Money Stuff: Coinbase Said Web3

The main point I want to make here is that this is all so dumb and I hate it. Kalshi, a commodities futures exchange registered with the US Commodity Futures Trading Commission, offers contracts on various unconventional commodities like election outcomes and football games. Last month you could buy futures contracts on commodities like a Coinbase Global Inc. representative saying the words “prediction market” or “Bitcoin” or “Web3” on its third quarter earnings call. Why? I don’t know. To hedge your risk that Coinbase wouldn’t say those words? To help people understand and price the future states of the world in which Coinbase did or didn’t say those words? Because you were bored and liked to gamble, probably.

At 9 a.m. last Thursday, “Bitcoin” was trading at about a 92% probability. Coinbase is a crypto exchange, Bitcoin is the biggest cryptocurrency, that makes sense. “Prediction market” was trading at about 71%. Prediction markets are hot right now and you might have guessed that Coinbase was thinking about them. [1] “Web3” was trading at about 15%. “Web3” is definitely a thing that crypto people said a lot in, like, 2022, but not so much recently. These probabilities all seem pretty reasonable.

The earnings call happened at 5:30 p.m. on Thursday; you can listen here, or here is the Bloomberg transcript. Near the end of the call, Coinbase Chief Executive Officer Brian Armstrong said this:

I hope we answered your question on that. I was a little distracted because I was tracking the prediction market about what Coinbase will say on their next earnings call. And I just want to add here the words Bitcoin, Ethereum, Blockchain, staking, and Web3 to make sure we get those in before the end of the call.

Okay? Sure. I don’t know. Total volume on the what-will-Coinbase-say Kalshi market seems to have been about $80,000, so you could have made, you know, probably tens of dollars on “Web3” when he said that. But he sure did say that.

Traditionally commodity futures markets reference commodities. There is a lot of real-world economic supply and demand for, like, wheat. If you bought a lot of wheat futures, you would want the price of wheat to be high at the expiration of the contract, because then you would make more money. But in the general case it would be hard for you to make the price of wheat high. There’s a lot of wheat in the world, and a lot of people who use it, and you can’t do that much to change the price. [2]

Every so often this situation breaks down: Sometimes there is a derivative market (a commodities futures contract) that is much bigger than the underlying market (for the commodity), so you can make a lot of money on the derivative by spending a little money to manipulate the underlying. We have talked about allegations that Jane Street manipulated the Indian options market; an essential element of those allegations is that the Indian options market is much bigger than the market for the underlying Indian stocks, so you could manipulate the options by buying the stocks. [3]  

But when the commodity derivative is “Brian Armstrong will say some words in public,” I mean … ? The Kalshi what-will-Coinbase-say market was not big in dollar terms. But the market for the underlying commodity was infinitesimal. It cost Brian Armstrong approximately nothing to say “Web3” at the end of his earnings call. (Maybe it cost him a little bit in terms of, like, lawyer heartburn; maybe it benefitted him by pleasing prediction-market gamblers. [4] ) He could just say that! And then the settlement values of those commodity futures would change.

The Commodities Exchange Act makes it illegal to “manipulate or attempt to manipulate the price of any swap, or of any commodity in interstate commerce, or for future delivery on or subject to the rules of any registered entity.” The standard definition of commodities market manipulation requires proof that “(1) Defendants possessed an ability to influence market prices; (2) an artificial price existed; (3) Defendants caused the artificial prices; and (4) Defendants specifically intended to cause the artificial price.” [5] An artificial price is one that “does not reflect basic forces of supply and demand,” one that is not determined “by free competition alone.” [6]

Was the final settlement price of the “Brian Armstrong will say ‘Web3’ on an earnings call” contract … artificial? Did he manipulate that price? If you had given him $5 and said “hey can you say ‘Web3’” and he was like “sure whatever I don’t care,” would that be manipulation? If you had not given him $5, but just pointed him to the prediction market prices and said “lol wouldn’t it be funny to say ‘Web3,’” would that be manipulation? If he had bought “Yes” contracts before saying it, would it be manipulation? 

I have no idea. I think the questions are incoherent. I don’t know what it could mean to manipulate the market for Brian Armstrong saying words. [7] If you and your buddy have a $20 bet on whether or not your boss will say some words in a meeting, and your buddy prods her into saying the words, and she says the words and your buddy wins the bet, I think you would probably be like “ahhh you got me” and hand over the $20. Your buddy would be kind of cheating, but not really. That is the sort of dumb bet that you and your buddy make because you are bored in a meeting and like gambling; in that context, a little market manipulation is a perfectly acceptable way to spice things up. 

Kalshi poses the question: What if dumb bets that you and your buddies make because you are bored and like gambling were regulated by the CFTC? I don’t think there are good answers for anything here. Bloomberg’s Emily Nicolle and Justina Lee write about this stunt:

Under the CFTC’s core guidelines, regulated platforms aren’t supposed to list contracts that are “readily susceptible to manipulation.” Mention markets may now test the limits of that definition, though rules surrounding this new era of prediction markets are still evolving.

“I’m hopeful that we see additional guidance from the CFTC and maybe a comprehensive regulatory framework that addresses these kinds of issues as they come up,” said Andrew Kim, partner at Goodwin Procter LLP. “A core question that needs to be answered first, is there something to regulate here? Or is this a flash in the pan feature that sounds nice in theory but doesn’t work in execution and you move on to the next thing.”

Right, the ideal answer — aesthetically and probably legally — is neither “this sort of thing is prohibited” nor “this sort of thing is fine” but rather “please stop talking to me about markets in what Brian Armstrong will say on a call.” What Brian Armstrong says on an earnings call is not intuitively a commodity, and having it trade on a regulated commodity exchange raises questions that are both hard to answer and also dumb.

Obviously this doesn’t matter. Again, the market for these derivatives was tiny. On the other hand, this was an earnings call for a $90 billion public company that reported $433 million of net income on $1.8 billion of revenue last quarter. Like, real business is happening here, and the Coinbase earnings calls was a way for investors and analysts to get some information about the fundamentals of the business that they are allocating capital to. Presumably people and algorithms used what was said on the call to inform how they traded Coinbase stock, and perhaps the stocks of other companies, or cryptocurrencies, or other financial instruments whose expected future cash flows changed based on what Coinbase said about its business. And then also, entirely separately, there were some prop bets on Coinbase saying “Web3.”

Everything, I increasingly find myself saying, is sports betting. Within just the past year, US financial markets have caught the sports-betting bug. You can bet on sports in your brokerage account — it’s an “emerging asset class” — and that behavior has infected the rest of financial markets. An earnings call is an opportunity to learn fundamental information about a real business, but you know what would make it more exciting? Having a little action on what words the company might say.

We talked on Thursday about a cool M&A deal. Metsera Inc. is a pubicly traded biotech company that, in September, agreed to sell itself to Pfizer Inc. for $47.50 per share in cash plus up to $22.50 in contingent value rights. Last week, it announced that it had a better offer from Novo Nordisk A/S, which would pay $56.50 per share plus up to $21.25 in CVRs. The Novo deal has a higher price, but also much more antitrust risk than the Pfizer deal: There is very little overlap between Pfizer and Metsera, but a lot between Novo and Metsera, so antitrust regulators are fine with the Pfizer deal but might block the Novo deal.

Novo addressed this problem by paying for the Metsera shares before the deal closes: It will give Metsera $56.50 per share in cash (about $6.5 billion total) when it signs the merger agreement, and Metsera will dividend that cash out to its shareholders. (They only get the $21.25 CVR if the deal closes.) In exchange, Novo will immediately get a non-voting convertible preferred stock giving it 50% economic ownership of Metsera. If the deal gets through antitrust review, Novo will acquire all of Metsera. But if the deal gets shot down for antitrust reasons, the Metsera shareholders will keep the cash and Novo will keep the preferred stock; if Metsera ultimately sells itself to someone else, it will have to pay back the $6.5 billion in cash before its shareholders get anything.

One way to think about this structure — and the way I wrote about it on Thursday — is that it is a clever allocation of antitrust risk. Novo wants to buy Metsera, Metsera wants to sell to Novo, but there is some uncertainty about whether US antitrust regulators will allow the deal to go through. Novo thinks it will. Metsera is worried, and does not want to abandon Pfizer’s sure-thing bid for Novo’s higher but riskier offer. So Novo has found a way to take all of the antitrust risk on itself: Metsera’s shareholders get paid no matter what, and if the deal gets blocked for antitrust reasons, that’s Novo’s problem. 

But there is another, more cynical way to think about it. The more cynical view is:

• Novo and Pfizer are giant global pharmaceutical companies.
• Novo is a big player in the weight-loss-drug space: It makes Ozempic and Wegovy.
• Pfizer is not a big player in weight loss, and in August it discontinued trials of its weight-loss drug.
• Metsera has a promising weight-loss drug.
• If Pfizer buys Metsera, it will be a big player in the weight-loss-drug space, and a strong competitor to Novo.
• If Novo buys Metsera, it will be even more dominant in weight loss.
• But if nobody buys Metsera, that’s just as good for Novo: Metsera is unlikely to commercialize its weight-loss drug on its own, so if nobody acquires it then the drug wil not be a viable competitor for Ozempic and Wegovy. In particular, Pfizer — a giant drug company with no weight-loss drug of its own — would be a competitive owner of Metsera, with strong incentives to commercialize its drug. Stopping Pfizer from getting Metsera would be very good for Novo’s market power.
• This deal structure might lead to Novo owning Metsera, sure, anything’s possible. But given the competitive dynamics it probably won’t.
• But this deal structure will stop Pfizer from owning Metsera: It will tie Metsera up for years in antitrust review; Novo might never end up owning it, but the important thing is that Pfizer won’t either.

On this view, Pfizer needs a weight-loss drug and Novo already has a weight-loss drug, so Novo doesn’t need another one, but it does want to stop Pfizer from getting one. And on this view, Novo doesn’t actually care about closing the merger and acquiring Metsera: It is willing to spend $6.5 billion just to stop Pfizer.

The more cynical view is, of course, Pfizer’s view. Bloomberg’s Madison Muller reports:

Pfizer Inc. accused Novo Nordisk A/S of trying to stifle competition in the weight-loss market by attempting to acquire obesity startup Metsera Inc., the second lawsuit in four days as Pfizer tries to retain its grip on a deal that Novo upended last week.

In a statement on Monday, Pfizer said it filed a new lawsuit against Novo and Metsera on antitrust grounds, saying the company’s taking action to “stop Novo Nordisk from illegally paying off Metsera and its controlling stockholders to gain control of, and impair and potentially kill, an emerging US competitor.”

Here is the first lawsuit, filed against Metsera in Delaware Chancery Court on Friday to try to stop Metsera from ending its merger agreement with Pfizer. Right now, Metsera still has a signed binding merger agreement to be acquired by Pfizer. The merger agreement allows Metsera to get out of the deal if it gets a “superior proposal” from someone else, and Metsera has announced that the Novo deal is likely a superior proposal. But Pfizer’s argument is that (1) the Novo deal is illegal or at least very unseemly and therefore (2) it is not a “superior proposal” that Metsera should be allowed to accept. From the complaint:

The Novo Nordisk proposal, reflected most recently in presumptive deal documents sent to Pfizer yesterday, involves a contrived scheme designed to evade antitrust review, breakup the Pfizer deal, and kill a nascent competitor in violation of the antitrust laws.

It works like this: Novo Nordisk will pay more than $6.5 billion in cash to Metsera, on signing, in exchange for Metsera’s termination of the Pfizer deal and newly issued non-voting economic interests in Metsera representing 50% economic ownership of the company. It is contemplated that Metsera will then immediately dividend that cash to its stockholders. All of this will happen within days and without a stockholder vote or regulatory approvals.

In an announced but illusory “second step,” Novo Nordisk would also acquire Metsera’s voting securities and provide a contingent valuation right (“CVR”), if the deal survives regulatory scrutiny and closes — in other words, never, given that Novo Nordisk’s dominant position in the market will prevent antitrust approval. An extraordinarily long outside date, more than two years from now, will lock up the company in the meantime. During this interim period, Novo Nordisk would fund Metsera’s ongoing operations in exchange for a debt instrument, rendering Novo Nordisk not only Metsera’s largest stockholder, but also its largest creditor, tightening its grip over Metsera’s competing products. And even if the rest of the Novo Nordisk proposal fails to pass regulatory review, Metsera and its stockholders get to keep the initial $6.5 billion payment. 

At its core then, Novo Nordisk’s proposal is nothing more than an old-fashioned bribe of Metsera and its key stockholders to not compete against Novo Nordisk, and prevent its pipeline product from reaching the market, to the detriment of Pfizer and ultimately consumers that are dependent on these life-changing drugs.

The second lawsuit was filed in federal court against Novo:

The lawsuit asserts that Novo Nordisk’s recent proposal to acquire Metsera constitutes an anticompetitive action by Novo Nordisk to protect its dominant market position in GLP-1s by capturing and killing a nascent American competitor before it gains the support of Pfizer, one of America’s leading pharmaceutical companies. The lawsuit alleges that Novo Nordisk’s proposed transaction violates Section 7 of the Clayton Act because of the anticompetitive effects it would have in the GLP-1 drug markets to the detriment of millions of Americans who suffer from obesity, diabetes, and other metabolic conditions, that it constitutes an anticompetitive conspiracy between Novo Nordisk and Metsera in restraint of trade in violation of Section 1 of the Sherman Act, and that it constitutes attempted monopolization and conspiracy to monopolize under Section 2 of the Sherman Act. 

Ordinarily, antitrust laws prevent companies from closing anticompetitive mergers. You sign a merger agreement, you submit it for regulatory review, and the antitrust regulators either say “yes that’s fine” or “no this merger would make you too big.” If they say it’s fine, you close the deal; if they say it’s not, you walk away. The bad antitrust effect would come from closing the deal, so regulators get a chance to prevent that. But Pfizer’s argument is that, in this case, the thing that is bad for competition is not Novo acquiring Metsera, but rather Novo breaking up the Pfizer deal, so even if the Novo never ends up getting Metsera it is still an antitrust problem.

We also talked on Thursday about Elon Musk’s pay package at Tesla Inc. Basically Musk has about 15% of the voting stock of Tesla, and he’d like to have 25%. If he doesn’t get at least a path to 25%, he might quit. To get him to 25%, Tesla’s board of directors has decided to give him a pile of restricted stock, contingent on meeting various operational and stock-price targets, with a headline value of $1 trillion. Seems high.

I argued on Thursday that it would be much better to just give him more votes, without giving him more economic ownership. Give him extra-voting stock to get to 25%, boom, done. Is this possible? I dunno; it raises obvious conflict-of-interest problems, but it seems to me that it’s strictly better for shareholders than giving him $1 trillion of stock, so it ought to be possible.

But it probably isn’t. Law professor Ann Lipton emailed me to point out that, whatever state corporate law might say, the actual constraint on Tesla is stock exchange listing standards. Nasdaq, where Tesla is listed, has a rule that says:

Voting rights of existing Shareholders of publicly traded common stock registered under Section 12 of the Act cannot be disparately reduced or restricted through any corporate action or issuance. Examples of such corporate action or issuance include, but are not limited to, the adoption of time-phased voting plans, the adoption of capped voting rights plans, the issuance of super-voting stock, or the issuance of stock with voting rights less than the per share voting rights of the existing common stock through an exchange offer.

Not only can Tesla not create a new super-voting stock and give it to him, it also can’t create a new non-voting stock and ask other shareholders to exchange into it. It is tempting to propose something like “you dividend one share of non-voting Class B stock for every share of existing stock (like Google did in 2012), and then Elon Musk does a public exchange offer where he offers to exchange 1.05 shares of his new pile of non-voting stock for 1 share of other holders’ voting stock, [8] and Tesla’s extremely Musk-friendly shareholder base tenders enough shares to get him there,” but even that seems to violate at least the spirit of the rule. [9]

And it’s hard to get out of this. The New York Stock Exchange, the other main US listing exchange, has a similar rule. Even the Texas Stock Exchange’s listing standards say that “Voting rights of existing Shareholders of publicly traded common stock registered under Section 12 of the Act cannot be disparately reduced or restricted through any corporate action or issuance”; this seems to be required of US exchanges. “We searched high and low” for a way to give Musk more votes without giving him more economic value, Tesla board chair Robyn Denholm told the Financial Times, but they didn’t find one, and the listing rules seem to be the main impediment.

Let me try one more. Maybe create new non-voting stock and do a 9-for-1 dividend of the new stock, so that each shareholder now has nine non-voting shares for every one voting share and the voting shares represent 10% of the value of the company. And then give Elon Musk a new pay package consisting entirely of voting stock, representing 12% of the votes, but only 1.2% of the economic value of the company. That saves shareholders $900 billion! But, again, seems to violate at least the spirit of the rule.

Two points here:

1. Hahaha what if Musk had taken Tesla private in 2018, when he said he was thinking about it? In a lot of ways Tesla perhaps should have dual-class stock to entrench Musk’s control forever, but that’s hard to do with an already-public company. But you go private, you zhuzh up the share classes, you go public again … I don’t know, seems too late now, but something to think about.
2. If you do have a really good way to get Elon Musk 25% of the votes without giving him more stock, do let me know! 

In 2004 and 2005, roughly one-third of US mortgages were adjustable-rate loans, with a low initial rate that would adjust after a few years to a (potentially much higher) floating rate. Adjustable-rate mortgages made it possible for people to buy houses that they couldn’t sustainably afford: They could make the low initial payments for a few years, but they probably would not be able to make the higher floating payments. The theory at the time was roughly that, in two years, your home’s price would be way up and you could refinance the mortgage and avoid the problem. This theory eventually stopped working and there was a mortgage-related financial crisis in 2008 for which ARMs took some of the blame. 

Since then ARMs have been considerably less popular. Why? Well, one set of theories might include “banks remember 2008 and don’t want to do that again,” or “borrowers remember 2008 and don’t want to do that again,” or “bank regulators remember 2008 and don’t want to do that again.” But I suppose another possible theory is “after 2008 there was a fall in house prices followed by a long period of low interest rates, so nobody needed ARMs, since the whole point of an ARM is to make it possible to finance an expensive house with a lower interest rate than a fixed-rate mortgage.” And now houses are expensive and interest rates are higher so, boom, ARMs are back:

The unaffordable housing market is causing a growing number of home buyers to take on a type of riskier loan to cut their borrowing costs. 

They are opting for adjustable-rate mortgages, or ARMs. These loans initially offer cheaper borrowing rates compared with a fixed-rate mortgage. But ARMs reset, usually after three to 10 years, which can saddle borrowers with higher monthly payments if mortgage rates have risen over that time. ...

About 10% of purchase-mortgage applications were for ARMs in the week ended Oct. 3, the highest rate since 2023, according to the Mortgage Bankers Association. In early 2021, when mortgage rates were near historic lows, less than 3% of purchase applications were for ARMs.

Stereotypically if you got an ARM in 2005 you were betting on house prices going up, but if you get an ARM in 2025 you’re betting on mortgage rates going down.

One aspect of the modern artificial intelligence boom is that the hottest AI-linked companies — particularly OpenAI and Nvidia Corp. — can make other companies’ stocks go up. The market assumes that there will be big winners from the AI boom, that OpenAI and Nvidia are among the likely winners, and that they have the power to create other winners. If OpenAI or Nvidia announces a commercial relationship with or an investment in some other company, that company is a winner and its stock goes up. Sometimes this fact can be exploited for financial gain: OpenAI bought some chips from Advanced Micro Devices Inc. that were kind of paid for by AMD’s stock-price rally when the deal was announced. Mostly though it is just a pleasant accident, the immense enthusiasm for Nvidia and OpenAI spilling over onto everything they touch.

It would be very funny if Nvidia announced a $10 billion commercial deal for fried chicken but no:

Jensen Huang’s Midas touch extends to Korean fried chicken.

Photos and videos of Nvidia Corp.’s CEO having beers and fried chicken at a local restaurant, Kkanbu Chicken, in Seoul — with Samsung Electronics Co. Chair Jay Y Lee and Hyundai Motor Co. Executive Chair Chung Euisun — went viral, and now investors are driving up the stocks that may be a beneficiary.

While Kkanbu is not publicly listed, shares of rival Kyochon F&B Co. briefly surged as much as 20% on Friday after photos and videos of the gathering circulated widely on social media. Korean poultry processor Cherrybro Co. also soared by the daily limit of 30%, with trading volume about 200 times its average. Neuromeka Co., a Kosdaq-listed company that makes chicken-frying robots, also jumped.

I mean! The AI boom’s appetite for chips and power and data centers is massive, and if OpenAI announces a deal to buy billions of dollars of chips from AMD then AMD’s stock kind of should go up. Jensen Huang’s appetite for fried chicken is … look, I don’t know him or anything, maybe he loves fried chicken more than anyone else on earth, but he is one guy, there is only so much chicken he can possibly eat. Jensen Huang is not going to materially increase the revenue of a Korean poultry processor. Nevertheless.

Trump Battles Tiny Toymaker Over Tariffs in Landmark Supreme Court Case. Amazon Inks $38 Billion Deal With OpenAI For Nvidia Chips. Microsoft Signs $9.7 Billion Deal With Data Center Firm IREN. Kimberly-Clark to Buy Tylenol Maker Kenvue for $40 Billion. Shutdown Stalls SEC Work on Private Credit in Retirement Plans. What next for Andrea Orcel’s UniCredit? Anthropic, AWS Give Customers of AI Agents a Helping Hand. Green Investors Enjoy Huge Returns as Stock Market Powers Through Trump’s Attacks. Short Seller Andrew Left Says Hedge Fund Lied to SEC About Payment in Trading Probe. Nvidia CEO Jensen Huang Completes $1 Billion Share Sale. An Unlikely Business Trend: Men’s Vulnerability Groups at Work. Sam Altman is trying to get his deposit back for a Tesla Roadster he ordered in 2018.

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!