For a while, sports betting in the US was regulated by state law. Some states banned sports betting; others allowed it under various regulatory regimes. But then Kalshi, the prediction market, started offering sports gambling nationwide and claiming, mostly successfully, that it was exempt from state regulation. Kalshi is a commodities exchange registered with the US Commodity Futures Trading Commission; its theory is that (1) sports bets are a sort of commodity “swap” subject to CFTC regulation and (2) the CFTC has exclusive jurisdiction to regulate those swaps. The states can’t regulate Kalshi, because the CFTC does.

But the CFTC didn’t, not really. CFTC regulation of sports betting was mostly hypothetical: Kalshi would “self-certify” the sports bets it wanted to list, and the CFTC wouldn’t say anything, and that was that. There were no federal statutes or regulations about what sports bets are allowed. Kalshi just decided on its own what bets to list, and if the CFTC objected it could let Kalshi know. I get the sense that the CFTC didn’t object much, though I don’t really know; for all I know there was an informal back-channel regulatory process in which Kalshi regularly asked the CFTC “hey can we take these bets” and the CFTC was like “nah.” But as a formal matter, Kalshi’s sports betting was essentially unregulated: It was exempt from state gambling regulation because it was subject to federal gambling regulation, but there was no federal gambling regulation.

Now that’s changing:

Trump regulators are proposing an expansive new set of rules on how they will govern the booming prediction markets, with parameters that will continue to allow most sports-related bets while trying to avoid inviting obvious manipulation.

The Commodity Futures Trading Commission is seeking the ability to block prediction wagers it finds aren’t in the public interest or that seem highly susceptible to manipulation, such as in cases where one person could have outsize impact on the result. The regulator proposed the new rules on Wednesday, confirming an earlier report in The Wall Street Journal.

Here is the CFTC’s announcement, and here is the text of the proposed rulemaking. These rules, if they eventually go into effect, will be the first set of federal sports gambling regulations, which is just interesting. The CFTC has to answer questions that no federal financial-markets regulator has ever considered, questions like: Should there be betting on Little League games? No, says the CFTC (page 122):

The Commission preliminarily believes that event contracts that settle solely by reference to games, sporting events, or outcomes in which participants are below the collegiate level raise public interest concerns. There are several factors that differentiate pre-collegiate sports from sports at the collegiate and professional levels. Since pre-collegiate sports have less extensive governing bodies and typically lack a rigorous integrity infrastructure, prediction markets would be less able to interface with the governing body. Also, the relevant data flows (to the extent formal data are collected at all) are decentralized and less reliable than for collegiate and professional sports. Similarly, broad and numerous groups of individuals would potentially have inside information about pre-collegiate sports and would be subject to little or no contractual limitations on information usage. … The Commission preliminarily believes that these differences from professional and collegiate sports raise particular concerns about manipulation, settlement integrity and information leakage.

Reasonable! 

There is a structural difference between Kalshi and traditional sportsbooks. The difference is that Kalshi is a market, where customers can trade either side of an event contract against each other, while traditional sportsbooks are principals, which take the other side of bets with their customers. I often write that this difference is overstated — sportsbooks take both sides of many bets, and a lot of Kalshi volume goes through market makers who function like sportsbooks — but it turns out to be quite important here. Sportsbooks are regulated by state gaming regulators, which are essentially consumer protection regulators. Kalshi is regulated by the CFTC, which is essentially a market integrity regulator.

So, for instance, state gaming regulators do not seem to have very sophisticated rules about insider trading (betting). [1] Why should they? A sportsbook will decide whether to take bets on some event based on whether it expects taking those bets to be profitable. If the event raises a huge risk of insider trading, the sportsbook just won’t list it. If it lists a bet, and then gets a ton of suspicious flow on one side of the event, it will probably limit the accounts of the suspicious bettors. If they win, it might report them to authorities or try to void the bets. “This market raises too much risk of insider trading or manipulation” is a problem for the sportsbook, which is on the other side of all the bets, not for the regulator.

Whereas on Kalshi, at least nominally, public customers are on both sides of every bet, and the CFTC’s job is to protect them from manipulation and insider trading. And so, when the CFTC thinks about which contracts to allow and which to forbid, its focus is on insider trading and manipulation.

So, for instance, the CFTC will let you bet on the outcome of a baseball game, because that is relatively hard to manipulate. But betting on whether an individual pitch — say, the first pitch of the third inning — will be a ball or a strike is easy to manipulate, so it’s forbidden (pages 120-121):

The Commission preliminarily believes that event contracts that settle solely by reference to a discrete action, event, or occurrence in sporting events, including, without limitation, event contracts settling on the type of a specific play called for or executed by a specific player or team, the type or outcome of a specific pitch thrown by a specific pitcher, the outcome of a specific shot taken by a specific player, or whether a specific player or team commits a specific foul or penalty, present public interest concerns.

Betting on player injuries, similarly, raises insider trading concerns. The CFTC’s general view is “that event contracts are more likely to be contrary to the public interest when any meaningful information about whether the underlying event will occur is unavailable to the broader market” (pages 97-98):

This includes events that are entirely random or where insight into the underlying event is highly concentrated—in a single individual, for example, or only individuals legally prohibited from transacting—and relevant information is necessarily concealed from the public. In such cases, the Commission would consider whether buyers and sellers have any basis to form a meaningful view on the underlying event, and whether the resulting prices can reasonably be expected to reflect informed market sentiment. This factor could also apply where the only market participants with insight into the underlying event would be legally prohibited from transacting in the event contract.

And so injury bets are bad (page 119):

The Commission preliminarily believes that event contracts that explicitly settle solely by reference to the duration, severity, occurrence, or medical diagnosis of an injury sustained by a specific athlete raise serious public interest concerns. First, such event contracts create perverse financial incentives that could  encourage or facilitate physical harm to athletes. Second, the settlement of such event contracts would likely depend on medical diagnoses, which raises public interest concerns about the confidentiality of medical information and the potential for such sensitive information to be leaked or exploited by insiders. Third, settlement conditions based on a physicians’ diagnoses or injury reports do not provide a sufficiently objective, verifiable, and manipulation-resistant basis for contract settlement. 

Though notice the first point there, that “such event contracts create perverse financial incentives that could  encourage or facilitate physical harm to athletes.” Part of what the CFTC is doing here is acting as a market regulator, trying to make sure that federally regulated sports bets are not rife with insider trading and manipulation. But part of what it is doing here is acting as a sports integrity regulator. The CFTC does not want federal sports betting to create incentives to injure athletes. And so, similarly, bets on fights — not boxing matches, I mean, but fights in baseball or hockey or whatever — are prohibited (pages 121-122):

The Commission preliminarily believes that event contracts that settle solely by reference to physical altercations, fights, or conduct between players or participants in the game that are subject to penalty, ejection, or disciplinary action raise public interest concerns. Such event contracts could create a direct financial incentive for both athletes and market participants to encourage, facilitate or provoke such conduct. Even if the probability that any athlete or market participant acts on such an incentive is low, the effect of a market in physical altercation contracts on the culture of athletic competition is inconsistent with the public interest. 

The point here is not so much to protect bettors from manipulation or insider trading; it’s to protect athletes — and “the culture of athletic competition” — from the bettors. It would be upsetting if the rise of federal sports betting leads to more fights at baseball games.

Also, I am somewhat kidding here in describing these rules as being about sports betting. Only somewhat: Perhaps 80% of Kalshi’s volume is sports bets, and a good chunk of the CFTC’s proposed rules are about sports. But technically prediction markets are broader than sports, and the CFTC’s proposal is about how it will regulate all sorts of prediction markets. The same principles apply broadly: The CFTC’s job is to stop event contracts that are “contrary to the public interest,” either because they are vulnerable to insider trading or manipulation, or because they might create perverse incentives in the real world. Not only is the CFTC in the business of protecting athletes from Kalshi bettors; it is also in the business of protecting Nicolás Maduro from Kalshi bettors. “All event contracts involving terrorism, assassination, and war are highly likely to be against the public interest,” says the CFTC (page 256), and (pages 68 and 71):

An event contract that settles on whether Nicolás Maduro dies as a result of an attack by an organized political or military faction by December 31, 2026, involves assassination. The settlement-determining event—his death—is an occurrence within the assassination activity. ... 

An event contract that settles on whether Maduro is out of office by a certain date, without further specification of the qualifying mechanisms, involves assassination within the meaning of the Special Rule because assassination is among the pathways by which the settlement condition can be satisfied. The same event contract, redrafted to settle only on whether the named individual ceases to hold office “by reason of electoral defeat, resignation, constitutional removal, negotiated departure, or natural death,” would not involve assassination, because the event contract’s terms specify the qualifying pathway and exclude the Enumerated Activity pathway. Similarly, an event contract that settles on whether Iran’s uranium enrichment facilities remain functional as of a certain date would involve war, because an activity of war is among the pathways by which the facility could cease to remain standing; the same event contract, redrafted to settle only on whether the facility is demolished pursuant to a government order, or to negotiated terms of a diplomatic deal, would not.

We talked about this in March, when Kalshi decided not to pay out on “Ali Khamenei out as Supreme Leader” contracts because Khamenei was “out” due to bombs. In some sense any event contract could be affected by war — Little Joe might escape if his zoo gets bombed; Taylor Swift might not get married in New York if she’s assassinated — so perhaps all of them need war-and-assassination carveouts. 

Elsewhere, it sure seems to me like “mention markets” — bets on whether some public figure will say a particular word in a public appearance — would be contrary to the public interest under the CFTC’s reasoning, as “insight into the underlying event is highly concentrated—in a single individual, for example, or only individuals legally prohibited from transacting.” (“Will George Santos attend the State of the Union,” similarly.) They are obviously ripe for insider trading or manipulation, and don’t seem to serve much public purpose. But the CFTC doesn’t specifically discuss them, so it’s hard to be sure.

Of course one might ask a broader question about what is “contrary to the public interest”: Is sports gambling itself contrary to the public interest? But here, again, the CFTC is essentially a markets regulator, not a consumer protection regulator. One important purpose of state gaming regulation is to protect people from some of the worst effects of gambling; thus, for instance, most states require bettors to be at least 21 years old. The CFTC is just in a different position: It is not in the business of stopping people from gambling on sports markets, any more than it is in the business of stopping people from gambling on interest rates or stock indexes. [2]

Instead, the CFTC asks if there is some useful economic function served by sports betting. I have previously argued that, come on, it’s sports betting; you should be embarrassed to go around saying “actually sports betting is useful for hedging and for informing economic decisions.” The CFTC is not embarrassed (pages 94-96):

Event contracts that involve sporting events can be used for price discovery in a variety of ways. Sports teams are economic enterprises and sports stadiums are regional economic anchors that generate economic activity and materially affect both regional and national markets. For these reasons, it is economically useful to know not only how a sports team is likely to perform in upcoming games, but also how the public believes that the sports team will perform in upcoming games.Thus, the price discovery utility of an event contract on whether a team will win a game this weekend arises not simply from whether the information about that particular game can be directly tied to a specific economic decision. Rather, the price discovery usefulness of all event contracts about a team may arise from other analyses, such as how their pricing and trading volume change over time, how trading in event contracts about one team compares to trading in event contracts about other teams, and so forth. Anyone interested in that team as an economic enterprise can use information derived from those event contracts as one factor in economic decision-making on a variety of topics. Just as the corn futures market is about more than just corn, a prediction market about one sporting event is about more than just that sporting event. …

Whether event contracts can be used directly for hedging is of limited importance in the public interest determination; rather, the question is whether the information derived from event contract pricing can be used to guide hedging decisions.

In short, to take just one example, the price discovery value of event contracts on how many points a basketball player will score in a game depends on more than whether the event contracts can be used to hedge the purchase price of a ticket to the game. In some circumstances, the prices of those contracts could also, along with other information including the prices of many other event contracts, be factored into models used for commercial forecasting or audience-demand analysis, which are economic questions.

Economic efficiency demands that we have national markets on how many points Jalen Brunson scores tonight.

On the other hand, I have previously half-joked that Kalshi should start offering roulette and blackjack, [3] but the CFTC rules that out (page 112):

Games of random chance are likely contrary to the public interest The Commission preliminarily believes that event contracts involving games whose outcome depends on random chance—e.g., pure luck—are likely to be contrary to the public interest. As discussed above, prediction markets function as information aggregation vehicles, meaning their usefulness depends in part on whether market participants can bring insight, expectations, or informed views as to whether the event underlying the contract will occur. When an outcome is dictated solely by luck and cannot be meaningfully predicted, participants have no insight to contribute, leaving their forecasts without any informational value. Trading in such event contracts therefore provides no meaningful information that could support decision making or market understanding.

So there are no federally regulated online roulette markets, for now.

Conceptually, the way an initial public offering works is:

1. A company hires a bunch of banks to sell stock to investors.
2. Each bank has a bunch of salespeople who call investors and say “how much stock do you want and at what price?”
3. Each investor says something like “if it prices at $130 per share, we’ll take 2 million shares, if it prices at $135 we’ll take 1 million, but if it prices at $140 we’re out.”
4. The salespeople write this down.
5. All the salespeople’s orders get collected in one central list — the “order book” — that shows who wants to buy the stock, and how much, and at what prices.
6. After the marketing is concluded, the senior bankers and the company look at the list to see how much demand there is, and then use that demand to set the IPO price.
7. Then they go through the list and figure out how much stock everyone should get. If there are three times as many orders as there are shares in the IPO — the book is “three times covered” — then on average every investor will get one-third as much stock as they ask for, but that is just an average. A long-term buy-and-hold investor who is friendly with the company and the bankers, who won’t flip the stock and will be a supporter of the company, might get its full order filled. A flighty hedge fund might get zero. The process will be manual and involve a certain amount of negotiation between the banks and the company.
8. Eventually the list will be done and the banks will go out and confirm all the allocations with the investors.

In some ways, the SpaceX IPO will be a simplified version of this process: It is marketing its shares at a fixed price, so the order book is much simpler. In other ways, though, it is a much more complicated version of this process: It is the largest IPO ever, there are a lot of banks, and there will be a lot of orders.

At a high level, the job of a bank or brokerage is to keep lists: of how much money customers have, of how many shares they own, of how many shares they want to buy. Banks and brokerages are technology companies, and their job is to have very robust computer systems to keep these lists. When they mess up the lists, that is newsworthy. Conceptually, “write down how many shares people want, and then figure out how many shares to give to the people who want them” is not a particularly hard problem. But practically, you know, that list is so long, and so many people are updating it so frequently, and what if you mess it up? Here is a fun story from Bloomberg’s Lu Wang and Isabelle Lee about how the SpaceX IPO list-keepers are practicing keeping the list:

S&P Global Inc.’s Equity Bookbuild group, which helps underwriters capture and allocate investor demand during initial public offerings, has spent weeks expanding the efficiency and capacity of its infrastructure ahead of SpaceX’s Friday trading debut. At Depository Trust & Clearing Corp., staff plan to spend the weekend monitoring systems and communicating with industry peers.

A blockbuster IPO is not just a test of investor demand. It is also a test of the technology, communications networks and risk-management systems responsible for processing millions of orders, messages and transactions at near-instantaneous speeds. Brokers, exchanges, market makers and clearing firms must all operate in sync as orders are routed, trades executed and transactions settled.

SpaceX amplifies those demands. The company is expected to attract outsized interest from institutions, retail investors and exchange-traded funds alike. The offering has already generated more orders than shares available, according to people familiar with the matter, raising expectations that trading volumes could rank among the largest ever seen for a newly public company.

Right, I mean, on the one hand all of this is true, but on the other hand if SpaceX’s banks come to it tomorrow and say “ahhh there were just too many orders for stock so we lost track of them” that will not be acceptable.

We talked the other day about how the stock market is a financing tool again. Back in ancient history, the point of the stock market was that companies that needed money could sell stock to raise the money. But for the past few decades, that has been less true. “The public stock markets are increasingly about capital return rather than capital raising,” I wrote in 2015: Companies went public to provide liquidity to their shareholders, and then used that liquidity to buy back stock, not to sell it. Selling stock was embarrassing — diluting shareholders, etc. — and also unnecessary for the large, money-spinning and not especially capital-intensive tech companies that make up a large chunk of the US market.

And then starting in, uh, 2026, that changed. Now the large, money-spinning tech companies that make up a large and growing chunk of the US market are hugely capital-intensive; they’ve gotta build data centers for AI. The Financial Times reports:

US markets are close to ending more than two decades of declining equity supply as a trio of mega initial public offerings brings a flood of new shares that investors warn could strain the limits of demand.

The listing plans of SpaceX, Anthropic and OpenAI come as Wall Street’s existing Big Tech groups look to multibillion-dollar share sales to fund their vast AI spending, in a reversal of decades of share buybacks that have helped US stocks more than triple in price since 2016.

Goldman Sachs estimates net supply of equity in the US — measured by new shares hitting the market less equity removed by buybacks or companies going private — will be almost flat in 2026, having been in negative territory since 2003. The bank expects an even greater influx of new shares in 2027, as lock-up periods on this year’s IPOs expire.

Without the tailwind of a shrinking supply of shares, some analysts and investors worry Wall Street’s tech-led rally could finally run out of steam.

“This is a sea change,” said Ajay Rajadhyaksha, global chair of research at Barclays, referring to AI spending that is leaving little room for buybacks and turning some of the biggest companies in the US into net equity issuers.

“It will give us a clean test case for how much of the broad stock rally over the past decade has been about a net reduction in shares.”

I have sometimes written that, if you buy the broad stock market (in an index fund or whatever), you are essentially buying “your share of economic growth”: The economy grows over time, corporate profits grow with the economy, stock values go up with profits, and so your returns are linked to economic growth. I have gotten pushback on that, and in fact the relationship of stock prices and economic growth is not perfect. For instance: To the extent that the stock market shrinks every year, then your stock ownership reflects a growing share of corporate profits (and not just a fixed share of growing profits). Part of the stock market’s return in recent decades has come from its shrinking. And now it is growing again.

The first thing that you learn in investment banking is that, if you are going to pitch a client on some sort of deal, your bank is No. 1 in the league table for that sort of deal. If you are going to pitch a Ruritanian widget company on acquiring a Freedonian sprocket company for $1 billion, you are the top bank for that sort of deal. But what sort of deal? Depends! Maybe you are the No. 1 bank in global mergers and acquisitions year-to-date, or in the last 12 months, or in 2025, or in the 2020s. Or maybe you are the No. 1 bank in Ruritanian cross-border M&A. Maybe you are the No. 1 bank in Freedonian sprocket equity transactions. Maybe you are the No. 1 bank in asset sales between $850 million and $2 billion, or between $500 million and $1.1 billion. Excludes companies with controlling shareholders. Excludes financial sponsor transactions. Ranked by number of transactions. Ranked by equity value. Ranked by enterprise value. Ranked by target revenue. 

You put together a list — the league table — showing how all the banks rank on your preferred metric, with you in first place and with a bland footnote explaining the universe of deals you considered. You become expert at “cutting” the league tables to make you look good. You find some set of transactions that includes (1) more transactions you did than anyone else and (2) the proposed transaction. This can be laborious, but it shouldn’t be hard. For any reasonably big bank, there is some way to define a set that works. Your job is to find the least embarrassing set, the one with the shortest and least stupid footnote.

I don’t know if this will persuade any clients to hire you, but it is excellent preparation for life. Anyway:

President Donald Trump turned into a M&A enthusiast, with a surprise shout-out for Citigroup Inc. Chief Executive Officer Jane Fraser.

Trump touted a metric cherished by investment bankers, congratulating Citigroup for its position as the top adviser on mergers and acquisitions. It’s unclear, however, what league table the president is relying on to put Citi in the leading position. The bank’s shares briefly spiked after the post.

“Wow! CITI was ranked Number 1 in topping M&A Advisory Market by Value in Q1,” Trump posted Wednesday morning. “Congratulations to Jane F and ALL of her great people. They’ve worked really hard! BIG comeback for CITI!!!”

The presidential proclamation will come as a surprise for Goldman Sachs Group Inc., whose executives have been touting their major lead in merger-advisory rankings, consistent with data compiled by Bloomberg.

I assume he just hallucinated this, but if someone did show him a league table cut to make Citi No. 1 in M&A, that is impressive.

SpaceX Set to Put 30% of Tradeable Shares in Passive Hands, Testing Markets. Musk Looks to an Army of Loyalists to Help Make Him a Trillionaire. SpaceX IPO Draws Billions in Orders From Middle Eastern Funds. How Nasdaq Turned Its Index Into a Listings Weapon. SoftBank Attempt to Get $6 Billion OpenAI Margin Loan Stalls. Ares Management raises $8.5bn for newest fund in spite of private markets woes. Apollo hunts for Japanese life insurer to boost growth. Kalshi Plans Workplace Disclosure Rule to Combat Insider Trading. “I worry that A.I. will be to high-school-educated women what deindustrialization was to high-school-educated men.” Wyoming’s Data-Center Boom Meets the ‘Man Camp’ Backlash. ECB moved to rein in Revolut’s ‘self-guided missiles’ in Europe. San Francisco Rejects a Tax Hike on Companies With Highly Paid Executives. Judge Punishes 4 Lawyers After Catching Both Sides Using A.I. in Lawsuit. Private Equity Founder Constantino Ran Firm in ‘ Drunken Haze,’ Ex-COO Says. Wedding Inflation Has Desperate Brides Paying Witches for Perfect $100,000 Days. 

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!