Matt Levine is simply the best

[ValueArb]

Quote

 

Over the weekend a friend pointed out to me that trucking company Yellow Corp.’s logo is orange. I don’t know what to do with this information, but I found it unsettling and have decided to inflict it on you. Anyway Yellow also made some other mistakes and is now bankrupt:

Yellow Corp. filed for bankruptcy and will remain shuttered — throwing 30,000 people out of work — after the trucking firm’s long-running financial woes were compounded by a dispute with the Teamsters Union.

The carrier will sell its warehouses and trucks in order to repay a pandemic-era, government loan of $737 million and another $485 million in debt it owes private lenders, according to court papers filed Monday in bankruptcy court in Wilmington, Delaware. The Chapter 11 petition gives Yellow breathing room from creditors as it winds down and solicits bids on its assets.

The company’s shares are already reversing some of last week’s gains, plunging as much as 40% as of 11:07 a.m. New York time Monday.

Yes “last week’s gains”: Yellow closed at $3.90 last Tuesday, its highest price since February, despite having shut down and being in plans for bankruptcy. Now that it is in bankruptcy — with a plan to liquidate and return nothing to shareholders — the stock is off a bit; it traded around $2.40 at noon today, still much higher than it was trading in early July, before it shut down operations. I don’t know, man. Here’s the Wall Street Journal on the return of meme stocks:

Trucking-business Yellow shut down operations around a week ago, but shares have jumped some 400% since. And shares of cash-strapped drugstore chain Rite Aid surged 68% in the past week—for no obvious reason. Their rallies contrast with the overall market. The S&P 500 just notched its worst weekly performance since March. ...

Ali Behzadpour ... started noticing moves in Tupperware and Yellow shares a couple of weeks ago through stock-scanning software and buzz on Discord. He said he has been “scalping” Tupperware and Yellow, or taking short-term positions as brief as seconds or minutes. In one recent session, he estimates, he made more than $3,000 from his short-term Tupperware trades.

“I knew there was some momentum behind it,” said Behzadpour. “And hype.” 

Individual investors’ daily net purchases of Yellow totaled nearly $5 million on Tuesday, according to Vanda Research. It had previously never cracked $1 million, according to data going back to 2014.

Terrific. The way it works now is that if US companies go bankrupt their stocks go up. The stock of a small-cap trucking company is just a boring stock, but the stock of a bankrupt company is an absurd lottery ticket, and that is just more fun. Yellow went from being a boring stock that had lost value to being a silly stock that has gone up, and that’s how modern finance works I guess. 

Though actually the Yellow story is stranger than that. The Journal also notes:

MFN Partners, a Boston-based investment firm, has accumulated more than 22 million common shares of Yellow in recent days, accounting for a 42% stake, according to Yellow’s securities filings. MFN didn’t return a request for comment on Tuesday. 

MFN took a significant stake last year in Yellow competitor XPO, according to filings by XPO. 

Here is MFN’s filing from last Tuesday, showing that it owned (as of last Tuesday) 42.5% of Yellow and that it bought almost half of that stake last Monday; its total cost was about $22.9 million, or a little more than a dollar a share. I have no idea what to tell you, except I guess that at today’s price MFN is sitting on about a $30 million profit, and more than 200 million shares have traded since last Tuesday. If you are a professional investor, “buy a ton of stock in a public company that has said it plans to file bankruptcy, and then sell the stock at a profit once it actually files for bankruptcy” is not, you know, a rational strategy, but it kind of works now? (Not investing advice!) But buying 42.5% of the company is too much, since it subjects you to short-swing profit rules that basically mean you can’t take your profits for six months. (Not legal advice!) Buying almost half a company a week before it files for bankruptcy as a long-term investment seems like a mistake.

 

 

From yesterdays Money Stuff column.

Edited August 8, 2023 by ValueArb
quoting

[aws]

It is strange which meme stocks catch on. Obviously most bankruptcy announcements do not have this impact even in this environment (see PTRA today for example), but once they do the pumps seem pretty durable. Almost seems like there is some paid promotion going on behind the scenes.  Just when it looked like Yellow would start to bleed out, it jumped as much as 50% right before the close on a report by Charlie Gasparino which essentially said: If you ignore the $1 billion in unsecured claims, there might be money leftover for shareholders. 

[Saluki]

I remember that right before it went bankrupt there was a day where it went something like 120%, then it opened the next morning down about 30%.  Stay away from the mosh pit. It's not for people who have a mortgage and a day job. 

[ValueArb]

I have a friend who was betting on BBBY, all I could do is review the balance sheet and give him my take. Don't even want to ask if he got out or not.

[aws]

Yeah, that one is especially cultlike, but you can at least see how it caught on since it originated with the Gamestop guy. Why they still think he's coming back to an empty shell to bail them out after pumping and dumping on them a year ago is another matter.

 

Others seem like they must have a promotional aspect behind them, like Yellow or Tupperware recently.

[ValueArb]

Quote

Look, as a trade, I am always going to love a founder selling high at the peak of the cycle and then buying back low when valuations have collapsed. (“The Wag trade,” I have sometimes called it, after the dog-walking startup that extracted a bunch of money out of SoftBank Group Corp. when SoftBank was particularly enthused about, you know, throwing hundreds of millions of dollars at dog-walking startups, and then bought SoftBank out cheap when that enthusiasm waned.) And I suppose if this is your plan, then it helps to (1) sell high, (2) continue running your startup as a division of the buyer and (3) be controversial and bad at it, so the value goes down and the buyer feels forced to sell. Crafty!

But “Dave Portnoy sold Barstool high at the peak of legalized sports gambling in 2020, and then bought it back low when the bubble popped in 2023,” does not seem quite right? I would not exactly say that Penn got what it paid for, with Barstool, but on the other hand Penn’s “interactive” segment — consisting largely though not exclusively of its “online sports betting and casino app called Barstool Sportsbook and Casino” and its in-casino Barstool-branded retail sportsbooks — went from $121 million in revenue in 2020 to $663 million in 2022, and was up another 66% year-over-year in the first half of 2023, for a run rate of almost $1 billion. “Barstool has been a great partner and we are thankful to Dave Portnoy, Erika Ayers, Dan Katz and their team for helping to rapidly scale our digital footprint across 16 jurisdictions in the U.S. and introducing their audience to our retail and digital products,” said Penn in announcing yesterday’s deal. Could Penn have added $500 million of annual sports gambling revenue without the marketing partnership? Could it have added more, without the public-relations and licensing problems that Barstool brought? Quite possibly! There really was a pretty big sports-gambling gold rush, and it does look like Penn spent lavishly and imprecisely to get in on it.

But basically Penn paid $550 million to Barstool to get into online sports gambling with a popular brand, and amortized that cost over about three years, or call it $180 million a year. And now it is paying ESPN $2 billion for a 10-year licensing agreement, or $200 million a year. Portnoy didn’t exactly sell Barstool at the peak and buy at the bottom: He sold Barstool at the point when it made sense for Penn to partner with Barstool, and bought it back at the point when it made sense for Penn to partner with Disney.

One assumes that Penn would have preferred to sell Barstool back to Portnoy for some price higher than zero, but:

It didn’t make sense for them to keep it,
It was perhaps too controversial for most other media or gaming companies, and
Presumably any buyer would have to come to terms with Portnoy for it to make sense.
So I guess Portnoy was the only bidder? I dunno, I just Googled “Dave Portnoy enemies” and got this list that he tweeted last year beginning with Henry Blodget, and if I were Penn I would have called Blodget (a wealthy media mogul!) and said “hey you can have Dave Portnoy for $10,000” just to see if I could spark a bidding war. It is possible though that Portnoy’s non-compete promises are worth more to Penn than a buyer would pay.

Also Penn gets 50% of the proceeds if Portnoy sells. He “has no plans to sell Barstool Sports again,” and why would he? He sold it for $550 million and got it back for free; selling it again would just be greedy. Once you’ve cashed out $550 million and kept the business, it can be a lifestyle business. I assume that Penn’s 50% is just schmuck insurance.3 It’s embarrassing enough to buy it from Portnoy for $550 million and then sell it back to him for $0, but it would be much worse if he turned around and sold it to someone else for $550 million again.

 

[ValueArb]

Quote

 

If you have a bunch of trees, and you chop them down to make paper or lumber or whatever, you can sell the paper or lumber or whatever for money, but on the other hand trees store carbon and cutting them down is bad for climate change. If instead you do not chop down the trees, that is good for the environment, and it is a great innovation of modern finance that, now, you can get paid for not chopping down the trees. This is called “carbon credits.” There are measurement problems.

 

If you mine Bitcoin, you use a lot of electricity to run computers to perform calculations to get Bitcoins for yourself, which you can sell for money. But this is bad for the environment, because it uses electricity that is probably generated in ways that release carbon.1If you were to stop mining Bitcoin, conversely, that would be good for the environment. Can you get paid, though, for not mining Bitcoin? Oh yes, modern finance has solved that one too:

Bitcoin miner Riot Platforms Inc. made millions of dollars by selling power rather than producing the tokens in the second quarter as the crypto-mining industry continued to grapple with the impact of low digital asset prices.

The Castle Rock, Colorado-based company had $13.5 million in power curtailment credits during the quarter, while generating $49.7 million in mining revenue. Riot booked $27.3 million in power curtailment credits last year and $6.5 million in 2021 from power sales to the Electric Reliability Council of Texas, which is the grid operator for the Lone Star state. …

The company had $18.3 million in power credits in June and July based on its latest monthly operational updates, including $14.8 million in power curtailment credits received from selling power back to the ERCOT grid at market-driven spot prices under its long-term power contracts and $3.5 million in credits received from participation in ERCOT demand response programs.

Here is the 10-Q; this stuff is described in Note 8. Some of what is going on here is that Riot has a long-term power supply agreement in which TXU Energy Retail Co. has to supply it with electricity at fixed prices through 2030, and Riot has the option to sell the power back to TXU, at market rates, for credit against its future electric bills, when the spot price exceeds the contract price. But part of it is demand response, where ERCOT pays Riot cash for using less than its typical electrical load during periods of peak demand. 

 

As with carbon credits, there are measurement problems; I have never mined a single Bitcoin, yet ERCOT has never sent me a penny for my forbearance. Still, how great is modern finance? Twenty years ago, if you had told people that one day they could get paid for not mining Bitcoin, they would have said “what?” But now it is possible. Modern finance created the problem (Bitcoin mining) and the solution (paying people not to mine Bitcoin); the overall result is that nothing happens and yet people get paid. Just a miracle of financial engineering.

 

Also: Riot is getting paid for not using electricity, but if you are an enterprising Bitcoin miner surely you should look into getting paid for not using carbon when you are not mining Bitcoin. Riot is not there yet, but it is possible to imagine a warming world in which energy prices go up and Bitcoin prices go down and Bitcoin miners can get paid more for not mining Bitcoin than for mining Bitcoin. Giant fortunes will be made by people who got in early to the business of not mining Bitcoin. The future is so good, man.

 

 

[ValueArb]

From a week or so ago, thought it was interesting in how ARM's price has been "set". 

 

Arm 

Schematically a very good trade is:

1. Buy 10% of a private company at a $1 billion valuation, paying $100 million.
2. Buy another 5% at a $2 billion valuation, paying another $100 million.
3. Now you own 15% of a company worth $2 billion. Your stake is worth $300 million, but you paid only $200 million for it.
4. You have a profit of $100 million.

The problem is that, while on a mark-to-market basis you have a $100 million profit, on a cash-flow basis you have paid $200 million and not gotten any cash back. If you sell your 15% stake for $300 million then, great, $100 million profit, but if you were the only buyer at the $2 billion valuation (or at the $1 billion one!) then you might end up with a loss. On the other hand, markets do tend to anchor on the last trade. If the last trade was you buying at a $2 billion valuation, then the best guess is that the next trade will also be at around a $2 billion valuation. If that’s you selling, then, good.

 

I think of this trade — buy a stake in a private company, then buy some more at a higher valuation — as being a specialty of SoftBank Group Corp., mostly because SoftBank did this in a very prominent way with WeWork Inc. It did not work: SoftBank invested in WeWork at a $20 billion valuation, and then a few months later it invested again at a $47 billion valuation, and then a few months later it utterly failed to sell WeWork to the public market at a $96 billion, or any, valuation. (And now WeWork has a public market capitalization of less than $300 million.) 

 

Still. In 2016, SoftBank took chip design company Arm Holdings Ltd. private at a $32 billion valuation. Then it sold about a quarter of the company to the Vision Fund, an investment fund managed by SoftBank, at that valuation. And then this month it bought that stake back from the Vision Fund for $16.1 billion, a $64 billion valuation. The Vision Fund doubled its money (in cash), while SoftBank doubled the valuation of its stake (on paper). Yesterday Arm filed to go public again. There are skeptics; the Financial Times’s Lex column writes:

SoftBank has tried to set the stage for a high price. The listing document reveals an internal transaction in which it acquired its own Vision Fund’s stake in Arm for $16bn, a deal that valued the chip company at more than $64bn.

Investors should take little notice of this figure. On a broader, average industry earnings multiple, Arm’s enterprise value would be closer to $30bn. This is not far off the price SoftBank paid when it bought the company in 2016.

To be fair, Arm says that itself; a risk factor in the prospectus warns:

The purchase price paid by SoftBank Group to acquire shares in Arm Limited from SoftBank Vision Fund may not be, and should not be treated as, indicative of the trading price of our ADSs following the completion of this offering.

In August 2023, a subsidiary of SoftBank Group acquired substantially all of SoftBank Vision Fund’s approximately 25% interest in Arm Limited at a purchase price of approximately $16.1 billion, with the associated payments to be made in installments over a two-year period. The purchase price for this transaction was established by reference to the terms of a prior contractual arrangement between the parties. Moreover, the transfer was part of a larger transaction that also involved transfers of certain other entities from SoftBank Vision Fund to SoftBank Group. The consideration for such transfers is not included in the purchase price paid for the shares of Arm Limited. In light of the foregoing, investors are cautioned that the purchase price paid in respect of the Arm Limited shares may not be indicative of, and is not intended to reflect, expectations regarding the trading price of our ADSs following the completion of this offering.

On the other hand, if you don’t think that the stake is worth at least $16 billion, it's a bit odd to pay $16 billion for it? Apparently the “prior contractual arrangement” was a cap on the Vision Fund’s return at two times its money; paying out the maximum return just before the initial public offering is a pretty bold bet that the IPO will go even better.

 

[ValueArb]

My Forex Funds

Two stylized facts about the foreign exchange markets are:

• It is very hard for retail traders to reliably make money trading currencies, and
• They keep trying?

From first principles, it seems like a good idea to be on the other side of their trades. If retail traders want to buy pounds, sell them pounds. If they want to sell yen, buy yen from them. Etc.

 

If you do this as a full-time business, you will want to make some refinements. Here are the three main refinements:

1. It can’t really be the case that retail investors lose money because they consistently bet the wrong way on currency movements. (That would take skill!) They lose money because (1) they make essentially random-chance bets on currency movements, with zero expected value, and (2) they lose a bit of money on each trade from commissions and spreads. So, if you are in the business of trading with retail traders, make sure to charge them lots of commissions and spreads.
2. Meanwhile you don’t want to incur the costs of going out and buying yen, pounds, etc., each time your retail customers do a trade. Ideally you will “bucket” their trades: A customer comes to you and says “I want to buy $1,000 worth of yen,” they give you $1,000, and you say “okay great you have 14,400 yen in your account.”1 You don’t actually go and buy them any yen (you just make a note in your books saying that you owe them 14,400 yen), and they don’t actually want yen (they just want to bet on the price). They will eventually close out the bet by selling you the 14,400 yen back for dollars, and when they do you will cheerfully deliver them, like, $970. You keep their $30 loss as your profit. No yen are ever involved.
3. You will want them to make leveraged bets: Instead of giving you $1,000 to buy $1,000 worth of yen, they give you, like, $100 to buy $10,000 worth of yen. Then if the yen moves against them by 1%, they have lost all their money. If the yen moves in their favor by 1% they have doubled their money, but the odds are in your favor. 

The straightforward way to combine Refinements 2 and 3 is to trade FX futures, where the customers put down a small amount of collateral to enter into a contract with you that pays out based on the price movements of some currency pair, so (1) the trade is leveraged and (2) you never have to go out and actually buy any foreign currencies. FX futures are a very standard way to trade currencies, so customers will be happy to trade them with you.

 

It should go without saying that none of these refinements are legal advice.2

You can make further, shadier refinements. For instance, if your customers are only trading with you — if you never go out and buy any actual currencies or do futures trades with outside dealers — you might think, well, it is not strictly necessary that the prices I charge customers reflect actual market prices. They could just be different prices. If yen futures currently trade at 147.4 and a customer comes to you looking to buy yen, you sell them yen at 146.9 to the dollar. Why not, who cares, it’s all just happening on your own computer systems. You have to be careful with this: You probably have to show your customers some screen of bid and ask prices before they trade, and if you show them that yen are 147.5 bid / 147.4 offered and they hit “buy,” they might be puzzled to get only 146.9 yen for their dollar.3 But (1) they might not be (not everyone pays attention) and (2) you can explain it away. “Ah, price impact, market slippage, the market was at 147.4 but then your big buy order moved it,” you can say.

 

Another important set of refinements is: Some customers might actually be good at trading foreign exchange, and will cost you money; what do you do about them? Your answer might be some combination of:

• Shut down their accounts.
• Jack up the commissions, spreads, fake prices, etc., that you charge them, so that they start losing money.
• Stop bucketing their orders, route them to real markets, and let them trade with their own money with someone else; you charge commissions but no longer take the other side of their bets.

The last important refinement, of course, is that you probably want to lie about all of this? “We charge zero commissions and all of our customers make money!” might be a lie, but it is a better pitch than, like, “we have found a maximally efficient way to fleece rubes like you.” Everything else I said above can get you in trouble — none of it is legal advice! — but the lying is its own separate way to get in trouble.

On Friday the US Commodity Futures Trading Commission brought a fun enforcement action against Traders Global Group Inc., which runs an online FX trading platform called “My Forex Funds” that “offers retail customers the opportunity to become ‘professional traders,’ using Traders Global’s money to trade against third-party ‘liquidity providers’ and sharing in any trading profits”:

In return for the opportunity, customers pay certain fees to Traders Global, and are required by Traders Global to maintain a certain minimum amount of account equity, referred to as a “drawdown limit.” Traders Global assures customers that “your success is our business,” and “we only make money when you do.” Traders Global’s pitch has proven appealing to customers; more than 135,000 of them signed up during the relevant period, paying at least $310 million in fees.

But Traders Global is allegedly on the other side of all the trades:

Traders Global is a fraud. In reality, Traders Global is the counterparty to substantially all customer trades. When customers make money, it means that Traders Global loses money. … Customers do not trade “live funds” against “multiple liquidity providers” nor do customers share in “trading profits,” as Traders Global claims. In reality, customers trade against Traders Global in an electronic trading environment that Traders Global controls.

The CFTC’s complaint hits all the highlights. There are commissions of course:

Traders Global claims on its website that customer trades are subject to “commissions” of $3 per lot. Because Traders Global tells its customers that they are using Traders Global’s funds to trade, the customer is led by Traders Global to believe that the commissions are those being charged to Traders Global by a liquidity provider, exchange, or other third party.

In fact, Traders Global is not charged commissions by any third party on “trades” for which it acts as counterparty. Traders Global fails to disclose, however, that it—not some third party—assesses these commissions. The so-called “commissions” are simply a charge against customer account equity imposed by Traders Global.

And customers make leveraged futures trades, maximizing their likelihood of losing all their money and having to put in more:

A customer who signs up for an account is required to stay within a “drawdown limit” imposed by Traders Global. The drawdown limit requires a customer to maintain a certain minimum amount of equity in his or her trading account. The amount varies with the size of account. …

In a December 11, 2021, YouTube video, Kazmi explained that the drawdown is designed “to force traders to get into that habit of locking in some of the profits . . . .” “If you lose all that money,” Kazmi explained, “it’s not only you that’s losing, it’s us as well right. Because we want to make you profitable so we can be profitable . . . .”

If a customer exceeds the drawdown limit, Traders Global will “disable” (i.e., terminate) the customer’s account. If the customer wishes to continue with Traders Global, the customer is required to re-start their account and pay another fee.

 

There is, uh, slippage:

Traders Global uses specialized software to automatically add “delay” or “slippage” to customer orders. Traders Global does this in order to reduce the likelihood, or amount, of a customer’s profitable trading….

By dialing the “slippage” up, Traders Global sets the software to automatically execute a customer’s market order at a worse price than the best bid or offer displayed at the time the customer entered the order. For example, if a customer entered an order to buy a leveraged contract for Euros when the offer price displayed on the customer’s computer screen was $1.759, imposition of slippage would result in the order being executed at a higher price, e.g., $1.800.

Traders Global subjects most of its customers to some amount of delay or slippage.

And there is the handling of profitable traders, which involves dialing up the slippage:

Customers whose trading generates consistent profits, however, are subjected to longer delays and increased slippage, implemented by Traders Global.

And then, if that doesn’t work, routing them to other counterparties so they are no longer Traders Global’s problem:

A very small number of customers trading “live” accounts manage to trade profitably despite Traders Global’s attempts to handicap them. After identifying such customers, Traders Global may route some or all of these customers’ orders to an off-exchange leveraged forex and commodities dealer outside the U.S. (hereinafter, the “Dealer”). Traders Global refers to this as “STP’ing” a customer’s account; “STP” stands for “straight-through processing.”

STP’ing a customer is very rare. Of 24,000 customers with “live” accounts during the Relevant Period, fewer than 100 of them had a single trade STP’d.

In one respect, the STP’d customers get what they were promised—the opportunity to trade against a liquidity provider using Traders Global’s money, which Traders Global has on deposit with the Dealer, in a margin account. …

For customers on STP, Traders Global uses its specialized software to impose an additional “spread” on the price feed visible to STP’d customers. This spread results in a customer seeing—and believing himself or herself to be executed at—a worse price than what Traders Global got from the Dealer.

All the good stuff. I guess one question is how much of this is illegal. The bones of it — “we will trade futures clients with our customers, charge them commissions, close their contracts when their equity gets too low, and route orders we don’t want to other deals” — seems kind of fine? As long as you don’t lie about it?4 But the combination, plus the alleged lying, is quite lucrative:

[Chief Executive Officer Murtuza] Kazmi has used the proceeds from the Traders Global fraud to fund a luxury lifestyle.

In April 2022, Kazmi paid $1.6 million to purchase a Lamborghini Aventador at auction. In December 2022, Kazmi paid $3.3 million for a Bugatti racecar.

[ValueArb]

First regulators broke trades to steal from those lucky enough to buy BRKA at $150 last week, now Schwab is screwing up once in a lifetime trades that would have made a customer millions.

 

Quote

Last week, Jason Leopold, at Bloomberg’s FOIA Files newsletter, rounded up a bunch of Trump Media shareholder complaints to the US Securities and Exchange Commission, and they are mostly what you’d expect: “The Market Makers in DJT are manipulating the OPTIONS prices,” that sort of thing. (DJT is Trump Media’s ticker symbol.) But I would not in a million years have predicted this complaint:

One person claimed that they would’ve reaped a windfall from the sale of Trump Media stock on the day it went public, but brokerage Charles Schwab screwed up and botched the sale. They said the value of their securities skyrocketed to $5.2 million on March 26 and that a person monitoring the stock with them “screamed at me at the top of her voice: ‘Sell it all! RIGHT NOW!!!!!!’”

 

When they tried to unload the stock, however, their account showed they sold their holdings for $68 a share, not $15,994 a share, which is what they claimed it traded at when they placed a sell order.

 

A spokesperson for Charles Schwab explained to FOIA Files what happened.

 

“On March 26, Nasdaq implemented a symbol change of the SPAC ticker DWAC to DJT,” the spokesperson said, referring to Digital World Acquisition Corp., the blank check company formed to merge with Trump Media & Technology Group. “A few of Schwab's platforms displayed a quote for $DJT (Dow Jones Transport Index) rather than the DJT. This issue only impacted quoting and not trading. Schwab promptly fixed the issue that same morning, and all client orders were filled at the correct price, which that day ranged from $57.25 to an all-time high of $79.38,” the spokesperson said. …

 

“I was so exceedingly devastated by the fact that I wasn't actually a multi-millionaire because of Schwab's website messing up that I simply closed the customer service chat box,” the person wrote.

 

Edited June 10 by ValueArb

[aws]

The display bug convinced him to sell at the top. Lucky guy

[Gregmal]

Reminds me of during the 2021 boom when there were screenshots of people ripping Robinhood for their orders not getting filled...on Saturdays. 

[Whensthepaintdry?]

His recent podcast had a segment on the recent Ackman listing news. 

[Xerxes]

1 hour ago, Whensthepaintdry? said:

His recent podcast had a segment on the recent Ackman listing news. 

 

Thanks. This looks great.

I reposted it on Pershing thread.

[ValueArb]

Quote

 

Brad Jacobs is a successful businessman “who built multibillion-dollar companies in logistics and other sectors through acquisitions.” Last year, he decided that he will next get into the business of rolling up companies that distribute building products. To do this, he started a company called QXO Inc., made himself the chairman and chief executive officer, picked a team, and funded the company with $900 million of his own money and $100 million from outside investors.

 

Jacobs has done well in the past, and people want to back him, so QXO was a hot enough ticket that last week — just a week after Jacobs put in his own money, and before he acquired any businesses — QXO was able to raise an additional $3.5 billion by selling stock and warrants privately to “certain institutional and accredited investors.” (Bloomberg News reported that the investors include affiliates of the 3G Capital founders and the Walton family, as well as T. Rowe Price.) That deal hasn’t closed yet — QXO expects to get the money “early in the third quarter,” so next month or so — but it’s a nice start. QXO has raised a roughly $4.5 billion fund to go out and buy buildings products distributors, and it “is targeting tens of billions of dollars of annual revenue in the next decade through accretive acquisitions and organic growth.”

 

Often, the way this would work would be that QXO would go buy companies and build a business, and it would deploy a lot of that $4.5 billion and achieve perhaps billions of dollars of revenue over a few years. Perhaps it would raise more money, or perhaps that initial $4.5 billion plus organic cash generation would be enough. And eventually QXO would be successful and growing and profitable, and Jacobs and his investors would start thinking about taking it public, so that they could sell their shares to other investors and monetize the business he built, and so that he could use QXO’s stock as a currency to acquire even more building products distribution companies.

 

QXO, however, took things in a slightly different order. When I say that Jacobs “started a company called QXO Inc.” with $1 billion of his own and his investors’ money, that is not quite right. What he actually did was find a teensy public company called SilverSun Technologies Inc., which he bought for $1 billion and renamed QXO, giving the company an instant public listing.

 

Even that is not quite right: Jacobs didn’t exactly buy SilverSun; he and his investors invested the $1 billion in SilverSun in exchange for new stock. They took control of the company, changed the board of directors and management and strategy, and now own about 99.8% of the stock,[1] but SilverSun’s previous shareholders stuck around. They didn’t get cashed out, and their stock still trades publicly. Or, rather, they did get cashed out — QXO/SilverSun will pay the pre-deal SilverSun shareholders a dividend of $17.4 million, which is roughly equal to SilverSun’s entire market capitalization as of last November[2] — but they kept their stock too. And that stock still trades on the Nasdaq.

 

Which creates a very weird dynamic. QXO is not really a public company. It is an investment vehicle for Brad Jacobs to acquire some companies, funded with $4.5 billion of commitments from him and his private investors. The acquisitions haven’t happened yet, and most of that $4.5 billion of investor money hasn’t come in yet. Jacobs and his investors did put in $1 billion for their stock already, but that stock is locked up for the long term and not available for trading on Nasdaq.[3]

 

But QXO is technically a public company, and there is a little rump of 664,284 shares that do trade publicly. And so if you are a public investor and, like those institutional investors, you like Jacobs’ chances as a roller-up of building products distribution businesses, you can go to your retail brokerage and buy QXO shares. But not very many of them. There are only 664,284 available, and they trade in the tens of thousands of shares per day.

 

This scarcity makes them expensive. Bloomberg tells me that QXO closed on Friday at $134.21 per share, for a market capitalization of $89.2 million. That reported market capitalization is just (1) the share price times (2) the 664,284 public shares outstanding; it is an accurate accounting of the market capitalization of QXO’s tradeable stock. But the implied value of the company is much higher, because those public shares are just a tiny sliver of its total capitalization. “On a fully diluted basis,” says QXO, taking into account the $3.5 billion capital raise and the conversion of various convertible preferred shares and warrants, “the Company would have approximately 821.6 million outstanding shares of Common Stock.” The public shares are just 0.08% of the total.

 

That is: If you multiply Friday’s $134.21 closing price by the 821.6 million fully diluted shares, you get a market capitalization for QXO of $110 billion.[4] Before it has even rolled up any building products distributors! Just a $4.5 billion pot of cash trading at $110 billion.

Or to put it another way, QXO sold $3.5 billion worth of stock, in the deal announced after the close last Thursday, at $9.14 per share.[5] The stock closed at $205.40 on Thursday (and $134.21 on Friday). QXO sold stock to private investors at a 95% discount to its public trading price. If you were a retail shareholder buying QXO stock at $205.40 on Thursday afternoon, you might be annoyed that QXO was selling it to institutions at $9.14 at the same time.

...

 

 

[ValueArb]

An outstanding detailed explanation of the MMTLP debacle.

 

TLDR: SEC claims managment was trying to engineer short-squeezes using worthless oil and gas claims, succeeding the first time but the 2nd time FINRA stepped in and stopped trading two days ahead of its conversion to an illiquid security to protect shareholders and now there is a mob of angry retail speculators stuck with worthless stock blaming it.

 

Quote

MMTLP

One of the leading financial conspiracy theories of recent years is MMTLP. I am not sure I can explain the theory very clearly, because it is the nature of conspiracy theories to be sprawling and slippery, but the central idea is that short sellers ganged up on a public company, and stock exchanges and regulators took the side of the short sellers at the expense of ordinary shareholders.

Today the US Securities and Exchange Commission brought fraud charges against Meta Materials Inc., the public company whose preferred stock, MMTLP, is at the center of a conspiracy theory, and its former chief executive officers, John Brda and George Palikaras. (Meta settled the case for a $1 million fine; Brda and Palikaras will fight it.) Here is the story, as alleged in the SEC’s complaint against Brda and Palikaras.

There was an oil company called Torchlight Energy Resources. By early 2020, “it had sold all of its revenue-generating oil and gas assets,” and was left only with speculative exploratory properties. It needed to drill those properties to keep its leases, but the drilling would not be profitable, and it needed money. Its stock fell below $1, and it was in danger of being delisted from Nasdaq.

Like the CEOs of many unprofitable public companies, Torchlight’s CEO, John Brda, had conceived a dislike for short sellers, who borrowed the stock and sold it to bet against the company. And technology existed to take revenge on short sellers.

When you sell stock short, you have to borrow it from a stock lender, and your deal with that lender is that you eventually have to return (1) the stock and (2) any dividends on the stock. If a company pays a $1 cash dividend, you pay $1 to your stock lender. If the company dividends out a share of preferred stock, you have to go buy that preferred stock and return it to your lender.

In 2019, Overstock.com, whose CEO Patrick Byrne had spent years fighting short sellers, made a plan to distribute a weird blockchain preferred stock to its shareholders that would not be tradable. This was meant to trap short sellers: They couldn’t go buy the preferred stock to return to their lenders, because it didn’t trade. They would have to close out their shorts — by buying back the stock — before the dividend was paid out, because afterwards they’d be stuck and would get in trouble. There would be a short squeeze: Before the dividend, the shorts would rush to buy back the stock, and they’d have to do it at very high prices.

Or this is the theory; it did not work for Overstock, mostly because the SEC considers it manipulation and told them not to do the dividend. Byrne was not pleased.[1]

I don’t know if Brda was inspired by Byrne’s example or what, but he allegedly developed an improved version of the idea. It goes like this (quotes from the SEC complaint):

• Torchlight would do a reverse merger with some other company that “desired Torchlight’s Nasdaq listing but not its oil and gas leases.” He found a good partner in Metamaterial Inc., a Canadian company that “focused on research and development of early-stage, applied materials technology.” (After the merger, the combined company changed its name to Meta Materials Inc., and I will refer to it that way.)
• Because Meta didn’t want the oil and gas assets, it would spin them back out to the pre-merger Torchlight shareholders. Meta would end up with a US public listing, and Torchlight’s shareholders would get (1) a stake in Meta plus (2) their oil and gas assets back, in the form of the spun-off company.[2]
• Except instead of just spinning out the Torchlight assets as a new company, Brda found a more complicated structure. At the closing of the merger, Torchlight shareholders would get, not actual shares of legacy Torchlight, but rather a new preferred stock representing a claim on the value of those assets. (This stock came to be called MMTLP, its ticker symbol.) Meta would put the Torchlight assets in a box and try to sell them after the merger. When it eventually sold them for cash, it would pay the cash out to the holders of MMTLP.
• Crucially, the plan was that Torchlight and Meta “would not register [MMTLP] or make [it] available for immediate trading on any exchange.” It would be a dividend of a non-tradable preferred, which means, in theory, that Torchlight’s short sellers would not be able to buy it to deliver to their stock lenders.
• The plan was, says the SEC, to “market and promote the Preferred Dividend to spread the narrative to select investors that short sellers would not be able to obtain the Preferred Dividend, thus pressuring short sellers to close their positions, leading investors to believe a short squeeze would occur, and artificially inflating the price of Torchlight common stock on a temporary basis.”
• And then, the plan was to take advantage of that inflated price by selling stock: Torchlight would “capitalize on Torchlight’s inflated common stock price by, among other things, raising capital through an ATM Offering,” that is, an at-the-market offering of shares directly to investors on the stock exchange, just before the closing of the merger. And then it would “use capital raised through the ATM Offering to drill wells required to maintain Torchlight’s oil and gas leases.”

So the plan was to punish short sellers with a short squeeze, and to use the short squeeze to raise money at an inflated price.

Brda and Palikaras set out to market that plan. Sort of. The SEC alleges that they went around telling investors that the shorts would get squeezed. Palikaras said, on a call with some shareholders: 

And there is one more element to add here, which is the, let’s call the x-factor. If you notice the Torchlight stock is massively shorted … This deal is set up not to give a [cash] dividend at closing. So, in order for the short positions to cover, they have to have the stock on their hand because the dividend will be paid out as a preferred share, not cash. As a result, there is no physical way for the shorts to cover the stock when the time to close, and we believe … there will be a potential jump towards the close, it’s called a short squeeze… (emphasis added).

But they couldn’t market the plan too explicitly, because, you know, this really is market manipulation. You can’t go around saying “we structured this merger to do a short squeeze”; you have to have some corporate finance reason for it. So the SEC also complains that they didn’t mention the short squeeze in their official communications to shareholders:

Specifically, Brda failed to disclose, and failed to cause Torchlight to disclose, the following in Torchlight’s proxy statements and other public filings and statements: (1) that the Preferred Dividend was intended to cause, and to lead investors to believe, that there would be a short squeeze and an increase in Torchlight’s stock price; (2) that Brda believed the Preferred Dividend would cause a short squeeze and/or temporarily inflate Torchlight’s stock price; (3) that the potential for a short squeeze and/or temporary inflation of Torchlight’s stock price were reasons that Brda and Torchlight considered in approving the merger and Preferred Dividend and recommending that shareholders approve the same; (4) Brda’s plan to, and/or the potential that Torchlight would, deploy the ATM Offering; and (5) that, in connection with negotiations over the merger and Preferred Dividend, Torchlight and Meta I discussed the Preferred Dividend’s potential to cause a short squeeze, its potential to temporarily inflate Torchlight’s stock price, and Brda’s plan to use an ATM Offering to raise funds from investors.

That is, it was market manipulation for Brda and Palikaras to go around telling some shareholders that they had concocted a short squeeze, but it was also fraud for them not to tell other shareholders that. Also, to be fair, they were telling shareholders this on Twitter, but subtly. Sort of:

On June 13, 2021—the day before Torchlight announced its Preferred Dividend Record Date—Palikaras tweeted a graphic of shorts-in-flames, kicking off a series of tweets designed to promote the short squeeze theory and encourage investors to purchase Torchlight’s common stock. ...

Palikaras testified during the SEC’s investigation that his tweet had nothing to do with the concept of a short squeeze, but the timing, circumstances, and content of his tweet imply that he intended to tout that the Preferred Dividend would set Torchlight’s “shorts” on fire by triggering a short squeeze.

Here is that picture:

This all basically worked. Torchlight and Meta Materials signed their merger agreement in December 2020; the merger closed on June 28, 2021. On June 14, two weeks before closing, Torchlight announced the record date for the MMTLP dividend: People who held Torchlight shares as of June 24 would get shares of MMTLP. The theory was that short sellers would have to buy back their Torchlight shares by June 22 (for T+2 settlement: If you buy on June 22, you get the shares on June 24), or else be stuck with no way to get MMTLP. So the short squeeze should peak on June 22. The SEC says:

Following the announcement of the Record Date on June 14, 2021, the price of Torchlight stock surged. The price at closing jumped from $3.58 per share on June 14 to $5.07 per share on June 15 to $5.99 per share on June 16. Torchlight stock price peaked at $10.88 per share on June 21—an increase of over 200% from its price at closing on June 14.

Torchlight did its at-the-market offering between June 18 and June 24, selling 16.2 million shares at an average price of $8.50 and raising $137.5 million. By June 28, after the merger closed, the stock had fallen to $1.99.[3] Torchlight’s marketing of the short squeeze worked to prop up its stock for a bit, and Torchlight took advantage to raise money.

But how did it work? Why did the stock go up? One possibility is that the shorts were in fact squeezed and had to buy back their Torchlight shares. Another possibility is that the shorts were fine — they were able to close out easily, or they rolled their Torchlight shorts into new MMTLP shorts — and what drove the stock was not a short squeeze but a meme-stock effect. Torchlight shareholders got Brda’s and Palikaras’s somewhat coy message — that there would be a short squeeze peaking on June 22 — and rushed to buy the stock to profit from the short squeeze.[4] Even if the short squeeze never materialized, if people believed that it would, that would push up the price.

Surprisingly, the SEC doesn’t know which explanation is right:

Torchlight’s stock price artificially increased as a result of Defendants’ scheme. However, the evidence available at this time is inconclusive as to whether, or to what extent, the trading volume was attributable to short sellers covering their positions versus defrauded investors purchasing Torchlight’s stock to “burn the shorts” or obtain the Preferred Dividend that they believed was worth $1–$20.

The SEC thinks that either of these is fraud. If the scheme succeeded in burning the shorts, then it is market manipulation: Brda and Palikaras artificially drove up the price of the stock by secretly engineering a short squeeze. If the scheme failed to burn the shorts, then it is still market manipulation: They artificially drove up the stock by openly claiming they were engineering a short squeeze. There is something logically inconsistent about this, but you get the idea. Brda and Palikaras were coyly half-marketing the short squeeze, both concealing it and advertising it, and either way it would be fraud.

Now, I will go ahead and guess that the answer is mostly the latter, and that in fact the Torchlight shorts mostly were not burned, and that instead the victims were Torchlight shareholders who believed in the short-squeeze theory. I am not confident about this, but it’s my best guess. 

And here’s why. This story has taken us through June 2021, when the merger closed and Torchlight/Meta raised all this money from shareholders. But there is a long subsequent history of MMTLP, which the SEC only hints at in its announcement today, saying: “A separate Commission investigation regarding subsequent events related to Meta Materials (MMTLP) remains ongoing.” 

But when we talked about MMTLP last year, it was mostly about post-2021 events. Meta never did sell those old Torchlight oil and gas assets. The SEC now says that it never planned to, that the properties were hard to value and nobody would want them, and that Brda’s claims that they’d be worth between $1 and $20 per MMTLP share were far too optimistic.

Instead — as the SEC says they planned to do all along — Meta ultimately announced that it would spin those assets out, in a separate company, to holders of MMTLP, to replace their MMTLP shares. The new company is called Next Bridge Hydrocarbons, and it is extra-super-duper non-tradable.

Because it turns out that MMTLP, which was intended not to be tradable, traded anyway: Meta never got it a listing on a stock exchange, but it traded in over-the-counter markets. And there were short sellers. Or at least, people thought there were. Maybe the short sellers had actually borrowed and shorted MMTLP, or maybe they were just old Torchlight short sellers whose short positions had transformed, by operation of the dividend, into MMTLP short positions. Or maybe they were imaginary. But in any case, by 2022, “many investors were convinced that short sellers had placed massive, hidden bets against MMTLP.”

This suggests that the 2021 MMTLP short-squeeze plan didn’t work: It turned out short sellers could still be short MMTLP. But the short-squeeze conspiracy playbook worked great: Torchlight was able to pump its stock in 2021 by predicting a short squeeze.

So they apparently tried it again. Replacing MMTLP with Next Bridge shares was a way to make the shares even more non-tradable, to squeeze the shorts even more. Unlike MMTLP, Next Bridge shares, by design, would “not be eligible for electronic transfer through the Depository Trust Company or any other established clearing corporation,” so there is essentially no way at all to trade them, even over the counter. In its prospectus, Next Bridge very explicitly promised that this would cause a short squeeze:

If any investors have sold shares of MMTLP short, then in connection with the Distribution such investors may feel compelled to buy shares of MMTLP to cover such sales before the Distribution. If this were to occur, given the potential high demand from buyers with a relatively low supply of MMTLP shares available for sale on the OTC market, the MMTLP price per share as shown on the OTC market may rise significantly but not be representative of the value of the underlying shares of our Common Stock that you will receive in the Distribution.

And, again, the plan worked. Not necessarily in the sense of causing a short squeeze — who knows — but in the sense of causing retail shareholders to believe in the short squeeze and buy the stock. In June 2023, the Wall Street Journal’s Alexander Osipovich reported:

In their last weeks of trading, buzz mounted that the shares were poised for a short squeeze, a phenomenon in which short sellers exit their bets against a stock, causing its price to rally. Fueled by such chatter, MMTLP surged more than 650% from the start of October [2022] to Nov. 21. That put a peak valuation of about $2 billion on the underlying assets—even though some oil and gas analysts have dismissed them as worthless. 

But then the Financial Industry Regulatory Authority stepped in:

On Dec. 9, 2022, Finra halted trading in MMTLP. Such halts are part of Finra’s tool kit for preventing market disruptions. But the halt surprised adherents of the short-squeeze theory, who expected two more days of trading in which to realize their gains.

Some investors had poured tens of thousands of dollars into MMTLP expecting big profits. Instead, they were stuck with shares they can’t sell.

“Given the lack of drilling success, production or cash flow at Orogrande, it is certainly possible the preferred stock transaction was simply a means to create the perception or reality of a short squeeze,” said Jeff Davies, a former energy hedge-fund manager.

If you bought MMTLP in November 2022, expecting it to surge in December as short sellers hit their deadline for getting out of the stock, and then you held on a bit too long, you were stuck with — exactly what you were promised, non-tradable shares in Next Bridge. You’ve still got them. There is “substantial doubt about the Company’s ability to continue as a going concern,” said Next Bridge, in its most recent 10-Q, from last September. 

What have we learned? I want to emphasize how intense the MMTLP conspiracy theory is. Osipovich wrote last June:

MMTLP investors have formed a noisy protest movement, wielding hashtags such as #FinraFraud. They allege that Finra’s halt covered up a conspiracy among hedge funds and market makers to suppress the price of Torchlight, and later MMTLP, through naked short selling. …

Some Finra employees and market veterans caught up in debate around MMTLP have received death threats. In one anonymous message to a market veteran seen by The Wall Street Journal, the sender alluded to mass shootings and vowed to come “piss on your casket.”

But what the SEC’s case today makes clear is that MMTLP was not a grass-roots movement of retail investors on social media who were aggrieved about short sellers. The whole thing was concocted by the CEO of Torchlight to pump up the stock to raise money. There really was a conspiracy against Torchlight/MMTLP investors, but it was not the short sellers who were conspiring, it was the CEOs of Torchlight and Meta Materials.

Because if you run an unprofitable publicly traded oil company with lots of cash needs and no proven oil reserves, a short-selling conspiracy theory is very valuable to you! If you can get shareholders to believe that (1) all your problems are caused by short sellers, (2) you have found a way to punish the short sellers and (3) your scheme will also push the stock price up, then those shareholders will buy a lot of stock and you can make a lot of money.[5] This is all just as cynical as can be, tricking people into believing that the public markets are rigged against them by shadowy short sellers, in order to take their money yourself.[6]

Or, of course, all of this is wrong, and the SEC is in on the conspiracy too. Always possible.

 

[ValueArb]

Catching up on my Matt Levine columns, from last week interesting discussion of the Berkshire trading glitch and how some Interactive Brokers customers put in market orders trying to get A shares for $200 ended up paying $720,000/share, and how IB covered their loss for them.

 

Quote

There are two ways to buy stock: market orders and limit orders.[1] A limit order says: “I would like to buy 100 shares of Amalgamated Widgets at $20 or less.” If the stock is available at $20 or less, you get your shares. If it’s not, you don’t. You don’t know for sure if your order will be executed, but you do know the maximum price. A market order says: “I would like to buy 100 shares of Amalgamated Widgets at whatever price I can get.” You know the order will be executed, but you don’t know at what price.

Market orders are famously risky, because occasionally prices surprise you. So 99.99% of the time, what happens is that you see Amalgamated Widgets stock trading at $20 per share, and you say “I would like some of that,” and you put in an order to buy 100 shares of Amalgamated Widgets at whatever the market price is, and by the time you press the button on your order and it runs through your broker’s systems and gets to the trading venue and gets filled, the price is, like, $20.01, or $20.02, or $19.98 or whatever, and you get your shares at a slightly different price from the one that you saw on the screen, and you say “ah that’s fine” or “oh well, slippage,” and you understand that pressing the buttons on your retail brokerage’s website is not an exact science but it’s good enough.

And then 0.01% of the time, what happens is that you see Berkshire Hathaway Inc. Class A shares trading at $185 per share, and you say “I would like some of that,” and you put in an order to buy 100 shares of BRK/A at whatever the market price is, and by the time the order gets filled, uh, 90 minutes later, the price is $741,971.39 per share, and you get a bill for $74 million instead of the $18,500 you expected:

On the morning of Monday, June 3, 2024, at approximately 9:50 am EDT, the price of Berkshire Hathaway Class A shares (“BRK A”) suddenly plummeted in the space of a few seconds from approximately $622,000 per share to approximately $185 per share. This occurred as part of an unspecified technical issue at the New York Stock Exchange (“NYSE”). This technical issue and dramatic price event led NYSE to promptly halt BRK A from trading.

News of BRK A’s anomalous price drop quickly spread across social media. Some of the clients of the various brokerage subsidiaries of Interactive Brokers Group, Inc. (together with its subsidiaries, the “Company”), in an apparent attempt to take advantage of this “opportunity,” submitted market buy orders during the trading halt, presumably expecting those orders to be filled at approximately $185/share when trading resumed.

Without any further notice and without addressing a substantial order imbalance that developed during the halt, NYSE resumed trading of BRK A at approximately 11:35:54 am EDT at a price of $648,000. Over the next 98 seconds, the price of BRK A rose to as high as $741,971.39 per share. Many of the Company’s clients that had placed market buy orders during the trading halt were filled at various prices during this run-up, including some who were filled at the peak price. 

Incredible stuff. The main point to make here is that, while $741,971.39 is in some sense the “wrong” price for BRK/A — Berkshire Hathaway Inc.’s extremely high-priced main share class — it is much, much, much closer to the “right” price than $185 is. The stock has traded in the low $600,000s all month, other than the June 3 anomalies, and of those anomalies $185 was considerably more anomalous than $741,971.39. But what seems to have happened is something like this:

1. For glitch reasons, the price fell from $622,000 to $185.
2. Recognizing the glitch, NYSE halted trading.
3. It took almost two hours to restart trading.
4. During that time, NYSE kept taking orders to trade BRK/A, which would execute upon the reopening.
5. Many of those orders came from professional market makers, high-frequency trading firms, etc.
6. Those professional orders would normally take the form of limit orders on both sides of the market, a sort of schedule of supply and demand. A market maker might put in orders like “I’d buy 2 shares at $621,000, and I’d buy 10 more at $620,000, and I’d buy 20 more at $619,000, or I’d sell 2 shares for $623,000, and I’d sell 10 more for $624,000, and I’d sell 20 more for $625,000.” And the exchange would build up an order book of limit orders around what market makers figured was the fair price, with some orders right around that fair price (the “midpoint price”) and others further away. 
7. Also, though, during that time, people “across social media” saw that BRK/A had traded at $185, a 99.97% discount to its real value.
8. They thought “ah, super, a sale on BRK/A.”
9. So they put in limit orders to buy the stock at a price no higher than, say, $190 per share, a bit above where it last traded.
10. No, I’m kidding. I mean, maybe some of them did. Maybe some investors put in limit orders to buy BRK/A during that halt, with limit prices of $185 or $190 or $200 or $1,000. And none of those orders were executed, because by the time the stock reopened it was trading at $648,000. Which is fine. You saw BRK/A trading at $185, you thought “ah, that would be a good deal, if I could actually get it,” you put in an order saying “if I can really get BRK/A at these prices, I’d like to,” and you were slightly disappointed but surely not surprised that you couldn’t. 
11. But at least some Interactive Brokers customers instead put in market orders to buy BRK/A at whatever price it traded at when it reopened. Which was obviously not going to be $185. It was going to be $622,000.
12. Except it wasn’t $622,000 either: So many people apparently put in market orders to buy BRK/A, because they saw it trading at a discount, that “a substantial order imbalance … developed during the halt.” And so the professional market makers, seeing that there were more buyers than sellers in the order book, raised their limit orders, so that the new midpoint price was closer to $648,000. And even so, the imbalance of market buy orders ate through the order book, so that all of the market makers’ sell orders were executed until the stock traded at $741,971.39, and some of those retail orders were filled at that price.
13. And then, once all the retail market orders were executed, the stock went back to normal. BRK/A was trading around $628,000 by 12:15.

In some sense, if you tried to buy Berkshire Hathaway A stock using market orders because you read on social media that you could get it at a 99.97% discount to fair value, and instead you ended up getting it at almost a 20% premium to fair value, that is your fault and you deserve what you get. These people, Interactive Brokers presumes, were not steamrolled by market volatility while they were trying to buy Berkshire Hathaway at a normal price: They were trying to take it from people who were accidentally selling it at a discount, and they deserved to get steamrolled.[2]

But in some other sense, that’s not good customer service, and it’s also kind of risky. (What if you really did have $20,000 in your account and tried to buy 100 shares? How is your broker going to get $74 million out of you?) So that Interactive Brokers press release that I quoted goes on:

The Company determined to take over a substantial portion of these trades as a customer accommodation. The Company also promptly filed claims for compensation with NYSE. On June 25, 2024, NYSE notified the Company that it had denied those claims in full. As a result, the Company has realized losses (including losses on certain hedge transactions) in the amount of approximately $48 million.

If you did this dumb trade, Interactive Brokers fixed it for you: They took over the trades and ended up buying a lot of BRK/A stock for roughly $48 million more than it was worth.[3] Interactive Brokers also “filed a clearly erroneous execution (‘CEE’) petition with NYSE and certain other U.S. exchanges, seeking to bust the trades that had occurred at anomalously high prices during the disorderly market that followed the resumption of trading.”

But the exchanges declined to bust the trades, apparently on the reasoning that those trades were not clearly erroneous. Paying $648,000, or even $741,971.39, for BRK/A shares is not crazy. Paying $185 is. The customers tried to buy the stock at clearly erroneous prices, but they ended up buying it at regular prices. Erroneously. They made the error, not the market.