Money Stuff: Predict the KPIs

Classically, a share of stock in a company is worth the present value of the expected future earnings of the company. Therefore, a share of stock is a bet on the future earnings of the company: If this quarter’s earnings are surprisingly good, or if the market comes to expect higher future earnings, the price of the stock will go up. I am using “earnings” in a loose and generic sense, just, like, some measure of the money that will eventually be available to shareholders. The stock’s value is a fairly straightforward function of future dividends, and a somewhat more complicated function of future net income or EBITDA or gross margins or revenue or units sold or whatever. Doubling a company’s future revenues probably won’t exactly double the future cash flows available to shareholders — it might even decrease them — but there’s some correlation.

Also, the price of an oil company’s stock might be a function of the price of oil: If oil prices shoot up, oil-company stock prices will probably go up. For that matter, the price of a sports company’s stock might be a function of sports performance: If the Knicks win the NBA championship, for instance, the stock of Madison Square Garden Sports Corp. might go up. In fact, the Knicks won the NBA championship in June, and MSGS is up more than 50% this year. I wouldn’t put a ton of weight on that one data point, and there are a lot of steps between “Knicks win games” and “future cash flows available to MSGS shareholders are higher.” But there is a connection, and in a very loose sense, MSGS’s stock is a bet on the Knicks to win a lot of future basketball games.

You could almost imagine MSGS using this for fundraising purposes. [1]  Like:

• MSGS goes out to investors to raise $100 million to spend on getting better at basketball or whatever.
• MSGS promises the investors “we will give you back $2 million for every game the Knicks win in the 2026-2027 season.”
• That has an expected payoff of about $106 million according to Kalshi, though with plenty of variance.
• The investors share in the upside (if the Knicks do great they make more money) and downside (if the Knicks are terrible they lose money) of the team.
• MSGS also shares in the upside and downside of the team: If the Knicks do great, MSGS makes more money selling tickets and whatever; if the Knicks are terrible MSGS makes less money.
• The investors do not own actual equity in MSGS; they do not own a residual claim on MSGS’s cash flow. They own a sports bet. But it has a certain family resemblance to equity; they share in the upside and downside of the company’s business performance. [2]

Why would MSGS want to raise money this way, rather than by issuing stock? Well, I mean, it wouldn’t; this is a silly hypothetical. The market for public-company stock is much deeper and more liquid and better priced than the market for sports bets, for now. But we have talked a lot, over the years, about related topics:

1. For a while, if you had a business idea, you could raise a lot of money at attractive prices by selling crypto tokens. What made this extra attractive was that you (maybe) didn’t have to comply with US securities regulations; you could sell the tokens to investors without filing public disclosures and getting audited financial statements and doing all the other stuff required of US public companies. You weren’t selling “stock,” see; you were selling “tokens.” US securities regulators sort of tolerated this for a while, and then stopped tolerating it, and are now maybe back to tolerating it.
2. More recently, there have been suggestions that “tokenized stock” of private companies could be sold to public investors without complying with securities laws, though I’m not sure anyone quite believes that.
3. We have talked recently about prediction markets on corporate valuations. You can go on Polymarket and bet on Anthropic’s future valuation, that is, its stock price. Again, Anthropic does not file public disclosures with the US Securities and Exchange Commission, and yet Polymarket offers those bets to everyone. (Except, very nominally, US traders. Very nominally!) 

None of these products, I think, quite works, as a matter of US securities regulation. The tokenized stocks and private-company-valuation bets are (supposedly) not sold to US investors. I think that tokenized stock is clearly a “security” under US law, and prediction markets on corporate valuations are clearly “security-based swaps,” and both of them are subject to the SEC’s disclosure and registration regime. They are very close substitutes for stock, which means that US securities law treats them like stock.

But what about slightly less close substitutes? Instead of selling bets on a company’s stock price, what about selling bets on its net income? Those things are different, but the stock price can theoretically sort of be decomposed into a series of bets on its future net income. Would a bet on net income — a bet that pays $1 per $100 million of Tesla Inc.’s 2026 net income, for instance, or a bet that pays $1 if that net income is above $4 billion and $0 if it’s below — be a security, or a security-based swap?

I’m not sure, but my impression is “kind of, yeah.” One piece of evidence is that Kalshi doesn’t list bets like that. Kalshi is a US regulated prediction market. It is registered with the US Commodity Futures Trading Commission, which allows it to trade all sorts of “swaps” (bets), but not security-based swaps. (It has proposed to list equity index bets, because weirdly broad-based equity indexes are commodities, not securities, under US rules.) So it doesn’t list the sort of Anthropic-valuation contracts that Polymarket (which is mostly not US regulated) does, and it doesn’t list Tesla net income contracts either.

It does list Tesla total deliveries contracts, though. You can bet, on Kalshi, on how many cars Tesla will deliver in 2026. There is probably some correlation between how many cars Tesla delivers in 2026 and its stock price at the end of the year; the stock price is in part a function of the car deliveries. But those things are different enough that Kalshi can reasonably argue — and the current SEC and CFTC seem to agree, or at least not care — that the total-deliveries contract is not a “security-based swap.” It’s a bet on cars, not a bet on Tesla.

That is: A bet on Tesla’s stock price is a close substitute for buying stock, and so is regulated by the SEC and off-limits, for now, to Kalshi. A bet on Tesla’s net income is a somewhat more indirect substitute for stock, and so is maybe regulated by the SEC and maybe off-limits to Kalshi. A bet on Tesla’s car deliveries is an even more indirect substitute for stock, and so is maybe not regulated by the SEC, and listed on Kalshi.

Kalshi definitely lists Knicks total wins contracts. 

Bloomberg’s Bernard Goyder reports:

Kalshi is asking regulators to delay products from one of its competitors, Cboe Global Markets Inc., as the friction between upstart prediction markets and incumbent financial exchanges heats up.

The event-betting platform sent a letter to the Securities and Exchange Commission this month, asking the agency to hold off on approving new binary options contracts tied to specific line items in corporate earnings reports — products that would compete with some so-called event contracts already offered by Kalshi.

The dispute flips the script on previous industry debates in which Cboe and CME Group Inc. have both argued that prediction market products have been approved too quickly by the main regulatory agency overseeing the platforms, the Commodity Futures Trading Commission. …

Previously, many financial products tied to public stocks — such as equity options — have been governed by the SEC. But Kalshi has operated its contracts under the oversight of the CFTC, which has said that prediction markets are derivatives exchanges that should come under its oversight.

Cboe, on the other hand, sought approval from the SEC for its new binary options tied to corporate performance metrics.

Cboe and CME have both complained that the CFTC has allowed new kinds of prediction market contracts to start trading with minimal scrutiny, in contrast to the SEC’s slower process, which, they argue, has been rigorous and attuned to investor protection.

Here is Cboe’s proposal, and here are the public comments filed with the SEC. (Here’s Kalshi’s, and here’s Cboe’s response.) Cboe wants to “amend its Rules to permit the listing of binary options overlying key performance indicators (‘KPIs’) reported by certain issuers of stock (‘binary KPI options’).” A binary option is a yes/no bet that pays $1 if the KPI is above the contracted level and $0 if it’s below. It seems like the bets would be on things like earnings per share, net income, total revenue and segment revenue. You could bet on whether Tesla’s earnings per share this quarter will come in above or below $0.40, for instance. For that matter, Cboe would also let you bet on Tesla’s “Model 3/Y Production (#),” that is, how many cars it produces.

These would, in Cboe’s view, be securities bets. They would be equity derivatives, just like options on a stock’s price. They would be regulated by the SEC. They would be part of the stock-market ecosystem regulated by the SEC, not part of the swaps/bets/prediction-market ecosystem regulated by the CFTC.

Kalshi disagrees, for somewhat obvious reasons. The point, for our purposes, is that, for now:

1. Stocks and stock options are listed on stock (and options) exchanges, regulated by the SEC, and subject to a regulatory regime in which companies need to disclose a lot of business information if they want public investors to be able to buy their stocks.
2. Like, sports bets are listed on prediction markets, regulated by the CFTC, and not subject to the same sort of corporate financial disclosure regime. You can bet on the Knicks, whose financial statements are public, or on the Lakers, whose aren’t. The CFTC, correctly, does not consider it essential that every sports team’s audited financial statements be disclosed to bettors. 
3. Bets on companies’ performance, bets that are correlated with their stock price but are not exactly their stock price, are up for grabs. A bet on net income or revenue or units sold is not the same as a stock investment, but it’s not so different either. Is is part of the stock market, regulated by the SEC, requiring corporate disclosures? Or is it part of the betting markets, regulated by the CFTC, not requiring those disclosures? Perhaps we’ll find out.

Of course my real interest is in sports teams raising money by selling sports bets. Or more pragmatically: What if net-income bets, or KPI derivatives more generally, are not securities? Could a private company raise money from public investors by selling net-income contracts on prediction markets? A share of stock is a bet on a company’s future income, but maybe a bet on a company’s future income is not a share of stock.

Here is one bad thing that could happen. A company raises money by issuing bonds. It wants to sell, say, $1 billon of bonds. In the bond offering, it tells investors that it has earnings of $300 million per year. Investors read the prospectus and think “ah, this company makes plenty of money to pay back these bonds,” and they agree to buy the bonds at, say, a 6% interest rate. In fact the prospectus is wrong and the company actually earns $0 per year. The first interest payment on the bonds comes due and the company says “whoops, no money.” It defaults on the bonds, it goes into bankruptcy, and the bondholders get back $0 of their $1 billion. I think it is self-evident why this is bad.

Here is another bad thing that could happen. A company raises $1 billion of bonds at 6% by telling investors that it has earnings of $300 million per year. In fact the prospectus is wrong and the company actually earns $100 million per year. The first interest payment on the bonds comes due, and the company pays it. In fact, it makes all of the interest payments when due — $100 million is much less than $300 million, but it is enough to pay $60 million of interest — and at maturity it repays the full $1 billion. The bondholders get back their $1 billion, plus the promised interest.

Is this bad? I mean, you could make an argument that it’s fine. Like:

1. The bondholders bought the bonds because the company promised to repay them their principal with an agreed interest rate, and it did, so that’s fine. No bondholder lost any money: They invested $1 billion and got $1 billion back with interest.
2. The bondholders agreed to the fairly low 6% interest rate because they concluded that the company was relatively safe, that it would be able to repay the principal and interest without too much trouble. In drawing that conclusion, perhaps the bondholders considered the (incorrect) disclosure that the company earns $300 million per year; who can say really. That disclosure was wrong, but the conclusion was right: Ex post, the company really was safe, and it really was able to repay the principal and interest. The company’s realized credit risk was low, and therefore its 6% interest rate was fine.

These arguments seem bad? You can probably spot the flaws. Here are a few, though you can doubtless add others:

• Most risky bonds do not default, so “this bond got paid back and therefore its realized credit risk was low” is not a real argument.
• The ex ante probability of default was higher than the market thought — the company had less cushion to pay its debts than investors thought — which is the bad thing.
• If, a year after issuing the bonds, the company had said “whoops actually we make $100 million per year,” the market price of the bonds would have gone down. (Their expected yield would have gone up.) Bondholders would have lost money, on a mark-to-market basis. Of course if they held to maturity they’d get their principal back, but that is not the only relevant measure.
• If, before issuing the bonds, the company had accurately disclosed its earnings, bondholders would have charged a higher interest rate. Therefore they did lose money, measured against the correct baseline: Had they known the true facts, they would have gotten paid more interest.
• In fact there are cases of companies getting in trouble for this sort of thing: If you make incorrect financial disclosures to bondholders, that’s arguably securities fraud, even if you pay the bonds back on schedule.

Here is another thing that could happen. An insurance company raises $1 billion of annuity money. That is, it goes out to 1,000 investors and says to each of them, “if you give me $1 million today, I will pay you a steady income for the rest of your life.” [3]  That is like a bond; the company is raising cash today by promising payments over time. You could imagine the investors evaluating it like a bond: “This company has plenty of capacity to make the promised payments,” the investors might think, “and therefore I will accept an expected annual return of about 6% to reflect the safety of this investment.” But that’s not a real thing. I mean, that’s how the bond market works, [4]  but it is not a reasonable thing to expect of retail annuity buyers. People looking to buy annuities do not, generally, scrutinize the financial statements of insurance companies and choose between buying an annuity from a safe company at 6% and buying one from a risky company at 8%. Evaluating the financial strength of an insurance company is a complex and specialized business, and even figuring out the implied yield of an annuity — figuring out what sort of credit spread is embedded in the annuity product — isn’t always easy. [5]  You can’t say “this company brings in $300 million a year so I’m happy to get a 6% yield on my annuity,” because you can’t figure out how much the company brings in or what yield you’re getting on your annuity.

Instead, there is a somewhat more binary system in which state insurance regulators decide which insurance companies are safe. Safe companies can sell annuities to raise money; unsafe companies cannot. Evaluating the financial strength of an insurance company is a complex and specialized business, so it is done by state insurance regulators using a risk-based capital framework. If you have $X of capital, you can raise up to $Y of annuity money, etc. Insurance customers do not have to evaluate an issuer’s financial strength, because regulators do. 

This is exaggerated — some customers and their advisers surely do consider the financial strength of insurance companies, and safer companies probably have a lower cost of capital than bare-regulatory-minimum companies — but it is a useful approximation.

So one bad thing that could happen is: An insurance company wants to raise $1 billion of annuity money, it tells its regulators “we earn $300 million a year so we can easily cover those annuity payments,” [6]  the regulators are like “you sure can, no problem here,” the insurance company sells $1 billion of annuities at market rates, but in fact the insurance company makes $0 per year and can’t make any payments on the annuities. It defaults, it becomes insolvent, it is seized by regulators, and a state guarantee fund pays out some but not all of the money owed to customers. Self-evident badness! 

And then another bad thing that could happen is: Raise $1 billion of annuity money, say “we make $300 million a year,” regulators say “sure,” but it turns out you actually earn $100 million per year. But it’s fine, and you make all the required payments on the annuities. The annuities were, ex ante, riskier than the regulator thought. Had the regulator known the true state of affairs, it wouldn’t have let you sell all the annuities; it would have required you to have more capital against your asset base. Ex post, everything was fine. But the regulation is risk-based, and if the regulator doesn’t have accurate disclosures then policyholders and state guarantee funds are taking more risk than they want.

The point here is that it is not the customers who were misled by the wrong disclosures; the customers didn’t read the disclosures. The regulator was misled. And, similarly, when the disclosures are corrected, what happens is not that the market price of the annuities drops; the annuities don’t trade or have a market price. What happens instead is that the regulator demands more capital, to reflect the higher-than-expected risk.

I’m just using the simplest possible bad thing here, the company saying that it has more money than it actually does. In the real world, there are subtler — and less bad — forms of badness. The company could say “we hold $2 billion of investment-grade corporate debt to back our  insurance obligations,” but actually some of that debt was downgraded, or the ratings agency had incomplete data or conflicts of interest when it assigned those investment-grade ratings. Or regulators might have questioned those investment-grade ratings if they had known that actually the corporate debt was issued by affiliates of the insurance companies. That’s all stuff that makes the insurance riskier ex ante, but in a diffuse, hard to measure way. If the company had disclosed everything to everyone perfectly, it probably would have been required to have more capital to back its insurance obligations. But it’s probably fine! The insurance will probably get paid! 

Anyway:

TWG Global, the holding company at the heart of Mark Walter’s empire, hit back against what it called “multipronged attacks” on its business as US prosecutors continue to probe the firm.

The company is working with both the US Department of Justice and the Securities and Exchange Commission to resolve their inquiries, according to a statement Wednesday. Its insurance business has also submitted plans to its regulators to try to eliminate any concerns they have, the company said.

“Despite what has been reported, there has been no fraud,” TWG said. “There is no victim here. No one has been harmed, and no one has claimed they were harmed.”

We talked a bit about the TWG situation yesterday: TWG’s insurance companies “ disclosed more than $20 billion of loans that should have been labeled as affiliated transactions, but weren’t,” which probably means they should have had more regulatory capital than they did; they are now working with regulators to fix the problem. From the statement:

As it relates to Group 1001 [TWG’s insurance group], at its core this is a regulatory matter with a straightforward plan that has been submitted to its regulator to promptly eliminate all of the affiliate exposure at the Group 1001 insurance companies.

There is no victim here. No one has been harmed, and no one has claimed they were harmed. …

Affiliated transactions are commonplace in the insurance industry, widely permitted subject to applicable regulatory requirements, and a part of the insurance industry’s normal course of business.

Affiliated transactions should be properly disclosed, but to state that they “generally” have the potential to “loot” the insurer is untrue.

The reality is that Group 1001 has invested in real assets that are performing well; the insurance companies have recognized significant income from the investments and no policyholders have lost money because of these transactions.

As part of the plan, TWG is proposing to purchase the affiliated assets from the insurance companies, reflecting its confidence in the quality and performance of those assets.

It’s probably fine!

The Wall Street Journal has a good profile of Leopold Aschenbrenner, the 24-year-old founder of Situational Awareness, the hedge fund whose semi-collapse we have discussed around here. Notably, Aschenbrenner got married earlier this month, just days after selling most of his public stock portfolio to Citadel to meet margin calls. The Journal story includes this story from the wedding:

When it was time to toast the newlyweds, one speaker thanked Citadel CEO Ken Griffin for making it all possible. 

Just great wedding-speech roast material. Also this can’t be the first time that Griffin has been thanked “for making it all possible” in a wedding speech; if you have further examples please email me.

Anthropic Expected to Tell Investors It Sees Over $30 Trillion in Potential Revenue. The multiplying risks of financing data centres. Mark Walter’s Troubles Are Disrupting the Insurance World’s Hottest Trade. Insurers Want a Bigger Slice of the Bank Risk Transfer Boom. How four Russian bankers made millions from EU sanctions. Meta Says It Will Pay Up to $18 Billion Over Social Media Claims. Vanguard Buys Wealth Management Platform Altruist in $4 Billion Deal. Unlikely Winners of the Data Center Boom: Sound Consultants. Americans Say They Feel Guilty About Spending Money on Fun.

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!