Money Stuff: People Are Worried About Stock Market Leverage

Companies issue debt and they issue equity. There are various corporate-finance-y reasons to choose one or the other or some particular mix: Debt is often cheaper and tax deductible, equity is less risky and good for motivating employees, etc. But you could also tell a sort of market segmentation story. Some people want to own debt: They have some money, they want to keep it, and they want to invest it in something that promises to repay the money with a reasonable return. Other people want to own equity: They have some money, they’re willing to risk it to get a high return, and they want to invest it in something that could go to zero but also offers a potentially unlimited payoff. Companies sell some debt because there are people who want to buy debt, and they sell some equity because there are people who want to buy equity, and companies that need a lot of money will sell the stuff that people will buy.

On the other hand, there are all those corporate-finance-y reasons. You could imagine a world in which:

1. There is an enormous amount of demand for debt, from various institutions that are quite averse to risk and want a steady known return; and
2. There is a small but quite intense demand for equity: Some people have a little money to play with, they want to invest that money in stuff that could go to zero or offer a huge payoff, and they want the most concentrated possible form of that. They want stuff that will double or go to zero, tomorrow. Not, like, earn 6% over Treasuries for 20 years: double or zero, now.

Companies could collectively cater to that desire by financing themselves with, you know, 70% senior debt and 28% mezzanine debt and 2% equity: The senior debt would be reasonably safe, the mezzanine debt would be risky but offer a steady known return most of the time, and the equity would be a lottery ticket. This would, however, be terrible corporate finance. That’s too much debt; companies would be going bankrupt constantly. This is not how the world works, at all. The debt-to-total capital ratio of US companies is about 15% and has declined over time.

But what if that imagined world is right, or right-ish? What if investors want more debt than companies are offering them, and also riskier and higher-octane equity than companies are offering? That sounds like a job for the financial industry: shifting risk around, giving more risk to people who want it. The financial industry is good at putting stuff in boxes and issuing tranches of those boxes. If you have too much equity, you put the equity in a box and issue debt and equity against the box. The debt of the box adds to the stock of debt in the world; the equity of the box is riskier — more likely to double or go to zero — than the equity you put into the box. You have taken the given mix of debt and equity and used it to create more debt and riskier equity.

Here’s a Bloomberg News story about stock market leverage:

The surge in market leverage, stemming in part from the massive growth of levered exchange-traded products, retail margin accounts and hedge fund deposits at prime brokers, is stoking worries that it may exacerbate the next crisis. The demand to borrow money has driven an unusual mid-year spike in financing costs, which have reached the highest levels since December 2024.

“Leverage has become one of the defining themes for investors,” said Andy Kent, a broker at Kyte. “Margin debt is elevated, borrowing across parts of the shadow banking system continues to expand.”

And here’s a Wall Street Journal story on similar themes:

U.S. margin debt, or what investors borrow from their brokerages to buy securities, rose 54% to a record $1.4 trillion in May from a year earlier, according to Finra data. Meanwhile, high-risk leveraged exchange-traded funds that produce double or triple the daily move of underlying stocks are growing rapidly, as is trading in options tied to them. …

“I’m fearful that we’re building unintended leverage that isn’t fully understood,” said Mark Hackett, chief market strategist for Nationwide’s investment management group. “You’ve got people with a lottery mentality using margin to buy options on levered ETFs. That’s three or four layers.”

Yes! A levered ETF is, conceptually: You put $200 of stock in a box, you issue $100 of debt against it, and you issue $100 of equity against it. [1]  Now that $100 of equity represents $200 of stock; it’s twice as risky. A call option is a similarly levered bet on stock: If the stock goes up, you double your money quickly; if it goes down, you lose all your money quickly. A margin loan is borrowing $100 to buy $200 worth of stock. If you stack them on top of each other, you get super-concentrated, super-risky bets on stocks. Companies don’t want to sell you those bets, but you can make them yourself.

“This would, however, be terrible corporate finance,” I said above; it would greatly increase the risk of defaults and crises and disruption. For essentially the same reasons, you can see why people are worried about stock market leverage.

How do you know if you own a share of stock? In the very olden days, you’d have a paper stock certificate, but those are just novelty items now. Now the basic answer is: A share of stock represents a claim on a company, and so a share of stock is, in essence, an entry on a list maintained by the company. The company has a list — its “cap table” — of who owns its shares and how many shares they own, and owning a share means being on that list.

Some small startups will keep this list in an Excel spreadsheet or on a napkin or in the founder’s memory, but at large public companies it will be maintained by a professional outside contractor called a transfer agent. If I sell you some stock in a company, we essentially send an appropriate message to the transfer agent saying “I transfer 100 shares of Amalgamated Widgets to you,” and the transfer agent makes a note of it and now you own the stock. 

In practice, this rarely happens because most publicly traded US stocks are owned indirectly. One big clearinghouse, the Depository Trust Co., owns most of the shares of most of the stocks; banks and brokerage firms have accounts at DTC; and investors have accounts at banks and brokerages. If you own 100 shares of Amalgamated Widgets, what you actually own is an entry in a database at your broker showing that you own the stock, and your broker has an entry in a database at DTC showing that it owns the stock (on your behalf), and DTC has an entry in a database at Amalgamated’s transfer agent showing that it owns the stock (on your broker’s behalf). (This is called owning the stock in “street name.”) If you want to sell me some stock, and we have different brokers, there is some theoretical complexity: My broker doesn’t have your broker’s list, so how does it know if you actually own the stock? But in practice there is an interconnected system between the brokers and DTC, you tell your broker to tell DTC to send some of its stock to my broker for my account, and it all more or less works.

Occasionally this all breaks down, and we talk about it around here. Sometimes brokers just plum forget about their customers’ accounts. Sometimes DTC and its participants can’t quite keep up with all of the transfers. Sometimes transfer agents accidentally give people stock who aren’t supposed to have it. I genuinely love all of these stories — I want to read an 80-volume mystery series where instead of country-house murders the detective solves transfer-agent errors — but they are quite rare. Mostly the system just boringly works.

Big modern private companies introduce another complication, though. Those companies, for various reasons, often don’t want people buying and selling their stock. [2] If you want to sell me 100 shares of Anthropic, and we send a message to Anthropic’s transfer agent saying “you transfer 100 shares of Anthropic to me,” the transfer agent will say “no you don’t” or perhaps “let me check with Anthropic about that” (and then Anthropic will say no). Ah well.

This has given rise to new forms of indirect ownership. Most notably there is the SPV, the special purpose vehicle: You set up a legal entity (the SPV) to buy Anthropic shares directly from Anthropic, the SPV owns nothing but those Anthropic shares, and then you sell shares of the SPV to me. I do not buy any Anthropic shares, meaning that as far as Anthropic and its transfer agent know, I don’t own any shares. But I own shares of the SPV: The SPV has its list of shareholders, and I am on that list. And the SPV owns shares of Anthropic: Anthropic has its list of shareholders, and the SPV is on the list. I have an indirect interest in the Anthropic shares.

Conceptually, this is not very different from me owning shares of, like, Nvidia Corp. I’m not on Nvidia’s list either. But I’m on my broker’s list, and my broker is on DTC’s list, and DTC is on Nvidia’s list, and I have an indirect interest in the Nvidia shares.

The difference is that the public-company DTC-based indirect ownership technology works really really well almost all of the time, to the point that almost nobody ever thinks about it, while the private-company SPV-based indirect ownership technology … ehhhh. It breaks down a lot in boring tawdry ways.

Sometimes an SPV will not keep very good track of its own shares and will end up selling indirect ownership to more of the underlying stock than it owns. Other times — quite a few times — the SPV simply won’t own the underlying stock: Some SPV promoter will come to you and be like “I own 100 Anthropic shares in my SPV, I’ll sell you an indirect interest in 10 of them,” and you’ll be like “sure,” and the promoter will be lying. How can you check? Are you gonna call up Anthropic and ask if the SPV’s name is on their list? Why would it tell you? (It’s trying to stop people from trading its stock.) The public markets have a whole invisible apparatus for making sure I own the stock that I am indirectly selling you; the private markets do not.

Three possible reactions to this state of affairs are:

1. “This is fine.” Private companies are supposed to be private; their stock is not supposed to trade. If you want to own Anthropic stock, the thing to do is wait until it goes public; then you’ll be able to buy real Anthropic stock easily. You shouldn’t try to buy weird indirect private Anthropic stock now, so the fact that it is difficult and risky is fine. 
2. “We should fix this.” This is a solved problem; the public markets have figured out how to transfer indirect ownership of stock reliably and easily. You build an interconnected system where the SPVs and banks and brokerages and transfer agents can all talk to one another, so that if you want to buy Anthropic shares (in SPV form), you can check — or rather, the system will check in the background — to make sure that you are getting actual shares of an SPV that actually owns Anthropic shares. This presumably requires some cooperation from the companies, because checking that the SPV owns Anthropic shares means checking that it appears on Anthropic’s shareholder list, but perhaps there’s a way to get that cooperation. (For instance, perhaps involving the companies in the indirect-transfer system would give them more control over those indirect transfers, which might be a reason for them to help.)
3. Same as No. 2, but with the word “blockchain.” (Remember blockchain?)

Yueqi Yang at the Information reported last week:

Citigroup wants to tackle the murky ownership problems that have proliferated among private tech firms as more special purpose vehicles have popped up promising to connect investors with the hottest startups. 

The bank this month said it is launching a new service that lets investors trade private company shares on a blockchain. It’s initially limited to foreign investors—it will expand to the U.S. later. And Citi needs the companies to agree. So far one firm has done so. …

The way the service works is that Citi will issue depository receipts—a security used for companies to access foreign investors—for private companies. It will apply a serial number for each security so ownership can be clearly tracked and transferred on both the blockchain and existing market infrastructures.  …

With Citi’s private company tokens, investors “will have legal, verifiable certainty that the underlying [asset] behind those is the actual shares in the company, with the consent of the company, directly on the cap table,” said Artem Korenyuk, the bank’s global lead for digital assets enterprise alignment and services enablement. “That’s the fundamental difference that you do not get when you invest through these multilayered third-party SPVs.” 

Citi said it will get explicit permission from each company before offering its shares on the blockchain. “We don’t want to go against the company’s desires or intent,” said Korenyuk. 

Of course one way for private companies to create a system for investors to trade their shares in a centralized depository is by going public, but apparently everyone wants another way.

The basic story of private credit is that private credit managers raise institutional funds that they use to make loans that used to be made by banks. This requires raising the funds and evaluating the loans, but it also requires finding people to borrow the money. People intuitively know, from long history, that if they want to borrow money, they can go to a bank. But if you’re a private credit firm looking to lend people money, you need to find them and tell them about your services.

The most stereotypical form of private credit is making loans to fund private equity leveraged buyouts, which makes sense if you think about the marketing channels:

• There are only so many private equity firms, and they all do lots of LBOs, so you only need a few salespeople to cover the private equity firms. And if you have a good relationship with a few private equity firms, you will get a lot of repeat business.
• Many of the big private credit managers are also big private equity managers, so you can market your loans to a private equity sponsor by walking across the hall. [3] Or, at least, they are alternative asset managers, and you are an alternative asset manager; you speak the same language and went to the same schools and did the same two-year leveraged finance analyst programs at the same investment banks. The marketing is fairly easy.

But as private credit has grown, funding LBOs is not enough, and private credit has moved into various sorts of financing for big investment-grade companies, AI data centers, small businesses, etc. How do the private credit firms pitch those borrowers? The answer is some combination of:

• Hiring former bankers to come work in private credit, pitch companies and originate loans; and
• Partnering with banks: A bank has client relationships and sales skills and can find borrowers and bring in the loans, but the private credit fund has money and can fund the loans.

But as private credit has grown, funding business loans is not enough, and private credit has moved into various sorts of consumer financing. How do the private credit firms pitch those borrowers? The private credit firms are not going to hire tons of consumer bankers. You’re not going to get a phone call from Blue Owl pitching you on a credit card. The answer has to be some combination of [4] :

• Partnering with banks: A bank has a credit card program and sells some of the credit card receivables to a private credit firm; and
• Partnering with fintechs: There are all sorts of companies offering consumer loans, generally in genre known as “buy now pay later,” and those companies tend to be capital-light tech-flavored firms that are not funding those loans themselves. They need money, the private credit firms need loans, and there is a natural synergy.

In some sense, the way Blue Owl pitches you on a credit card is that PayPal pitches you on a buy-now-pay-later product. BNPL is the credit card of private credit. I wrote last year:

The modern rise of BNPL in the US is not so much a story of “fintechs offer a better user experience than credit cards” or “people are going into debt for burritos,” and more a story of “banks are retreating from consumer lending risk, and private credit firms, with their long-term capital, are better bearers of that risk.”

Anyway Bloomberg’s J.J. McCorvey and Rene Ismail report:

The private credit industry was dubbed “shadow banking” as it took business away from traditional lenders. Buy Now, Pay Later companies have been referred to as hawking “phantom debt” that falls outside Wall Street’s typical tracking methods.

Now, these two more opaque corners of finance are overlapping in a big way — and catching the attention of credit raters, former regulatory chiefs and others on guard for potential risks as US consumers show mounting signs of strain.

Officially known as “forward-flow agreements,” investing heavyweights like Blue Owl Capital Inc., KKR & Co. and Elliott Investment Management are increasingly agreeing to pre-purchase billions of dollars worth of loans before they’re made, in a bet that consumer assets will outpace returns elsewhere. That’s been a boon to the likes of Klarna Group Plc, Affirm Holdings Inc. and PayPal Holdings Inc., offering fuel for the origination machines at the heart of a business more Americans are embracing.

Right, yes, in the olden days, banks gave people credit cards; in the modern system of narrow banking, private credit firms give people BNPL loans.

I mentioned last week that private equity funds, unlike hedge funds, tend to charge performance fees only on realized gains. If a fund buys a company for $1 billion, spruces it up a bit, and decides it’s worth $2 billion the next year, it can’t book a $1 billion gain and send its investors a bill for a $200 million (20%) performance fee. It has to wait until it actually sells the company — probably years later — to realize the gain and collect its fees.

One reason for this is that valuing private companies, outside of a sale, is hard and conflicted. But another reason is just the structure of private equity funds: They raise money from investors, deploy it to buy companies, spruce the companies up, sell them for money, and return the money to the investors. Each fund has a finite life; investors put money in at the beginning and get their money out at the end seven or 10 or however many years later. The manager’s payoff tracks the investors’. (If the manager wants to buy new companies and deploy new capital, it will raise a new fund.) This is different from a hedge fund, which will generally have an indefinite life: When a hedge fund sells some stocks, it will normally use the money to buy other stocks, not return it to its investors. 

This difference is not a law of nature, though it is sort of a convenient way to run institutional capital. It is maybe not a great way to run retail capital. Calling up 50 big institutional allocators every two or three years to say “hey, doing a new fund, are you in?” is a fine approach, but marketing to retail is generally going to be smaller-scale and more continuous; you want to be able to take in new money whenever a new investor discovers your fund. Locking up the institutions’ money for 10 years is also fine; they are big and diversified and can knowingly take illiquidity risk to improve returns. Retail investors love to be able to get their money back whenever.

And so the basic private-equity fund lifespan — raise a lot of money at once, deploy it over time, return it when you sell companies — doesn’t really work for retail investors. You could imagine other approaches (run an evergreen fund, raise money continuously, reinvest sales proceeds in new investments, let investors cash out at net asset value in limited size); there are some well-known playbooks for selling private investments to retail investors. 

But the point here is that those other approaches break the link between performance fees and realization of gains. If an investor buys into the fund in Year 1 and sells in Year 3, before all of the gains are realized, should she simply not pay any performance fees? That seems very unfair to the investor who buys in Year 3: Should he have to pay the fees on all of the gains? It makes more sense for the selling investor to pay her share of the performance fees, and the simplest way to do that is to charge performance fees on mark-to-market gains. Every quarter, you see what the portfolio is worth, and if it’s up, then the manager collects 20% of the gains. 

This all makes sense, I think, but it is also convenient for the manager: Instead of waiting years to earn performance fees, the manager can collect them every quarter. Plus the difficulties and conflicts of valuing private assets remain: If you collect performance fees on mark-to-market gains on assets with no market, your marks might be a bit high.

At the Wall Street Journal, Jonathan Weil writes about those funds:

It is one thing for fund managers to record large unrealized gains on hard-to-value, illiquid assets and report stellar investment performance. It is quite another to also charge investors large performance-based fees before the profits are actually locked in.

This happens all the time at private-equity funds that cater to wealthy individuals. And the numbers in some instances are eye-popping. …

As an evergreen fund, Spring raises capital monthly while offering existing investors liquidity through quarterly repurchase offers. The incentive fees are accrued throughout the year and baked into the net asset value, or NAV, that Spring updates routinely. If the fund charged such fees only on realized gains, that would reward departing investors at the expense of those who stay for the long haul, assuming the paper gains eventually turn into real money. ...

The semiliquid nature is what makes performance fees on unrealized gains necessary, if the fund manager is going to charge them at all.

The simplest way to understand the rise of retail-oriented private investment vehicles is that those vehicles charge higher fees than, like, index funds. Sometimes they also charge nicer fees — for the managers — than institutional private investment vehicles.

At the Wall Street Journal, Jason Zweig reports on the meticulous records that a con man kept of how he ran his financial fraud. Obviously not best practices in any number of ways, including this one:

“I have to tell you that there’s risk,” Regan told one client. “But at the end of the day with insurance wrappers and protection, y’know, it’s the same risk that we have of, y’know, being ripped apart by a saber-tooth tiger or, y’know, stepped on by a dinosaur.” 

When I eventually teach a financial literacy course, I’m gonna spend some time on that sentence. People pitching real investments to real investors will spend some time on a clear-eyed examination of the risks, which always exist. People pitching Ponzi schemes will say “there’s no risk.” People pitching slightly more sophisticated Ponzi schemes will say “I am required to tell you that there is risk, but really there’s no risk.” If someone tells you that there’s no risk, including by using amusing dinosaur metaphors, run!

Comcast to Spin Off NBCUniversal and Sky, Sending Shares Soaring. King Street’s Poor Returns, Partner Exits Spur Firmwide Revamp. Chinese Hedge Funds Warn the AI ‘Super Bubble’ Is Ready to Burst. AI ‘exuberance’ risks ending in lengthy investment bust, BIS warns. OpenAI Limits Release of New Model Under Pressure From US. Anthropic’s Mythos 5 AI Model Cleared by US for Wider Use. China Has Matched Anthropic in Cybersecurity, Resetting AI Race. AI fuels record $200bn M&A boom in US power sector. Rocket Lab to Buy Iridium for $8 Billion in Challenge to SpaceX. James Sprayregen’s thirst for legal showdowns prompts late career move to Paul Weiss. Novak Djokovic Joins General Atlantic in His Wall Street Debut. Hacky Sacks Are Cool Again and Somehow Cost $23. “For many founders and builders, keeping pace with AI has become life’s singular focus.”

If you'd like to get Money Stuff in handy email form, right in your inbox, please subscribe at this link. Or you can subscribe to Money Stuff and other great Bloomberg newsletters here. Thanks!