I feel like, if you ran a big technology company, and you were looking to expand your artificial intelligence capabilities and needed to rent access to graphics processing units, and a GPU-as-a-service/AI-cloud company came to pitch you, and you said “so tell me a little bit about your company,” and the company said “well two weeks ago we were a sneaker company but we have since pivoted to AI,” you might say “huh, thanks but no thanks, we’re going to go with someone with a bit more AI experience and, you know, an actual data center.” Maybe that’s wrong; maybe the sneaker guys are great at AI. But you might worry.

But! If you said “so tell me a little bit about your company,” and the company said “well two weeks ago we were a sneaker company called Allbirds, but we have pivoted to AI,” your reaction might be different. Because, in this hypothetical, you run a big technology company, and you probably spent years wearing Allbirds. “I used to love Allbirds,” you might say; “high five!” And then you might sign a long-term cloud hosting agreement with Former Allbirds, because like many tech executives you have a nostalgic fondness for their brand.

I don’t think this is that likely, but worth a shot▸ AI summary: 

Allbirds Inc., the once-buzzy maker of wool sneakers valued at more than $4 billion in its heyday, announced a new business plan just days before it planned to close down for good: AI computing infrastructure.

And in a market that’s reacted in knee-jerk fashion to just about anything related to artificial intelligence, it did just that — sending Allbirds stock up as much as 461% after the opening bell.

The response underscores the intensity of the speculative mania around AI, which has fueled stampedes into would-be winners and panicked rushes away from any industry that seems poised to be hit by the competitive threat.

Of course there are two levels of analysis here. One is, sure, Allbirds is pivoting its business to AI compute infrastructure. That seems like a competitive and capital-intensive business in which Allbirds has no obvious expertise but, whatever, nostalgic fondness for the sneakers, maybe it’ll work out.

The other level is that Allbirds is pivoting its stock to being an AI meme stock. That definitely worked out! The stock closed yesterday at $2.49 per share, for a market capitalization of about $22 million. Allbirds previously agreed to sell its sneaker business▸ AI summary for $39 million and pay out the net proceeds to shareholders as a dividend; the $22 million market capitalization represented, roughly, the expected value of the dividend. At noon today, the stock was at $18.82, up 655% from yesterday’s close, for a market cap of about $164 million. That represents, uh, I guess it represents the expected value of the future AI-native cloud solutions business? Let’s go with that.

The key point here is the financing:

Allbirds, Inc. (Nasdaq: BIRD) (the "Company") today announced the execution of a definitive agreement with an institutional investor for a $50 million convertible financing facility (the “Facility”). The Facility, which is expected to close during the second quarter of 2026, will enable the Company to pivot its business to AI compute infrastructure, with a long-term vision to become a fully integrated GPU-as-a-Service (GPUaaS) and AI-native cloud solutions provider. In connection with this pivot, the Company anticipates changing its name to “NewBird AI.”

Allbirds/NewBird also filed a proxy statement today, laying out the convertible terms. The gist^[1]:

1. The investor will buy “up to” $50 million of convertible notes, $5 million initially and the rest later at the investor’s option.
2. The notes will have a 12% interest rate and a 5% original issue discount.
3. The investor can convert at any time at a conversion price of about $2.99.^[2] Again, the stock is at $18.82 today. A $50 million convertible — for which the investor is paying $47.5 million because of the original issue discount — would convert into about 16.7 million shares; at today’s price, those would be worth about $315 million, or considerably more than the market capitalization of the company.
4. There is also an “alternate conversion” mechanism, where the investor can convert at 93% of the lowest stock price during the 10 trading days before any conversion.^[3] We talked about this mechanism a couple of weeks ago; it is standard in micro-cap meme financing, and is sometimes rudely called a “death spiral convertible.” If the stock trades down to, say $3, and then back up to $4, the investor can convert at $2.79 per share (93% of the lowest price over the previous 10 days) and then sell the stock for $4. If the stock is volatile — if NewBird sometimes trades up a lot on AI enthusiasm, and sometimes trades down a lot on “wait, what?” — then this option is very valuable. (If the stock stays high, it doesn’t matter much; converting at $2.99 per share when the stock is at $18.82 is a great deal.)

The proxy does not disclose who the “institutional investor” is but, you know: great trade! The investor is essentially buying $50 million worth of stock at the old, defunct-sneaker-company price, and selling it at the new, AI-neocloud-company price.

Maybe! Presumably the investor can’t sell the stock yet: The convertible financing hasn’t closed, the company needs shareholder approval to allow conversions into stock, and probably selling twice the market cap of the company into the market would take time. (Though more than 200 million shares had traded by noon today, up from about 63,000 yesterday.) The investor has a lot of downside protection, but fundamentally it is betting that the new, AI-neocloud-company price will hold up for a while.

When we last talked about these sorts of convertibles, two weeks ago, I wrote:

This technology has existed for a long time, mostly for distressed companies, but in theory you would think that the meme moment would make it better. There is more upside now: If you are a small distressed public company, there is always a chance that your stock will moon and the hedge fund will make a fortune. So in theory a hedge fund should want to assemble a portfolio of these trades ... and hope that a few of them return 1,000%.

And … right? Notice that the investor in this deal is putting up very little money up front: Only $5 million is committed, and the remaining $45 million is at the investor’s option. If you had $100 million, a pretty good trade would be:

1. Go find 10 near-defunct public companies that still have stock-exchange listings, are not doing much, and ideally have some sort of nostalgic name recognition.
2. Offer each of them $50 million (10% upfront) to pivot to AI and sell you a convertible bond on very favorable terms.
3. Maybe they’ll all say yes; what do they have to lose?
4. You put up $50 million upfront, $5 million per deal you signed.
5. Each of them announces the pivot, the stock moons, and each $50 million deal is now worth $300 million on paper.
6. There is a delay — to get shareholder approvals, to close the financings, to sell the shares — in each deal, and not all of them will actually be worth $300 million by the time you are able to convert your bonds and sell your stock.
7. That’s okay: You’ve only put up 10% of the money, and don’t have to put up the rest if a deal doesn’t work out.
8. But if one works out, you close that deal for another $45 million, get a pile of stock, sell it for $300 million and triple your total investment.
9. If more than one works out, you use the $300 million from the first deal to fund the others, etc. If they all work out you make billions.

This is not investing advice and, of course, your risk is that none of the AI pivots work and none of the stocks go up long enough for you to profit. You are betting on a few more months of AI froth. Seems like a reasonable bet.

Also of course maybe some of them will actually succeed as AI neocloud businesses and you will have a good long-term investment.

Avis Budget Group Inc., the car rental company, has 35,324,685 shares of common stock outstanding. Its shares trade on the Nasdaq and anyone can buy them. But two hedge funds, between them, own 69.3% of the stock^[4]:

• SRS Investment Management LLC, run by Avis board member Karthik Sarma, owns 17,430,882 shares (49.34%).
• Pentwater Capital Management LP owns 7,048,300 shares (19.95%).

But SRS and Pentwater also have other bets on Avis’s stock. Specifically, they both own cash-settled total return swaps on the stock: They have contracts with banks providing that the banks will pay them any increase in the price of the stock, and they will pay the banks any decrease in the price of the stock.

An approximate way to think about these swaps is that the bank is providing financing to the hedge fund to buy more stock: The bank buys the stock and holds it on behalf of the hedge fund, which posts collateral for part of the value of the stock. (This is how Archegos Capital Management financed its big stock positions that went wrong a few years ago.) This is not strictly true: The swap agreement provides only for an exchange of cash flows, the bank has no obligation to buy the underlying stock, and it might hedge its swap position in any number of ways. But it is often useful and approximately correct to think that, if a bank writes a hedge fund a swap on 1 million shares of stock, the bank will hedge that swap by going out and buying 1 million shares.

Anyway here are the swaps:

• SRS owns 2,862,283 shares (8.10%) on swap.
• Pentwater owns 10,177,500 shares (28.81%) on swap.

So, economically, SRS and Pentwater between them own 106.21% of Avis’s stock outstanding. Huh! Actually the number is a bit higher: Pentwater also owns call options to buy another 775,800 shares (2.20%), for a total of 108.41%. Even that is fuzzy: Pentwater has sold some put options obligating it to buy more shares in some circumstances, and sold some call options obligating it to sell stock at fairly high prices, so you should not take this 108.41% number as exact. Also these numbers (from Avis’s proxy statement) are current as of a few weeks ago, and things might well have changed. Still:

1. It is unusual for the two biggest shareholders of a public company to own 108% of its stock. 108% is, for instance, more than 100%. Between them, SRS and Pentwater own more than all of Avis’s stock, or at least they did recently.
2. They’re not the only holders! Avis is in various stock indexes, and there are various normal index-fund holders. (Bloomberg tells me that BlackRock Inc. owns 1,625,627 shares, or 4.6%, Vanguard Group owns 1,525,988 shares, or 4.32%, and State Street Corp. owns 627,961 shares, or 1.78%.) Also active institutional investors own some, and retail investors. Just adding up SRS, Pentwater and the Big Three index investors, you get to 119.1% of the stock. 

How do people end up owning more than 119% of a company’s stock? Well, short sellers. Some people want to bet against the stock, so they borrow shares from people who own them, and sell them again. This creates new shares. Schematically:

1. A company issues 100 shares. Four investors — call them A, B, C and D — each buy 25 shares.
2. Investor X wants to bet against the stock, so she borrows 20 shares from A and sells them to B.
3. Now A still owns 25 shares (she loaned 20 out to X, but expects them back), as do C and D, while B now owns 45 shares (25 she bought from the company and 20 she bought from X). Thus people own a total of 120 shares.
4. But the books balance, because X owns negative 20 shares. There are 100 shares outstanding, and people are long a total of 120 and short a total of 20.
5. Investor Y can borrow 40 shares from B and sell them to C, etc., creating just as many shares as you want.

People get very antsy about this, worrying about “phantom shares” and “naked shorts,” but it is quite normal and legal and most of the time▸ AI summary the books all balance. (It came up a lot during the 2021 GameStop Corp. saga, because in that case short interest exceeded 100% of shares outstanding▸ AI summary — unusual and antsy-making, but again perfectly legal and straightforward.)

Still it might make you nervous, not so much if you are a citizen or regulator or Avis shareholder, but if you are an Avis short seller. To short the stock, you had to borrow it, probably from your bank or broker, who borrowed it from someone who owns it. Stock borrow is usually “open term,” meaning that the owner can demand that you return it any time. Who owns the stock that you borrowed? I mean, probably BlackRock or Vanguard or State Street or a retail investor with a margin account. But if the two big Avis shareholders own 108% of the stock, in some loose probabilistic sense they own the shares you borrowed. If they converted their 108% partly-synthetic position into all stock, and stopped lending it out, you would have to buy back stock to return to them. Who would you buy it from? Well, they own 108% of the stock. You’d buy it from them. How much would you have to pay them? Well, whatever they wanted to charge.

This is not quite right — BlackRock will keep owning stock, etc. — but it is a bit alarming. The supply of stock to borrow is constrained and risky. I talk from time▸ AI summary to time▸ AI summary around here about my favorite short squeeze, when Phil Falcone of Harbinger Capital Partners bought more than 100% of an issue of bonds called the “MAAX zips.” Some banks were short the bonds, and Falcone called in all of the borrow, forcing them to buy back the bonds. But there were no sellers, because he owned all the bonds. Then he conducted the coolest phone call in the history of Wall Street:

At some point, the conversation turned to the trading in the MAAX bonds. The senior officer asked Falcone how the Wall Street firm might satisfy its obligation to Harbinger. Falcone stated that the Wall Street firm should just keep bidding for the bonds. Falcone acknowledged that the Wall Street firm would suffer some losses doing so, but told the senior officer and the others that sometimes you are just on the wrong side of a trade.

In the course of this discussion, Falcone stated that he knew that the short position in the MAAX zips had created a “long” position in excess of the issue size. When the senior officer asked how he could possibly know this, Falcone stated that he was working the position himself and that he (i.e., Harbinger) had acquired approximately 190 million bonds. The senior officer and the other the Wall Street firm personnel were stunned.

Again, this is not quite possible in the Avis case, but it is, uh, an interesting precedent.

Anyway last week Bloomberg’s Sridhar Natarajan and Hema Parmar reported▸ AI summary:

Avis shares soared more than 150% in the past three weeks after Pentwater Capital Management said it acquired a sizable stake in the rental-car company, which already has an outsized shareholder in SRS Investment Management.

That’s making it hard for short sellers to continue borrowing the stock, and the cost of doing so has skyrocketed, according to financial analytics firm S3 Partners. S3 Managing Director Ihor Dusaniwsky says he is watching to see whether traders betting against the shares will keep holding on or cut their losses.

“This is a stock that has been short squeezed,” Dusaniwsky said. “This is like going to a Nascar race. The crashes sure are really interesting.”

Bearish investors have shorted about 43% of the tradeable shares of Avis as of Wednesday, down from 49% in mid-March, according to S3.

That was last Wednesday, when Avis closed at $261.35 per share, up 162% since March 20. Yesterday it closed at $411.56, up another 57% in four trading days and quadruple where it was a month ago. (It’s down a bit today.) It is hard to be short a stock when its main holders own more than 100% of it.

Incidentally! I mentioned that SRS and Pentwater, between them, own about 36.91% of the stock via cash-settled swaps, and I wrote that, while often banks hedge swap positions by buying the underlying stock, “the bank has no obligation to buy the underlying stock, and it might hedge its swap position in any number of ways.” The banks are effectively short stock to SRS/Pentwater, so the natural hedge is to get long by buying stock.

But there are other natural hedges. For instance, if you are a bank, and you have hedge fund customers who are short Avis stock, and you are also writing swaps to SRS or Pentwater, there is an obvious trade: Write swaps both ways. The short sellers go short on swap to you, SRS/Pentwater goes long on swap with you, and you just sit in the middle. SRS and Pentwater have cash-settled bets that the stock will go up, the short sellers have cash-settled bets that the stock will go down, and nobody needs to trade any stock at all.

These days I often find myself drawing distinctions between traditional financial markets, where assets are valued based mostly on their expected future cash flows, and weird postmodern financial markets that are all, like, memes and sports bets. But even in traditional financial markets, sometimes what matters is not the underlying business but the structure of the bets.

We have talked a couple of times this week about correlation in sports betting. Basically I have said that sports results tend not to be functions of some set of underlying fundamental factors, and that the outcome of one game is unlikely to tell you much about the outcome of a different game. Of course there are exceptions. Several readers emailed to mention end-of-season playoff/seeding situations: If Team X loses in the afternoon, guaranteeing Team Y a playoff spot, Team Y might rest its starters and lose the evening game. And one reader emailed this anecdote:

[A sportsbook trader] once told me that the thing that [troubled] them the most was weather in golf tournaments. They would have some bet like "pick 5 players that will shoot under/over par" and then some shark would come just blast them with under par bets when it wasn't supposed to rain at night (according to public predictions) but did (so the greens were softer and everyone putted better). He actually tried to look in to weather futures to hedge their risk but it was too complicated since the golf tournaments changed location every week.

Yes, fine, weather is probably going to be a generally applicable factor in your sports betting model.

A good heuristic is that if someone is selling a $1,000 online course teaching you to make a lot of money by trading stocks, that person is not making a lot of money trading stocks. If she were, she would be too busy making money trading stocks to bother teaching you.^[5] This is true of people selling $1,000 online courses teaching you to make a lot of money by ______, for most of the things you might put in the blank. But not all of them! For instance, a person selling a $1,000 online course teaching you to make a lot of money by selling $1,000 online courses is, in a sense, proving out the concept. If you’re paying $1,000 to a person for an online course in how to sell $1,000 online courses, you know that she is successfully selling $1,000 online courses. At least to you.

Anyway at Business Insider Emily Stewart reports:

Want to build a business? Get more TikTok followers? Learn to flip Goodwill finds? There's almost certainly an online class for that — or, likelier, dozens. An estimate from Grand View Research puts the global digital education market at $134 billion in revenue by 2030. Companies that let people create courses online, such as Kajabi and Squarespace, are worth billions of dollars. Newsletter publishing platform Beehiiv is building out ways for its writers to sell educational products. It seems like half the celebrities are teaching something on Masterclass. There are even a plethora of courses that teach you how to … make courses.

There's an extremely low barrier to entry in creating and selling online courses, which is good in that plenty of people have knowledge to share, but less optimal in that on the internet, "expertise" can simply be code for "confidence." The space is broadly unregulated and unsupervised. 

I assume there are courses on how to sell courses on how to sell courses, really infinite recursion is possible here.

Anthropic Attracts Investor Offers at an $800 Billion Valuation. Hedge Fund Leverage Powered by a Few Key Banks Sparks Concerns. Morgan Stanley Stock Traders Join in Wall Street’s Windfall. Hedge Funds Built for Crypto Turn to Oil and Gold on 24/7 Venues. Silver market seen as ripe for another price squeeze. BofA Discloses $20 Billion of Loans to Private-Credit Firms▸ AI summary. Pimco Buys All $400 Million of Bonds Sold by Blue Owl BDC▸ AI summary. US banks make record buybacks▸ AI summary on Trump’s looser rules and choppy markets. Atkins Faces Ticking Clock as He Reshapes Rules▸ AI summary for Wall Street. ‘Bifs’ replace ‘Piigs’ as Europe’s bond market▸ AI summary whipping boys. Stock-Clearing Firm DTCC▸ AI summary Taps Amazon to Aid Shift to Cloud. ICE coffee grading▸ AI summary. SantaCon Organizer▸ AI summary Charged In Wire Fraud Scheme Targeting Attendees And Host Venues.

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!