The artificial-intelligence boom is most often seen as a Wall Street story.
Nvidia (NVDA) rallies. Big Tech spends billions. Data centers multiply. Investors debate which company will become the next major AI winner.
But Bank of America says something more substantial is starting to unfold under the surface.
AI is splitting firms into two paths: those turning the technology into tangible growth and those being disrupted by it.
The BofA research note of Aug. 20, shared with TheStreet, paints a stark picture of this gap in the leveraged-finance sector.
And although bond spreads and leveraged loans may seem distant from everyday life, the companies borrowing in these markets employ workers, buy equipment, and supply products and services that consumers use.
If one group can develop and invest and the other faces slower revenue growth and more expensive financing, the implications can ultimately reach Main Street through hiring, wages, company investment, and corporate cutbacks.
BofA finds a widening gap between AI winners and losers
BofA separated issuers of high-yield and leveraged loans into those benefiting from AI (the “tailwind” group) and those facing AI-related disruption (the “AI-risk” or “headwind” group).
The data reveal a shocking gap.
Among high-yield borrowers, BofA’s AI-tailwind group posted 16.2% year-over-year revenue growth in the second quarter and 15.2% growth in adjusted EBITDA.
The AI headwinds companies grew their revenue by only 4.1% and EBITDA by 7.6%.
The disparity was considerably more pronounced for leveraged-loan debtors.
AI beneficiaries saw a 26.8% revenue increase compared with a 3.6% gain for enterprises vulnerable to AI disruption, according to BofA.
The AI-tailwind cohort’s EBITDA growth was 23.4% vs. 3% for the AI-risk cohort.
BofA describes the emerging pattern as a “Credit-K,” basically two groups going in significantly divergent ways.
That’s important because companies that make more money tend to have more room to hire, invest, and service their debt.
Companies on the weaker side of that split may have to make some tough decisions.
AI hardware is where the money is showing up first
The tech sector was again one of the best-performing sectors in leveraged finance during the second quarter.
But not all companies shared equally in the spoils.
Hardware companies in the leveraged-loan market recorded nearly 48% revenue growth and more than 50% EBITDA growth.
Hardware sales among high-yield issuers climbed 32.1%, and EBITDA jumped 85.6%.
Software and services were much softer.
Revenue rose just 4.5% among loan borrowers and 5.8% in high yield. That gap helps explain where the AI boom now stands.
The first to benefit are the companies that sell servers, data center gear, and other physical infrastructure, as huge sums of money are being invested to build out AI capability.
The advantage for firms only now rolling out AI is less clear.
BofA’s investigation identified substantial AI benefits at 31% of S&P 500 companies, compared with just 18% of leveraged-finance issuers.
Thus, the AI cash bonanza is still concentrated in the companies closest to building the technology.
Big Tech is borrowing billions to build the AI economy
In the debt market, the scale of the investment becomes evident.
Companies have sold over $335.7 billion in AI-linked U.S. dollar debt year-to-date, according to BofA.
Investment-grade debt accounts for some $255 billion of that.
High-yield borrowing is adding another $40 billion. Loans and direct lending each are contributing about $20 billion.
Some of America’s most known corporations are in the middle of that funding tsunami.
- Alphabet (GOOGL) alone sold almost $25 billion of investment-grade debt focused on AI on Aug. 6, according to BofA’s transaction tracker, after raising nearly another $20 billion in February.
- Amazon (AMZN) sold $25 billion of debt in July, following an even larger debt sale in March.
- Nvidia (NVDA) raised around $25 billion in June.
- And Meta Platforms (META) sold almost $25 billion of bonds in April.
Those numbers are another way to think about AI.
Big Tech isn’t only hooking up chatbots to current products.
Companies are pouring unprecedented amounts of capital into data centers, semiconductors, electricity, and infrastructure.
AI borrowing comes with a price
And bond investors aren’t perceiving that expenditure as risk-free.
Investment-grade AI debt now trades at spreads of roughly 119 basis points, or around 46 basis points wider than comparable non-AI debt, says BofA.
High-yield AI spreads are around 320 basis points, a premium of about 147 basis points over similar non-AI debt.
Related: AI agents create new problem for enterprise software
That suggests lenders understand the tension at the heart of the AI boom.
If built, the infrastructure could generate huge long-term profits.
More AI:
- Nvidia just made a move Wall Street wasn’t ready for
- Microsoft just took sides in AI policy fight
- OpenAI just disclosed something genuinely alarming
But someone has to pay for it first.
And those costs of financing weigh more for weaker borrowers than they do for cash-rich technological companies.
The AI boom is still leaving many companies behind
The underlying fundamentals in leveraged finance remain pretty solid overall.
High-yield revenue increased 7.5% year over year in the second quarter, with adjusted EBITDA up 9%.
Leveraged-loan borrowers delivered 8.6% revenue growth and 9.5% EBITDA growth. BofA expects the broader market to maintain solid momentum in the third quarter. But the headline numbers hide increasingly different realities.
Technology and energy are performing strongly. Real estate remains pressured by elevated rates and housing affordability. Retail high-yield revenue increased only 1%, while earnings declined 2%.
Food producers had a modest 2% increase in revenue and an 11% fall in earnings as rising commodity and freight prices squeezed margins.
That’s why BofA’s AI gap matters more than traders monitoring bond displays.
If the businesses that can do AI keep pushing away, cash could progressively flow into areas that are already benefiting from the technology.
That might decide which companies expand, whose factories and data centers are built, and eventually where new employment is produced.
The first stage of the AI boom was about exhilaration.
The following stage is getting a lot more tangible.
Now firms have to prove that standard artificial intelligence can deliver real revenue, improved margins, and enough cash flow to justify hundreds of billions of dollars of investment.
And Bank of America’s most recent data reveals that the winners are starting to separate from the rest of the pack.
