I think a lot about the time that Mark Zuckerberg was hauled before Congress to answer questions about whether Facebook (now Meta Platforms Inc.) is a monopoly. He said:

We face a lot of competitors in every part of what we do, from connecting with friends privately to connecting with people in communities to connecting with all your friends at once to connecting with all kinds of user-generated content. … Congressman, the space of people connecting with other people is a very large space.

“See,” I wrote, “Facebook also has a tiny market share, if you define its market as … human interaction? Human behavior? ‘Only like 5% of human life consists of Facebook, how could we possibly be a monopoly?’” 

Of course, when a chief executive officer is testifying in a congressional antitrust hearing, he will want to downplay the possibility that his company is a monopoly. When he leaves the hearing and goes back to his office, the incentives will reverse, and he will work to seize the biggest possible share of his market. If his market is all human interaction, there’s a lot of work to do. I went on:

When Facebook says “we’re not a monopoly, there’s lots of competition in the connecting-with-people space, some of our most fearsome competitors include Having Dinner With Friends and Reading Stories to Your Children,” you could interpret that as a threat. Facebook is good at disposing of competitors! Maybe it will copy enough features from Reading Stories to Your Children so that people will abandon it for a Facebook product, Instagram Reads Stories to Your Children or whatever. Maybe it will acquire Having Dinner With Friends so it can merge it into the Facebook experience. These companies are where they are because they dominate their markets; if they define their markets as the world, then the world had better watch out.

That was five years ago. The Wall Street Journal reports:

Mark Zuckerberg wants you to have AI friends, an AI therapist and AI business agents.

In Zuckerberg’s vision for a new digital future, artificial-intelligence friends outnumber human companions and chatbot experiences supplant therapists, ad agencies and coders. AI will play a central role in the human experience, the Facebook co-founder and CEO of Meta Platforms has said in a series of recent podcasts, interviews and public appearances.

“I think people are going to want a system that knows them well and that kind of understands them in the way that their feed algorithms do,” Zuckerberg said Tuesday during an onstage interview with Stripe co-founder and president John Collison at Stripe’s annual conference. 

Zuckerberg said on a podcast last week that he thinks the average person wants to have more friends and connections with other people than they currently do—and that AI friends are a solution. …

“The average American I think has, it’s fewer than three friends, three people they’d consider friends, and the average person has demand for meaningfully more, I think it’s like 15 friends,” he said in the interview with podcaster Dwarkesh Patel. 

The total addressable market for friends is 15 per person, and right now humans are only providing about three friends per person, which leaves an opening for Meta to capture as much as 80% of human (well, human-ish) interaction. [1]  That would be pretty good.

The basic trick in distressed debt is this:

1. A company borrows $100 from a bunch of different lenders.
2. Time passes, and eventually the company, for reasons, would prefer not to pay back the $100.
3. It goes to a group of lenders holding 51% of the debt ($51) and offers them a deal: The company will pay them back 110 cents on the dollar ($56.10), if it can pay back the other lenders (who hold 49%) zero dollars.
4. The majority lenders say “sure, $56.10 is better than $51.” 
5. The company puts the matter to a vote of its creditors, and 51% of them vote in favor of the deal, so it passes.
6. The majority lenders get their $56.10; the minority lenders get $0.
7. The company has saved $43.90; the majority lenders have made a $5.10 bonus; the minority lenders have lost $49.
8. Everyone is mad for a little while.
9. Eventually they forget and you do it again.

This is called a “liability management exercise,” or LME, and we have been talking about the basic idea for years. My schematic outline above is too cute; you can’t exactly do that. The actual trick is in figuring out what you can do. Most credit agreements and bonds will allow the company to do some stuff with the approval of a majority of the lenders, and other stuff only with the approval of every lender. Cancelling the debt is in the latter category: A company can’t literally zero 49% of the debt with the approval of the other 51%. Changing the maturity or interest rate is usually also off limits.

But it can make the debt pretty unpleasant for the minority holders. Often, for instance, the credit agreement will say “the company can’t issue any debt that is senior to this debt,” but that can be amended by a majority vote. So you get 51% of the creditors to agree to allow new more senior debt, and then you give them the senior debt: They vote to allow new debt that is senior to the old one, and they exchange their $51 of old debt into $51 of new senior debt. The remaining $49 of old debt is now worth less, since it is behind the new $51 senior debt in line for repayment. Perhaps the majority creditors give the minority an extra kick on their way out the door, stripping out covenants from the old debt so the minority creditors have fewer legal rights. Depending on how troubled the company is, that $49 of old debt might trade down to pennies on the dollar. (And then, if the company is feeling really feisty, it can buy it back cheap. [2] ) The company can’t really zero the minority debt, but it can shift a lot of value from the minority to the majority, and the majority creditors might be willing to pay the company for that value. Generally they pay in the form of new financing: The company can borrow more money and keep itself going for a while longer, because the majority lenders will kick in more money in exchange for seniority on their old money.

If you are a private equity firm, you are in the business of buying lots of companies with lots of debt, and eventually some of them are going to run into trouble. One thing that you could do in that situation is say “ah well, we gave it our best shot,” file for bankruptcy and hand the company over to its creditors. But probably you should get good at optimizing this situation: It will come up a lot, and there are trades to be made. Some of your companies will run into trouble and be unable to pay back 100% of their debt. If you go to all of the lenders and say “we’ll give you 80% if you let us keep the company,” they might say yes. But if you go to 51% of the lenders and say “we’ll give you 110% if you agree to stiff the other guys and let us keep the company,” they’ll probably say yes, and you can keep more of the company for yourself. 

And so these LMEs are stereotypically associated with private equity ownership, and private equity sponsors — who are repeat players with smart employees and good lawyers — tend to be good at them. In theory, you do not want to be too good at them, as a private equity sponsor: If you are too good at extracting value from your lenders, people will stop lending to you, and you will be unable to do any more leveraged buyouts. In practice memories are short and this theoretical limit has probably never been reached, but you never know.

Bloomberg’s Reshmi Basu has a story today about the state of LMEs:

Tropicana Brands Group is among a stampede of recent cases that reveals a class system is taking root after lots of these refinancing deals, cementing in place an approach that tends to favor bigger lenders and penalize the rest. The orange-juice giant and its owner PAI Partners discussed plans with a select band of creditors — including heavyweights such as Carlyle Group and CVC Credit Partners — to divvy up its loans into three tiers, people with knowledge of the talks say. Each rank will be worth less than the one above. Much less.

The inside group is in line to swap most of its existing debt for so-called first-out and second-out loans, putting them at the front of the repayment queue, and they’re being priced respectively at 95-97 cents on the dollar and 61-63 cents. Other lenders are braced for a hefty portion of third-out debt valued at just 28.5-31.5 cents. Owning supposedly safe first-lien loans won’t save them.

Similar cases abound. Newfold Digital, a web company owned by Clearlake Capital and Siris Capital, is in talks to possibly split its loans into four tiers, according to people familiar. Marquee names Pimco and Blackstone Inc. are part of the creditor in-crowd, and likely in the box seat for the skinniest losses.

A couple of points here. First of all, conceptually, the basic trick that I described above shouldn’t work. As a US appeals court said about Serta Simmons Bedding LLC’s 2020 liability management exercise:

Ratable treatment is an important background norm of corporate finance. Pursuant to this norm, a borrower must treat all of its similarly situated lenders, well, similarly. … Ratable treatment is such an important norm that it is often described as a lender’s “sacred right” under syndicated loan agreements.

This norm is, I think, much stronger in equity than in debt, but it really is an important background norm. And then the point of many LMEs is to find loopholes in the loan documents that let you violate this norm and treat some of your lenders better than others. In Serta Simmons, the trick was that the company was allowed to do “open market purchases” of its debt without offering the same deal to everyone: If it bought one lender’s loans back in the open market, that didn’t require it to buy back everyone’s loans. The company argued that this allowed it to do a non-pro rata negotiated exchange, and a bankruptcy court agreed, but ultimately the appeals court reversed that and eliminated the “open market” loophole. 

So the market quickly moved on to find other loopholes. Basu:

On New Year’s Eve, the Fifth Circuit Appeals Court in the US offered a glimmer of hope to downtrodden lenders when it ruled that Serta Simmons Bedding’s so-called “open market purchase” of discounted debt to complete a 2020 restructuring — a famous test case in the world of distressed borrowing — wasn’t allowed under the terms of its pre-existing loans.

The wish was this might finally curb a standard modus operandi in LMEs, where some lenders hand fresh cash to a company and then get the chance to swap their old loans for new debt that gets priority over other creditors.

Any celebrations by the left-behind have been short lived. Better Health Group and Oregon Tool Inc., a chainsaw maker, have already found workarounds in ongoing restructuring deals, according to lawyers.

The workaround seems to be what is called “extend and exchange,” and it is a neat little trick. Loan documents often require the company to treat everyone in a class of loans similarly, but they allow the company to treat different classes of loans differently. For instance, if you have 2-year loans and 3-year loans, you can offer the 2-year lenders something different from what you offer the 3-year lenders. And the agreement might allow individual lenders to extend their loans — to go from 2 to 3 years, for instance — without offering that opportunity to everyone. But of course if you extend your loan, you are in a separate class from the un-extended loans.

That’s all you really need. You go to 51% of your lenders, you ask them to extend their loans by a month, now they are a separate class, and now you can give them a better deal than the other 49%. [3]  This approach seems to have been deployed within weeks after the Serta Simmons decision, which suggests that private equity firms and their lawyers already had it ready. The market for this sort of thing is very efficient: Finding ways to shift value from one set of creditors to another is very valuable to private equity firms, and their demand creates supply.

Second, you could imagine lenders fighting back. Ex ante, this sort of thing probably does take value from lenders and distributes it to private equity sponsors and their lawyers. Basu:

Moody’s Ratings reckons only half of distressed exchanges help businesses avoid defaulting again. These swaps handed creditors average losses of 7% to 21% last year, Fitch Ratings estimates, worse still when it was an LME. That’s better than the 60% hit on bankruptcies, but lasting recoveries are rare. …

While participation rates in restructuring deals are soaring as the tiered model emerges as the go-to strategy, that’s largely because the alternative for those down the pecking order might be getting nothing at all. A key problem for lenders trying to navigate the new landscape is the tangled math needed to price their first-lien debt when it may end up sliced, diced and devalued. …

Legal and other fees increase the company’s burden. Buyout firm Platinum Equity is a regular LME protagonist, and early term sheets for its refinancing of digital PR group Cision Ltd. called for $30 million of professional fees to be funded from a new loan, according to people with knowledge of the situation. 

And there is some pushback; some loan agreements now have “Serta blockers” that require the consent of all creditors (not just a majority) to subordinate the existing debt to new debt, which prevents a lot of this sort of up-tiering. Or creditors could just refuse to participate in these sorts of LMEs, but they mostly don’t:

More of a threat comes from early signs that creditors en masse have grown sick of the LME merry-go-round and are becoming more willing to stare down private equity sponsors. In talks around ailing software group RSA, several lenders refused to accept Clearlake’s proposal of a double-digit haircut, and its reluctance to put in more cash. They say they’d rather wait it out and own the company if it eventually runs out of road.

The sheer amount of money sloshing around the private markets makes this a brave gambit. Clearlake, for instance, has built strong relations with many direct-lending funds — a ready alternative to traditional syndicated loans.

For private equity sponsors, the attraction of of private credit is that (1) private credit funds are desperate to make loans, so they value good relationships with sponsors who will bring them deals and (2) private credit funds value long-term relationships and will be nice to companies that run into trouble. If you are a public lender, you have to compete with that.

Third, I said above that, in theory, if you are a private equity sponsor, you don’t want to get too good at extracting value from your lenders, because then they might stop lending to you. Sometimes you might have to walk it back. Basu reports:

Clearlake, criticized by some lenders for its chunky management fees and dogged defense of its own equity’s value, has been involved in multiple grueling LMEs. It has spent time this year in one-to-one creditor meetings to offer a mea culpa of sorts, according to people familiar with the matter, acknowledging gripes while stressing its need to protect its interests. The firm wants to improve lender communication, a person with direct knowledge says. ...

Private-markets titan Apollo Global Management Inc. started a similar PR tour in 2015 after skirmishes with creditors, according to people familiar. 

Good for Clearlake. We talked about the Apollo apology tour a while back, and I wrote [4] :

If you build a reputation for extracting a lot of battle by being cleverer, harder-working and more ruthless than everyone else, well, in large swathes of finance that’s just good. Investors will want to give you money, because you are clever and ruthless. People will want to work for you, because they fancy themselves clever and ruthless, and because you pay well. Creditors will fear you, which is not all good (you have to pay them more to finance your deals), but which is maybe helpful if you actually do end up fighting with them.

If you develop a reputation for being ruthless in LMEs, that will probably make your LMEs easier: Your lenders will want to be in the group that cooperates with you, because they know that being in the other group will be bad. In theory this will make it harder for you to borrow money in the first place, but in practice there’s a lot of private credit money out there.

MicroStrategy Inc.’s core discovery is that the US stock market will pay $2 for $1 worth of Bitcoin. MicroStrategy (now called Strategy) is essentially a pot of Bitcoins, and its stock trades at roughly twice the value of its Bitcoins. So it issues more stock and buys more Bitcoins and goes up in a bizarre perpetual motion machine that I write about a lot but cannot explain. Many people have noticed, though, and we talk a lot about companies that copy MicroStrategy’s approach — being a “Bitcoin treasury company” — with various minor twists (being a Japanese company, doing Solana or Trumpcoin instead of Bitcoin, etc.). I am not sure that this always works, but it does seem to work more than I would have expected. The stock market pretty consistently pays $2 for $1 worth of crypto.

MicroStrategy’s other good discovery is that convertible bond investors will pay a lot for Bitcoin treasury company convertible bonds. Why? I think there are two conceptually related but very different-sounding answers:

1. If you like Bitcoin but worry about its volatility, a Bitcoin convertible is a good story: You can buy the convertible for $1,000; if Bitcoin (and MicroStrategy stock) goes up, you will get (some of) the upside; if Bitcoin (and MicroStrategy) crashes, you will get your $1,000 back. [5]
2. If you don’t care about Bitcoin but are a convertible arbitrageur, you make money on volatility: You buy convertible bonds, sell some of the underlying stock short to hedge your stock-price risk, and adjust the hedge over time as the stock moves. The more the stock moves, the more money you make adjusting your hedge. (We discussed this in more detail last year.) Bitcoin is stereotypically volatile, so intuitively you would expect a Bitcoin treasury company’s stock to be pretty volatile.

You might think that crypto treasury companies would also have an easy time copying this. Actually, though, Bloomberg’s Yiqin Shen reports:

The success of Michael Saylor’s fundraising playbook for Strategy is causing trouble for its imitators, as the market for the convertible bonds they are using to fund Bitcoin buys is getting tighter — and a rival with heavyweight backers is about to start pumping out its own paper. …

“There’s a bit of exhaustion in the new deal space. I don’t think convertible bond investors necessarily want new names since they have quite a bit to play with”, said Bank of America Corp. Head of Global Convertibles and Preferred Strategy Michael Youngworth.

The big new Bitcoin treasury convertible issuer is Twenty One, the pot for Tether’s and SoftBank’s Bitcoins, which is going public in the US by merging with a special purpose acquisition company and which plans to issue convertibles. (We talked about Twenty One last month.) But while “MicroStrategy owns Bitcoin, Bitcoin is volatile, convertible investors like volatility, I can own Bitcoin, so convertible investors will like me” seems like a plausible reasoning, it isn’t quite right, for two reasons.

First: MicroStrategy is not volatile just because it owns Bitcoin. MicroStrategy is much more volatile than Bitcoin, and the reason for that seems to be partly that it is a meme-ish stock but mostly that there are big popular double levered exchange-traded funds that own a lot of its stock. These ETFs rebalance every day, which means that, whenever MicroStrategy’s stock goes up, they have to buy more stock; whenever it goes down, they have to sell. This amplifies MicroStrategy’s volatility. Convertible investors do the opposite trade — they buy back some of their hedge shares when the stock goes down, and sell more when it goes up — which means that they profit from volatility but also dampen it. Having a big supply of ETF volatility makes MicroStrategy’s volatility very attractive to convertible investors. But not every crypto company can replicate that.

Second: To profit from volatility, convertible arbitrageurs need to sell the underlying stock short, which means that they need to borrow it. If they can’t borrow the stock, there’s no trade. Shen:

The trade depends on the hedge funds’ ability to borrow a target company’s shares at a reasonable cost. 

While that’s not a problem for Strategy, small issuers with few shares available for shorting are less likely to draw the interest of the hedge fund crowd.

“If short borrow does not exist, the arbitrageurs can’t get involved,” Bank of America’s Youngworth said.

MicroStrategy really is a finely tuned machine for extracting value from Bitcoins in the public markets, but it’s not that easy to copy it. Honestly it’s still pretty easy though.

My impression is that most senior executives at big banks think that working from home is bad and that everyone should be in the office every day. For a while after the pandemic, that was a hard sell to employees, but norms have shifted and now it is easier to tell employees to return to the office. The prior regime had one big advantage for executives, though, which is that they could save money on office space. The problem is that office space is not instantaneously flexible, so if you made office-space decisions in the old regime, in the new regime you are haranguing your employees to come back to desks that don’t exist:

As HSBC Holdings Plc prepares to move into a new London headquarters as early as next year, executives are grappling with how to handle a more severe desk shortage than they previously anticipated in the new, smaller building.

Senior managers have informed executives that under one relocation scenario being explored, the new Square Mile headquarters building would face a shortfall of 7,700 desks for staffers, according to people familiar with the matter. That’s an even more acute shortage than earlier in-house projections of less than 5,000, the people said, asking not to be identified discussing internal deliberations.

A combination of the firm’s return-to-office mandates and a larger-than-expected London-based workforce led to the revised estimates, the people said. … 

Estimates made at a time when most staff were spending a significant portion of time working from home have been upended both by the gradual return-to-office-based working and a trend for more amenities that have increased the overall amount of space needed.

Oops! It will be strange if the medium-term effect of the pandemic is that offices get more crowded.

Banker Bonuses Set to Drop as Tariffs Cause Economic Uncertainty. Citi’s Banamex Focused on IPO as CEO Rules Out Direct Sale of Bank. BlackRock orders managing directors back to office five days a week. Coinbase Buys Derivatives Venue Deribit for $2.9 Billion. France Blasts EU Plan for Retail Investor Protection as Costly. Bill Gates accuses Elon Musk of ‘killing’ children with USAID cuts. Disney Plans New Theme Park in Abu Dhabi. AIgatha Christie. Florida Judge In Ethics Case Defends Remarks As ‘Dad Jokes.’ Habemus papam. “When I was watching [American Psycho], I was like, Damn, I wish I had his skin-care routine, his morning routine. The only part that isn’t perfect is his psychopathic tendencies.” 

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!