The basic idea is that the US stock market will pay $2 for $1 worth of crypto. If you have a pot of crypto, you should merge it with a small US public company, because then your pot of crypto will be worth twice as much. This creates weird dynamics. I mean, it is a weird dynamic, but it creates further weird dynamics. Two are:

1. If you have a big pot of crypto, you will want a public company, so there is a good business in supplying public companies to crypto investors. (Particularly small public companies that don’t have much in the way of existing operations, so that they can easily pivot to just holding crypto.) If you have a $100 million pot of Bitcoin, merging with a public company will make it worth $200 million, so you would happily pay, like, $40 million to the people who own the public company, even if it has no value other than its public listing.
2. If you have a big pot of crypto, you should not sell it to a crypto buyer. You should sell it to the stock market. If you own 1,000 Bitcoins, those are worth $118 million if you sell them on the Bitcoin market, but $236 million if you take them public as a crypto treasury company.

On the first point, we talk a lot around here about small public companies that get gobbled up by crypto entrepreneurs so they can pivot to being crypto treasury companies. But this is inefficient and haphazard: If you want to take a pot of crypto public on the stock exchange, why should you have to find some defunct public biotech company, negotiate with its executives, strike a deal, lay off the biotech researchers, etc.? Why shouldn’t an investment bank just be in the business of supplying pristine public listings, so instead of pivoting some biotech/toy/liquor/whatever company to crypto, you can just start with a blank slate? 

Of course banks are in this business. This business — the business of supplying a publicly listed shell company — is the SPAC business (special purpose acquisition companies). In particular, Cantor Fitzgerald LP is in this business: the general business of raising SPACs, and the specific business of merging SPACs with pots of Bitcoin. We talked in April about Cantor Equity Partners Inc., a Cantor SPAC that announced a deal with Bitfinex/Tether and SoftBank to wrap their Bitcoins in a public company. The company will be called Twenty One Capital Inc., and the SPAC’s trading price currently implies about a 200% premium on the value of its Bitcoins. That’s a great trade for Tether and SoftBank. And for Cantor, which sponsored the SPAC and stands to make a lot of money on the deal.

And here’s a press release from this morning:

Bitcoin Standard Treasury Company to go Public Through Business Combination with Cantor Equity Partners I, Inc.

BSTR to launch with 30,021 Bitcoin on its balance sheet — which would be the 4th largest public Bitcoin treasury — and up to $1.5 billion of PIPE financing, the largest ever PIPE announced in conjunction with a Bitcoin Treasury SPAC merger, with the SPAC contributing up to an additional ~$200 million, subject to redemptions

BSTR, Holdings Inc. (“BSTR” or the “Company”) today announced it has entered into a definitive agreement for a business combination with Cantor Equity Partners I, Inc. (“CEPO”) (Nasdaq: CEPO) (the “Business Combination”), a special-purpose acquisition company sponsored by an affiliate of Cantor Fitzgerald, a leading global financial services and real estate services holding company. Upon closing, the combined company is expected to trade under the ticker symbol “BSTR.” …

Net proceeds will be used to acquire additional Bitcoin and to build a suite of Bitcoin-native capital-markets products and advisory services.

Sure, of course. BSTR has, let’s say, 30,021 Bitcoin. Those would be worth about $3.5 billion if it sold them, but if it takes them public they should be worth at least $7 billion. Cantor Fitzgerald is in the business of supplying public companies to pots of Bitcoin, so BSTR is combining with Cantor Equity Partners I — not to be confused with Cantor Equity Partners, which did the Twenty One deal — to take its pot of Bitcoins public. BSTR’s Bitcoins will be worth more on the stock exchange than they are as Bitcoins, and Cantor, as the SPAC sponsor, will get a cut.

I want to come back to the second weird dynamic, though: “If you have a big pot of crypto, you should not sell it to a crypto buyer.” Cantor Fitzgerald already sponsored a crypto treasury company, Twenty One. Twenty One is in the business of (1) accumulating Bitcoins, (2) having a public listing and (3) saying stuff about innovation, the future of money, Bitcoin capital markets services, etc. BSTR is in an essentially identical business. Twenty One is a pot of Bitcoins, BSTR is a pot of Bitcoins: Why do they need separate public listings? Why do investors need a choice of two more-or-less identical pristine pots of Bitcoin? (Not to mention all the other ones out there, plus MicroStrategy Inc.: I just mean that there are two big pots of Bitcoin that are going public via Cantor Fitzgerald SPACs within a few months of each other.)

You could imagine a world in which the BSTR people came to Cantor Fitzgerald and said “hi, I have a pot of Bitcoins and need capital markets advice,” and Cantor said “ah, super, we actually have a great relationship with the people at Twenty One. We took them public. They have a big pot of Bitcoins, ‘a corporate architecture capable of supporting financial products built with and on Bitcoin,’ and a business built on ‘strategically allocat[ing] capital to increase Bitcoin per share.’ They also trade at a big premium to the value of their Bitcoins. So what we’ll do is, we’ll call them, they’ll sell some stock, raise some money, and buy your pot of Bitcoins at a very nice price. How’s that sound?”

But that would be dumb, because BSTR doesn’t want to sell its Bitcoins as Bitcoins. If you sell Bitcoins on the Bitcoin market, you only get the price of Bitcoin! If you package your Bitcoins into a stock, you can sell them at a 100% premium! BSTR wants to go public itself, not sell its pot of Bitcoins to someone else. [1]

Here is how the Financial Times described this deal on Tuesday:

Cantor Equity Partners 1, a blank cheque vehicle that raised $200mn in cash in an initial public offering in January, is in late-stage talks with Adam Back, founder of crypto trading group Blockstream Capital, to buy more than $3bn in the digital currency, according to two people briefed on the talks.

The deal, which mirrors a $3.6bn crypto buying venture Brandon Lutnick struck with SoftBank and Tether in April, would advance Cantor Fitzgerald’s strategy of using publicly listed shell companies to buy bitcoin as it aims to take advantage of a surge in digital currency prices amid US President Donald Trump’s deregulatory push. … 

Back and Blockstream Capital would contribute their bitcoin in exchange for shares in the Cantor vehicle, which would be renamed BSTR Holdings.

That is, one way to look at this deal is that Cantor is “buying” Blockstream’s Bitcoins, as it “bought” SoftBank’s and Tether’s. But Cantor can buy them at a large premium: Instead of paying cash, it can pay Blockstream in shares of a (new) public Bitcoin company, which normally trade at 100+% premiums to the underlying Bitcoins. Cantor’s currency — shares of a public Bitcoin company — is worth more than cash. 

Twenty One has a similar currency (shares of itself). You could imagine a deal in which Twenty One paid for Blockstream’s Bitcoins in its stock. But at what price? The point of the Bitcoin treasury business model is that Twenty One can sell stock at a premium and use the money to buy Bitcoin, increasing its Bitcoin per share. But that’s the point of Blockstream’s model too. Blockstream wants to get paid the premium for its Bitcoin holdings, but Twenty One doesn’t want to pay the premium to acquire more Bitcoin. (It wants to “strategically allocate capital to increase Bitcoin per share.”) There is enormous value to be captured by moving Bitcoins into the public market, and everyone with a pot of Bitcoins wants that value for themselves.

Doesn’t this seem unsustainable in the long run? How are all of the Bitcoin treasury companies going to buy more Bitcoin, if every big holder of Bitcoin can make more money by starting Bitcoin treasury companies themselves? I mean, there are subscale holders; if you own 0.1 Bitcoins you’re not going to take that public so you might as well sell to Strategy or Twenty One or BSTR or whoever. And I guess eventually there will be stock-for-stock mergers of Bitcoin treasury companies. The ones that trade at lower premiums will sell to the ones that trade at higher premiums. I am excited to read the fairness opinions for those. 

Anyway, so far, the BSTR trade doesn’t seem to be working particularly well: At noon today, the SPAC’s stock was trading at around $13.93 per share, implying only about a 39% premium on BSTR’s stash of Bitcoins, well below my normal expectations of 100+% premiums. Perhaps the market for this sort of thing is finally getting a bit saturated?

There is a traditional form of financial structuring where you work at a bank, and a client comes to you and is like “I want a tradable instrument X that reflects the price of some other thing Y,” and you have to think carefully about how to link X and Y. Perhaps, you think, there could be some sort of arbitrage mechanism where holders of Y could exchange it for X, to keep the prices of X and Y in line. Perhaps you could build some basket of Y, and then X could be tradeable ownership shares of the basket. Perhaps you could call up three banks every day and ask them for bids on Y, and take the average of the bids as the daily settlement price for X. Perhaps the settlement price of X could be derived from historical levels of some non-tradable index of Y. Depends on the situation, but as a general problem it is hard. “I want a tradable instrument that reflects US home prices”: Sure, great, we all do, but which homes, and how can you make sure that the thing will accurately reflect home prices?

And then a genuinely great financial discovery of crypto is that you can do without any of that and just rely on vibes. [2]  You announce “I am launching a new token, HomePriceToken, that will reflect US home prices.” And then boring old-timers like me ask questions like “wait but how does it reflect home prices?” And you say “it just does, it’s in the name, it’s HomePriceToken, what is even the problem here?” And I say “is there some sort of arbitrage mechan—” and you say “no, shut up, it’s just HomePriceToken.”

This discovery is conventionally called “memecoins,” and I make fun of it a lot, but it really is an interesting conceptual novelty. The core ideas of a memecoin are that (1) it has a name that associates it with some other underlying thing and (2) its trading price has something to do with the underlying thing, not because of some arbitrage mechanism but because of the name. The price of Dogecoin goes up when people are thinking more about Doge, etc.

It’s a fascinating discovery because it opens the door to financialization of all sorts of things that don’t normally have any prices at all. Home values are one thing — there are complicated questions of liquidity and aggregation — but memecoins are not limited to conventional assets. Memecoins could reflect the song of the summer, the popularity of an actor, the viability of American democracy. Not in a prediction-market sort of way — not in a way that resolves based on some external fact — but just, like, within the world of the memecoin. If the democracy coin is up then democracy is up, and vice versa, and stop asking so many questions.

I am not saying that this isn’t stupid, it’s very stupid, but it’s interestingly stupid. Here’s Taylor Lorenz on Gen Z slang memecoins:

Every day, Boeshi, a 20 year old college student, scrolls through social media, looking for new words and phrases. He tracks usage of words like huzz, soyboy, baddie, or mewing, not just to use in text messages with friends, but to invest in financially. 

As new Gen Z and Gen Alpha slang words continue to go viral, an entire financialized ecosystem has emerged below the surface. Young people are investing real money into meme coins tied to slang words in order to capitalize and profit from their virality.

“These brainrot words, the more they get used, the price of the coin goes up,” Boeshi said. “The more the word gets popular, the more the coin gets popular.”

“Is there some arbitrage mechanism between the usage of the word and the price of the coin,” I interject, but then my brain rots entirely out of my head and Boeshi cheerfully ignores me.

In this emergent attention economy, virality equals monetary value. If a word is trending, its associated coin pumps. When interest drops, so does the price. “When a word is trending you’ll see that it's relative to the peak of the Google searches,” said Boeshi. “And then the drop is also relevant to the coin’s history.” Currently, there are dozens of slang word meme coins available for purchase on the alt crypto site Pump.fun.

I mean, okay. Sure. Huzz. Rizz. Skibidi. Elsewhere, here is a whole paper about memecoin manipulation by Alberto Maria Mongardini and Alessandro Mei:

Unlike utility-focused crypto assets like Bitcoin or Ethereum, meme coins derive value primarily from community sentiment, making them vulnerable to manipulation. This study presents a cross-chain analysis of the meme coin ecosystem, examining 34,988 tokens across Ethereum, BNB Smart Chain, Solana, and Base. We characterize the tokenomics of meme coins and track their growth in a three-month longitudinal analysis. We discover that among high-return tokens (>100%), an alarming 82.6% show evidence of extensive use of artificial growth strategies designed to create a misleading appearance of market interest. These include wash trading and a form of manipulation we define as Liquidity Pool-Based Price Inflation (LPI), where small strategic purchases trigger dramatic price increases. We also find evidence of schemes designed to profit at the expense of investors, such as pump and dumps and rug pulls. In particular, most of the tokens involved had previously experienced wash trading or LPI, indicating how initial manipulations often set the stage for later exploitation. These findings reveal that manipulations are widespread among high-performing meme coins and suggest that their dramatic gains are often likely driven by coordinated efforts rather than natural market dynamics.

Imagine launching one of the 17.4% of memecoins that weren’t manipulated. Seems lazy.

Elsewhere in small public companies getting gobbled up by crypto entrepreneurs, we talked yesterday about a tiny liquor company called LQR House Inc. A crypto entrepreneur named Robert Leshner spent about $2 million buying 57% of LQR’s stock in the open market, then announced his stake and his plans to “replace the board and help the company explore new strategies.” He didn’t say “crypto treasury company,” but I think at this point we all know what that means. You might think that the executives of LQR would be thrilled by this attention — presumably pivoting to crypto treasury would involve some payoff to them — but they were not.

But what could they do about it? Leshner owned 57% of the company, so he could replace the board and implement his own strategy. [3] But they had at least a little window between when Leshner bought a majority and when he actually replaced the board, and they sprang into action. Specifically, they started selling a ton more stock, to take Leshner down below a majority and make him not the controlling shareholder anymore. Leshner tweeted that he didn’t think that was productive, and I agreed with him, but I said “it is extremely cool and funny though.” As a takeover defense mechanism, just selling as many shares as possible is quite crude, but it does the trick. Also it raises a lot of money for LQR to do whatever it is doing other than pivoting to a crypto treasury strategy.

I also wrote:

In many cases the board can sell approximately as much stock as it wants, so it can always dilute you down. … The board can print new stock for free, and your money might be more limited than its ability to issue new stock.

This was approximately true, and in a footnote I qualified it in a few ways. For one thing, companies normally have only a fixed number of authorized shares, so the board doesn’t actually have unlimited ability to sell stock, though LQR’s cap on authorized shares (10 million, versus 1.06 million outstanding as of its last quarterly report) is not particularly constraining.

There is another, more technical limit that I did not mention. US Securities and Exchange Commission rules for “shelf registration statements” — the form that LQR uses to sell stock in at-the-market offerings — limit sales to “no more than one-third of the aggregate market value of the” stock, unless the company has more than $75 million of public float. [4] Again, as far as anyone knew last Friday, LQR had about 1.06 million shares outstanding, and its stock closed that day at $2.60 per share, giving it a public float of well under $3 million. That would seem to limit it to selling one-third of its stock: enough to dilute Leshner down below 50%, probably, but not enough to do it over and over again; he could buy a little more stock and get control.

But LQR filed to sell $46 million of stock, a very large multiple of its market cap. I don’t know how that works. One weird possibility is that the stock has traded up a lot since Leshner announced his involvement — it closed at $10.73 yesterday — and maybe LQR has sold enough stock to retroactively make it legal? (If there are 7 million shares outstanding now, boom, more than $75 million of float. [5] ) “I’m consulting with lawyers,” Leshner tweeted, but on the other hand what can you do about it? If LQR has sold the stock — on the exchange, to the general public, in open market transactions — it can’t really take it back, even if there are some technical legal questions. 

It is generally illegal for two widget companies to get together and agree not to sell widgets for less than $10 each. This is pretty well-known, and so companies tend not to meet like that. On the other hand, it can be mutually beneficial, for the companies, to signal to each other that they won’t lower prices: If I know you won’t lower your prices, I don’t have to lower mine, so we can both make more money.

And so there are various theories about how companies might collude to keep prices high, without meeting each other in smoked-filled rooms. One popular theory is called “algorithmic collusion,” which means roughly: Companies set their prices, their prices are public and on the internet, some computer service scrapes all of the prices, all the companies subscribe to the service, and the service tells them what price to charge (based on what everyone else is charging). Sometimes there are other elements to this, but the basic theory is that this sort of coordination kind of looks like meeting in a smoke-filled room to agree on prices.

Another popular theory is: earnings calls. Companies do earnings calls every quarter, where they talk about their business and answer analysts’ questions. The analysts ask questions like “are you still able to charge high prices for widgets,” and the companies give answers like “we have been very successful at maintaining price discipline and in our experience our competitors have also been quite rational, which is good, because price discipline is very important to us.” And then everyone can read the transcripts of those calls, and maybe the competitors do, and maybe they say “ah, they’re keeping their prices high, so we can keep our prices high.” Again, not quite a smoke-filled room, but it sort of rhymes.

One problem with that approach is that the antitrust regulators can also listen to the earnings calls, or read the transcripts, or for that matter have an artificial intelligence system read thousands of transcripts and find the ones that sound bad. Here’s a Freshfields memo about a European Commission antitrust case:

The case concerns the Commission’s 2024 dawn raid at Michelin. The underlying suspicion: several tyre manufacturers may have used earnings calls to signal pricing strategies to competitors, in potential breach of Article 101 TFEU. …

The Court … recognised as plausible the Commission’s view that certain public phrases (“we expect the industry to follow”, “we will maintain pricing discipline”) may serve a signalling function. ...

The case originated from an algorithmic screening of hundreds of thousands of earnings calls, across industries and jurisdictions.

The Commission used bigram frequency analysis to detect phrases associated with pricing intentions or competitor references (i.e., statistical screening of two-word combinations such as “price discipline” or “market leader”).

Tyre manufacturers were flagged for comparatively high densities of such expressions.

A manual review of the flagged calls followed. That qualitative screening eventually led to the dawn raid.

In the future, earnings-call scripts will be written by companies’ AI models to be interpreted by investors’ AI models, but the companies’ AI models will also have to be careful not to anger the regulators’ AI models.

The way the internet used to work was that you would write blog posts on your blog, and you would try to make them interesting and engaging for humans, and then you would include tags and metadata that were not part of the main text and that nobody read, and they would say “finance financial advice investing investment advice money make more money get money be rich,” just a sea of repetitive gibberish, for the search engines. You wanted to please your readers, so you gave them pleasing human-readable text, but it was also very important to please Google, because most of your readers came to you through Google search, and what Google wanted was not what humans wanted. When users searched for “investment advice,” Google wanted to send them to pages that offered investment advice, so a page that mentioned “investment advice” was more promising than one that didn’t.

Or that was my vague understanding of things; I was never an expert at search engine optimization, and I gather that this sort of “keyword stuffing” is now disfavored by search engines. (It is not hard to guess why: The search engines have gotten more sophisticated and discerning, and if you are spending a lot of time on this sort of spammy SEO, how good is your investment advice going to be?) But the general point is that there is some tension between what humans want and what computers want. Computers are much more patient than humans, but also more rigid. A human can read a blog that doesn’t mention the words “investment advice” and get investing ideas from it; a computer wants to see those words before classifying the blog as investment advice. But a computer doesn’t mind if “investment advice” is in a list of 50 keywords at the top of the page, while a human will see that list and be annoyed.

Anyway here’s a Wall Street Journal story about résumés:

The job seeker’s gospel commands that a résumé fit on a single page. It’s time to rethink that tenet as artificial intelligence screens more job applications.

A one-pager is designed to highlight your credentials for busy hiring managers who won’t take time to read a second page anyway. But there’s no need to cater to a human glance if a bot is going to read your submission instantaneously.

In fact, a longer résumé can increase the odds of getting through an initial review by giving you more space for the relevant words and phrases AI is trained to spot.

So you have ChatGPT read the job description and then write a résumé for you that uses all the keywords in the job description. “Demonstrated that I was a motivated self-starter by writing a blog,” etc., your bullet points will say. It can be as long as it needs to be to fit all the keywords, because it will be read by an AI too. Eventually the AI screeners will so perfectly mirror human hiring managers’ preferences that they’ll get a three-page résumé and think “ugh this person is wasting my time,” but for now, the bots don’t care that much about wasting their time, and they do care about finding the right keywords. So you should just tediously include all the keywords in your résumé, which fortunately your bot can do for you. Your AI writes the résumé, their AI reads it, and if your AIs come to an agreement then maybe you can meet in person. 

Kevin Warsh Says Fed Independence Is ‘Essential,’ But Limited. Post-Crisis Rules to Keep Banks Safe Are on the Way Out. Citi’s new banking chief steps up poaching of JPMorgan dealmakers. Morgan Stanley CEO Sees Banks Clawing Back Private Credit Gains. Texas Law Firm Heads Toward Trial Over Ex-Judge’s Romance. U.S.-Based Wells Fargo Banker Blocked From Leaving China. Substack Raises $100 Million, Betting on Subscriptions but Coming Around to Ads. “While regular-looking people fiercely compete for tables on Resy, models and influencers say they can easily book entire days of meals and experiences on Neon Coat.” Dark Web Travel Agencies Take Flight. “A natural part of the aging process—skin laxity, a sudden interest in World War II history, and a creeping sense of dread about the debt.”

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