, the head of the trading desk at Goldman Sachs, wrote in a client memo that multiple traders felt "uncomfortable" with this, pointing out that the credit market has become a key window for judging the AI narrative.

This is not a problem unique to Nvidia. The entire financing model for AI infrastructure rests on one assumption—that the cost of capital is near zero, and the market will always be willing to pay for the "future." When CDS starts pricing, this assumption faces its first real test.

Chain Reaction

The credit market's skepticism doesn't need to verify each company one by one. It propagates down the financing chain.

Oracle was the first to hit the tripwire. On July 9, S&P downgraded its rating from BBB to BBB-, citing three clear reasons: AI business cash burn exceeding expectations (FY2026 capex up 162% YoY to $56 billion, free cash flow negative at -$23.7 billion, projected to rise further to $95 billion in FY2027); excessive client concentration (OpenAI accounts for about half of its remaining performance obligations); and insufficient internal load to absorb excess computing capacity. Oracle's CDS has risen from 144 basis points at the start of the year to 215 basis points. The cost of insuring the debt of a company with $67.4 billion in annual revenue and a 77% surge in cloud infrastructure revenue has doubled.

This risk points directly back to OpenAI itself. According to Fortune and the Financial Times, in 2025, OpenAI had revenues of $13.07 billion, an operating loss of $21 billion, and a net loss of approximately $39 billion after accounting for one-time expenses. A company losing nearly $40 billion a year is Oracle's largest client and the core beneficiary of Nvidia's revolving financing scheme.

The chain continues. Alphabet's free cash flow turned negative. Meta's $12.55 billion bond for its Texas data center carried an interest rate of 7.534%—a coupon typically seen only in the high-yield bond territory—and the subscription multiple was only 1.6 times, less than half the average for US corporate bonds this year. The average borrowing cost for AA/A-rated tech companies rose from 3.9% a year ago to 5.1%, while BBB-rated companies saw their costs jump from 4.7% to 7.0%. For lower-rated, AI-related companies, unsecured loan rates have already touched 10.5%-12%. Meanwhile, the capital expenditure guidance for the five big tech giants continues to be revised upwards.

This isn't a problem with one company. It's the process of the entire financing model switching from "infinitely cheap" to "market pricing." Every widening of spreads at one level transmits to the next borrower, the next market.

The Bill from 2000

This script isn't new.

In March 2000, Cisco became the world's largest company with a market cap of $555 billion. Its growth engine was called "vendor financing"—lending money to unprofitable internet companies to buy Cisco equipment. At its peak, the exposure was about $4 billion, covering over 700 clients. Chambers called it a "once-in-a-century revolution."

After the Nasdaq peaked, clients began failing en masse. In May 2001, Cisco wrote down $2.2 billion in inventory and set aside $800 million for bad debts. Its stock fell from $79 to $9.50. In December 2025, Cisco's stock price finally surpassed its 2000 high, taking 25 years. However, due to continuous share buybacks reducing the share count, its current market cap is still over 40% lower than its 2000 peak—the stock price recovered, but the company's size did not.

From 1996 to 2000, vendor financing was priced as "ecosystem empowerment" and a "demand catalyst." After the Nasdaq crash in 2000, as clients died off, Cisco's bad debts surfaced—the same numbers were reinterpreted as "when are you actually going to pay back the money?"

The shift in logic is abrupt: before the break, the equity market provided narrative legitimacy to vendor financing; after the break, the narrative disappears, and the bill is just numbers. Jim Bianco, President of Bianco Research, asserts: Before the credit market has pricing power, equity valuation is always debatable. Once the credit market starts pricing, the debate ends.

Cisco's exposure was $4 billion. Nvidia's potential arrangement related solely to OpenAI exceeds $600 billion. This time, the CDS market and debt buyers reacted much faster—it took Cisco 14 months from its peak to the write-down; Nvidia's CDS began repricing the very next trading day after the news was disclosed.

The Rewarded "Holdout"

Apple reclaiming the title of the world's most valuable company today is no accident.

Apple's latest quarterly capital expenditure was only $1.97 billion, down 36% year-over-year. In contrast, Amazon set its full-year capex guidance at $200 billion, and Alphabet raised its to $205 billion. Apple's choice was to return cash to shareholders—announcing an additional $100 billion share buyback in the second quarter, along with a dividend increase. Cook's explanation was "taking a deliberate and thoughtful approach to capital expenditures."

This doesn't mean Apple isn't investing in AI. The A19 chip has a built-in neural accelerator, Apple Intelligence is embedded across its entire system, and the new Siri is rebuilt based on Gemini. The difference lies in the path: on-device inference combined with external model access, refusing to build large-scale in-house training clusters.

From 2024 to 2025, Wall Street constantly criticized Apple for under-investing; the prevailing view was that building your own computing power equaled a moat. By the summer of 2026, the credit market reversed that same judgment—large-scale capex had transformed from a competitive advantage into a debt burden, and a restrained balance sheet went from conservative to flexible. Jay Woods, Chief Market Strategist at Freedom Capital Markets, said: "Apple was criticized for under-investing in AI, but in hindsight, it successfully avoided the capital expenditure trap."

This reversal was confirmed in price action on July 28. The market isn't just penalizing the excessively leveraged; it's also rewarding those with intact balance sheets. These two outcomes are two sides of the same coin.

The 8th Circuit Breaker of the Year

The reassessment done by the credit market on Monday took only one day to reach the accounts of South Korean retail investors.

The KOSPI hit its 8th circuit breaker of the year. Samsung and SK Hynix account for 60% of the index weight, and these two stocks contributed approximately 70% of the KOSPI's gains over the past two years.

In May, regulators approved 16 2x leveraged ETFs linked to them. Their assets grew from 4.9 trillion won to over $9 billion in two months, with retail investors holding 92%. On the way up, leverage amplifies gains, attracts new capital, and pushes the underlying stocks higher. On the way down, the same mechanism works in reverse: underlying stocks fall → ETFs are forced to deleverage → underlying stocks fall further → margin calls → forced liquidations → continued decline. In the first two weeks of July, 1.2 million leveraged accounts received margin calls, and between 320,000 and 460,000 were fully liquidated. Retail investors in their 20s and 30s accounted for 62% of the losses.

This group couldn't possibly have known what was happening before Monday's open. They didn't know the structure of Nvidia's $600 billion revolving financing scheme. They didn't know the reasons for Oracle's downgrade. They didn't know the credit market was repricing AI assets using a framework completely different from theirs. They only knew Samsung and SK Hynix were falling, and their leveraged accounts were receiving margin calls.

This is precisely the structural feature of the entire transmission chain: Information flows from professional buyers in the credit market to the equity market, from the equity market to the global semiconductor supply chain, from the supply chain to the index, from the index to leveraged ETFs, and finally to retail investors' accounts. Each layer is more lagged, each layer adds its own leverage, and each layer bears a greater impact than the one before. A 14 basis point jump in Nvidia's CDS on Monday translated into a circuit breaker for the KOSPI. For leveraged accounts, it translated into a total wipeout.

The credit market is not repricing a single stock. It's repricing an entire pricing framework that assumed "capital is always cheap" as a default. It recalculates from the root. Information propagates step-by-step along the leverage chain, amplifying at each stage. Those who knew first set the direction. Those who found out last paid the price.