Apollo Global Management will lend against the New York Yankees at 10 to 20 cents on the dollar of the franchise's value. In most industries, Jim Zelter said, a lender is at 50 to 70 cents.
Wall Street has spent about 18 months arguing whether to own the equity or the debt of the artificial-intelligence build-out. Zelter's answer is that it depends on whose balance sheet the money comes off. Half of Apollo's capital sits on regulated insurance balance sheets that have to earn roughly 7% a year, and that mandate rules equity out.
"I'm often fond of saying we don't want to take equity risk for a fixed coupon."
Zelter is President of Apollo, which this year financed Intel, Broadcom and Nvidia, and he has been on this program 15 times since last January — he counted — mostly to say that rates would stay high.
The full segment is covered here so you can skip it.
Here are the 7 insights that matter.
👤 Guest: Jim Zelter, President of Apollo Global Management, which manages regulated insurance balance sheets alongside its private equity funds and has arranged debt financings this year for Intel, Broadcom and Nvidia
🎙️ Host: Jonathan Ferro, Bloomberg TV anchor, who put the questions with two Bloomberg colleagues at the table
📰 Published: 16 September 2026 on the Bloomberg Talks feed
🟣 Apple Podcasts | ⏱️ 14 min
Key Takeaways
Apollo bought into sports because it is the business least likely to be made obsolete by artificial intelligence
Of the 100 most-watched shows in a year, he said 95 to 97 are sporting events
A loan against a team is secured at 10 to 20 cents on the dollar, against 50 to 70 cents elsewhere
The highest gross margins in artificial intelligence are not at the model companies
Which is why Apollo financed Broadcom and Nvidia rather than picking a winner in large language models
Nobody can lend OpenAI or Nvidia money at the scale their equity investors have committed
Four or five of Nvidia's top holders own between 2% and almost 9% of a company whose stake he sized at $500 billion
Artificial intelligence is on course to be a tenth of the whole investment-grade bond market
Access to capital has become the competitive advantage, because the industry turned asset-heavy
He expects the 10-year Treasury yield to sit between 4.25% and 5.5% for a long time
A year ago the market was priced for three or four cuts
1. Sports as a Hedge
The segment opened on Apollo Sports Capital and whether the sports push and the artificial-intelligence push are the same trade. Zelter said they are.
The screen he applied to sports was obsolescence risk, not growth. He said the question he has been asking for several years is: "What are industries, what are activities, what are businesses that will have a very low chance of obsolescence?"
His evidence is the concentration of live audiences. Leisure time has grown over the last 50 to 100 years, and sport now dominates what people watch together: "If you look at the top 100 shows in a year, 95, 97 are sporting events."
He said that is unlikely to be disrupted, and that the view is three or four years old — a deliberate decision to put Apollo in the middle of those activities rather than an opportunistic one.
Where Apollo differs from other firms in sports is that it lends as well as owns. He said most of the industry has taken equity stakes; Apollo has done equity, debt financing and capital-expenditure expansion.
2. Yankees Loan-to-Value
Asked why the debt is the more interesting side, Zelter gave the collateral math.
Teams have historically had few financing alternatives, which leaves the lender with an unusually low loan-to-value. "You're loaning 10, 15, 20 cents on the dollar of value versus other industries, 50, 60, 70."
He framed that gap as a margin of safety on assets he called world-class and unique, and said Apollo wants to be associated with them for decades.
The same logic produced the Atlantic Aviation transaction, a fixed-base operator serving private aviation at airports, which he also put in the category of businesses he does not expect to be disrupted.
3. Calling All Precincts
A host raised the debate that has run on Wall Street for 18 months: debt or equity in the industrial build-out. Zelter said the capital dictates the answer.
Half of Apollo's capital comes from regulated balance sheets with a target of about 7%, plus or minus, depending on rates. With three- and five-year yields where they are, he said the firm can hit that with high-quality spread at short duration and take no equity risk.
His phrase for the financing need is "calling all precincts." He said the capex cycle is unprecedented in scale and will take public equity, private equity, private capital, investment-grade debt "and everything else in between."
That is why Apollo shows up as a lender rather than an owner in artificial intelligence, and he named the Intel, Broadcom and Nvidia financings as the examples.
The claim that most cuts against the consensus trade is where the profit is. He said the gross margin today is highest away from the models: "And so companies like Broadcom and many others are doing very, very well, notwithstanding who's the winner of the LLM race."
4. Why Not Fund OpenAI
Asked whether a large bond slate from OpenAI and Anthropic would interest him, Zelter said the constraint is capacity, not appetite.
His illustration is the size of Nvidia's equity register. Four or five of its top-10 investors hold between 2% or 3% and almost 9% of the company, a stake he put at $500 billion.
No lender or group of lenders can write that ticket. "The biggest companies in the globe are not going to have people provide that scale of debt."
He said Apollo does want to be front and center in financing OpenAI and Anthropic, and has been, but that it comes down to sizing and diversification.
The structural difference he drew is that equity tolerates concentration and debt does not. A host noted the convexity in equity; Zelter said debt gets paid back at par, so it is "much more of a diversification game."
5. AI as 10% of the IG Market
A host asked why Nvidia has been trying to anchor borrowing costs across its ecosystem if demand for the paper were really there. Zelter called it a complex question and answered it with issuance data.
The public investment-grade market has taken a large amount of supply this year, and he said forecasters now expect the artificial-intelligence complex to reach 10% of that market: "People are predicting that that system will be 10% of the IG market."
No single market is big enough on its own. "So there's no one asset class that's large enough to fulfill the aspirations of all these companies."
His example of doing it properly is Alphabet, which he said went to the debt market and to a mandatory convertible earlier this year rather than relying on one channel. The hosts added that Amazon and Alphabet have also tapped equity and different currencies.
On who gets crowded out, he said private capital itself has changed shape. For 20 years the industry bought asset-light businesses — software, distribution — and it has now moved into asset-heavy ones.
The conclusion is a winner-take-most market where funding is the moat. "There's the winners and the haves and the have-nots, and having access to capital in scale is certainly a competitive advantage."
6. Mortgages North of 7%
A host drew the contrast between an artificial-intelligence company growing 30% borrowing at 6% or 7% and a small manufacturer that has absorbed one supply shock after another, with an investment-grade index yielding close to 6% and refinancings coming.
Zelter's first answer was housing. "I do worry about where mortgage rates are right now. I think that's going to have a greater impact than north of 7% right now, which is not great for overall strength." He added that it is not good for the administration heading into the midterm elections either.
He put the level of long-term rates down to three things at once: inflation, fiscal concerns and a large amount of bond supply hitting the market.
On when the cost of capital finally bites: "But you're asking the good question about at what point does the economy say no mas, and the actual rate of cost of capital is going to slow down things. We're not seeing it right now."
The risk he does flag is narrower participation rather than a stop. He said financing costs could reduce the breadth of the capital-expenditure cycle, and called it a macro concern investors have to hold.
He argued the transmission from rate rises to slower growth barely works any more, because the structure of the banking system is very different from 20 years ago — something he said few people spend time on.
His bottom line on rates: "Rates are going to be higher for a while, and whether that's one rate hike today, which I suspect the Fed will do, or a handful over the next six to 12 months, we're in a higher rate environment, and I don't see it going back anytime soon." He twice credited Torsten Slok, Apollo's chief economist, for making the higher-for-longer case ahead of the market.
7. The 4.25-5.5% Zip Code
Asked whether this is a good time to be a bond investor, and how much further the Fed could go, Zelter answered from Apollo's own book.
He said Apollo is not a macro investor: 80% of its assets are credit or credit-like, with heavy investment-grade exposure held against regulated balance sheet and insurance liabilities.
On that mandate the current setting is close to ideal. "If you're in that world, the world we are in right now, and base rates are between 4.5% and 5%, it's an amazing time to be a credit investor." He ran through what a credit investor wants — a strong economy, strong activity, strong and broad equity and merger markets — and marked each one off.
He refused to forecast far out, and used the market's own recent error as the reason. "Well, you know, I think whenever anybody in these seats starts to extrapolate for years and years, that's a mistake." A year or so ago the market was pricing three or four cuts, he said, and that was wrong.
What he will say is that almost nothing costs less than it did one, three or five years ago, anywhere he travels in the G7.
The range he expects on the 10-year Treasury yield is between 4.25% and 5.5%: "I think that's a zip code we're going to be in for quite some time."
Bonus Insights
The segment opened with the hosts reading Apollo's own earnings-call language back to him: "The sheer size of the AI infrastructure build-out is unprecedented. We see an enormous opportunity for private capital to finance a portion of this, along with public capital."
Zelter said Apollo took no August break — "It was pretty much pedal to the metal" — through a run of announcements, travel and financings.
He brought Yankees caps to the studio, and one of the hosts said he would embroider a couple of extra lines around the Y before wearing it.
On why the United States still draws the capital despite having higher rates than the rest of the world, he said it remains "the beacon of economic activity," which he takes as evidence of underlying strength rather than a warning.
Zelter's bottom line is that the artificial-intelligence build-out is too big for any single pool of capital, so the money has to come from everywhere at once — and that an investor with a fixed-return mandate should take the coupon rather than the upside, because rates are staying high enough to make credit pay.
Products, Companies & Tools Mentioned
Apollo Global Management (Half its capital sits on regulated balance sheets targeting about 7%, and 80% of assets are credit or credit-like, which is why it lends to the AI build-out rather than owning it)
New York Yankees (The financing he used to explain loan-to-value on a sports franchise: 10 to 20 cents on the dollar against 50 to 70 elsewhere)
Atlantic Aviation (A fixed-base operator for private aviation at airports, which Apollo announced a transaction with — another business he does not expect to be disrupted)
Nvidia, Intel and Broadcom (The three AI financings he named; he also used Nvidia's equity register to argue no lender can match the scale of its shareholders)
OpenAI and Anthropic (He wants to be front and center in financing both, but said sizing and diversification limit how much debt any one lender can provide)
Alphabet and Amazon (Cited as companies raising across debt, equity, convertibles and different currencies because no single market is deep enough)
Listen to the full episode:
If this was worth your time, send it to someone who has to have a view on this.
Get the latest market chatter as it happens: