Money Stuff: Susquehanna Sold Some Suspicious Puts

The basic job of a market maker is to buy low, sell high, and not get adversely selected. You bop along, buying stock at the bid (say, $9.99), selling it at the offer (say, $10.01), and collecting the bid/ask spread ($0.02) on each pair of trades. If people trade with you essentially at random, you keep the whole spread as profit: Someone comes in to sell at $9.99 one second, someone else comes in to buy at $10.01 the next, and you keep the $0.02. 

But if someone sells you a lot of stock at $9.99 one second, and the next second the stock crashes to $9, then you lose $0.99. That is, roughly, adverse selection: buying stock when it is about to go down, or selling it when it’s about to go up. What you want is to trade with people at random; what you don’t want is to trade with people who know which way the price will move.

And so some of the business of market making is about avoiding adverse selection by knowing what the price should be, some of it is about charging a big enough spread to cover your adverse selection losses, and some of it is about trying not to trade with people who know too much. In the US stock market, one way to do that is to pay retail brokers to trade with their customers’ orders: Retail traders, stereotypically, trade at random, so if you buy some stock from them you have no reason to think the stock will go down. [1]

That’s stock market making. Options market making is the same but more so. [2]  Options tend to have wide bid/ask spreads and be bought by retail gamblers, so market makers make a lot of money on options. (Robinhood makes about four times as much money from payment for order flow on options as it does on stocks.)

But options also have a lot of risk of adverse selection. If you sell someone short-dated out-of-the-money call options, they will probably expire worthless and you’ll make money — unless the company announces a merger tomorrow. Then you’ll get blown up. But the people buying those options might know something about a merger. There’s a reason that my Second Law of Insider Trading is about buying short-dated out-of-the-money call options on merger targets. It happens a lot!

That is arguably an annoying example. Most of the time, “adverse selection” doesn’t mean “insider trading.” The more standard sort of adverse selection in market making is, like, a big institutional investor wants to buy a lot of shares, so the price goes up a lot after you sell some. Or a sophisticated hedge fund has done research suggesting that the fundamental value of a stock is above its price. Or a high-frequency trading firm knows that futures have moved up a tick before you do. Stuff like that.

On the other hand, if you are making markets in options to retail customers, adverse selection maybe … does … kind of mean insider trading? The customers are not managing billions of dollars or doing deep sophisticated research or running low-latency feeds from Chicago. If they buy a ton of options and then a merger is announced, that’s probably either luck or insider trading, and if they do it multiple times — or if many of them do it — then perhaps you can rule out luck. 

Perhaps this is unfair. Perhaps it is based on broad rude stereotypes about retail investors. But it is a nice theory for the market maker. Like:

1. You buy and sell options at big markups.
2. Most of the time, the customers are putting in random noise bets and lose money.
3. Sometimes, the customers have big wins.
4. But those don’t count: When the customers have big wins, it’s because the customers cheated.
5. So you shouldn’t have to pay them.

If you ran a literal casino, this logic would be very convenient. They’d win their bet, you’d say “nah you cheated,” and you’d hold onto their money. Perhaps they’d sue you, or file a complaint with the regulators, and you’d eventually have to give them the money. But perhaps they wouldn’t. For one thing, they probably did cheat, so they’d lose in court or with the regulator. And even if they didn’t cheat, what are they gonna say? “No I just did fundamental research and had a hunch that something good would happen to that stock”? Sounds fake.

If you are an options market maker, it’s not quite as easy as that. The options trade on an exchange, the customers aren’t your customers exactly, you don’t control their accounts and you can’t just keep the money. But here’s this:

Susquehanna Investment Group is attempting to unmask the identities of individuals it claims made at least $100 million trading on inside information about a Chinese government crackdown on cross-border brokerages last month.

The Pennsylvania-based market-making firm, which says it was the counterparty on most of the alleged insider trades, sued 100 John Doe defendants in Manhattan federal court on Monday. Susquehanna is seeking to recover more than $70 million it says it lost to what it believes is one of the largest insider-trading schemes in recent memory.

According to Susquehanna, many of those trades were made from accounts at Interactive Brokers Group Inc., as well as the platforms of two firms targeted in the Chinese crackdown, Futu Holdings Ltd. and Up Fintech Holdings Ltd.’s Tiger Brokers. Susquehanna is seeking an order freezing certain accounts at those brokerages and authorizing subpoenas of them.

Here is Susquehanna’s complaint, and here’s its motion to freeze the brokerage accounts. Of course Susquehanna has no direct evidence of insider trading. It doesn’t even know who these traders were, so it certainly doesn’t know whether they were insiders at Chinese regulatory agencies or brokerages, or what information they had when they made their trades. But it knows that they bought short-dated out-of-the-money put options on Futu and Tiger just before China announced the crackdown, that Futu’s and Tiger’s stocks crashed after the crackdown and that the puts made a lot of money, at Susquehanna’s expense. “Insider Trading is the Only Plausible Explanation for the Subject Trades,” says the complaint:

Upon information and belief, there was no new public information about FUTU or TIGR released between May 7, 2026 and May 21, 2026 that would provide a reasonable basis for the Defendants to place so many high-risk, high-reward purchases of short-dated FUTU and TIGR put options. … Yet despite the lack of negative public news about Futu or UP between May 7 and 21, the Defendants placed unprecedented numbers of trades that would only be profitable if new information emerged in a matter of days or weeks that had a significant negative price impact on the stocks. Indeed, many of the options were purchased the day before the Crackdown News on May 22. Those facts provide powerful evidence of trading based on MNPI [material nonpublic information]. …

The Defendants’ identities are unknown to Plaintiffs. However, under any plausible scenario, the Defendants placed the Subject Trades in breach of a fiduciary duty or a duty of trust and confidence. The facts of this case suggest two categories of insiders who could have either traded directly on the MNPI about the imminent Crackdown News or tipped others: (i) Chinese securities regulators; and (ii) Futu and UP personnel who had knowledge of discussions with Chinese securities regulators about the enforcement action.

And so it is (1) suing the people who made the trades, (2) asking the brokers — Futu and Tiger and also Interactive Brokers — to reveal the identities of the people who made the trades, so it can sue them more effectively and (3) asking the court to order Futu and Tiger and Interactive Brokers to freeze their accounts so that, if it wins, it can get the money back. The only possible explanation for these trades is cheating, says Susquehanna, so it shouldn’t have to pay.

“There are few ways in which a man can be more innocently employed than in getting money,” Samuel Johnson said, and I think about that a lot. Here is my extremely biased stylized history of the vibes in tech and finance over the last few decades:

1. After 2008, finance was Evil, and smart quantitative people wanted to be in tech, which was Good. “Don’t Be Evil” was literally Google’s mission statement, back when Goldman Sachs’s was “relentlessly jamming our blood funnel into anything that smells like money.” [3]  Tech was about Changing the World and Building the Future; finance was just about seeking money in exploitative ways.
2. In the subsequent decade or so, people became a bit disillusioned with Big Tech. Google deprecated “Don’t Be Evil” in 2018. A certain cynicism set in about the problems tech was solving; Facebook’s vision of the future was serving up ever more addictive phone content to maximize advertising revenue. Political and cultural and social things happened that I won’t get into. A smart quantitative person in 2020 might have thought “well, if I go to Google I will be building systems to maximize advertising revenue, and if I go to Hudson River Trading I will be building systems to maximize trading revenue, and those things are roughly morally equivalent but HRT doesn’t go around moralizing about it, so I’ll go to HRT.” Finance hadn’t become Good, but it was back to being Neutral; tech had gone from Good down to Possibly A Bit Evil.
3. Then, in about November 2022, the modern artificial intelligence boom started and everyone wanted to work at AI labs, in part because they could make incredible fortunes overnight but also in part because of a real sense of mission. Building artificial general intelligence might be the most important thing humanity ever does, so a smart quantitative person would rather work on that than on extracting short-term trading signals for stock options.
4. Like 20 minutes later everyone started worrying that AI labs, instead of being Good, might actually be Incredibly Incredibly Evil. Like, if you go to work at Anthropic or OpenAI or Google or xAI, you probably get good free snacks, but are you possibly working toward human extinction? Seems bad.
5. If you are a cutting-edge AI researcher, you can also be very useful to a quantitative trading firm or hedge fund, and those guys probably aren’t going to wipe out humanity. [4] They just want to make money.

Have the vibes swung back to finance? I wouldn’t go that far, in part because there’s just soooooooooo much money in AI labs, and the people who like finance mostly like money more. Still, Bloomberg’s Nishant Kumar and Liza Tetley report:

Millennium Management is setting up an artificial intelligence laboratory to expand the development and application of cutting-edge technologies at the firm.

The new lab will become operational over the next few weeks, according to a memo seen by Bloomberg News. It will focus on accelerating early access and assessment of AI products, the memo said, as well as collaborating with AI firms on projects and attracting top AI talent.

The facility would “provide a highly entrepreneurial environment to attract and retain AI talent,” Vlad Torgovnik, Millennium’s chief information officer, said in the memo.

Right, hedge funds do need to be at the forefront of AI. Obviously their AI ambitions are a lot more modest than those of the big frontier AI labs. But maybe that’s good.

The general story about modern hedge funds is that bigger is better. Modern multistrategy multimanager hedge funds identify a mysterious quality known as investing skill, they hire people who have that skill, and they apply those people’s skill to the largest possible opportunity set. Those people are expensive, and the funds hire lots of them to make lots of uncorrelated bets and maximize their risk-adjusted returns.

This is not the only way to run a hedge fund. The Financial Times has a story about alumni of Elliott Management, which is also a giant ($80 billion under management) hedge fund, also runs a bunch of strategies and is also hyper-focused on risk management. But its approach is a bit different from the classic multimanager pod shops:

Elliott now does everything from boardroom fights to takeover battles and distressed-debt brawls, all while maintaining an unusually intense eye on levels of risk.

The phrase Singer has used over the years to describe this phenomenon is “manual effort”, or the attempt to eke out better returns by sheer force of will and resources. Elliott will often have as many as 50 employees devoted to one investment. …

Elliott’s hallmark investments also often combine legal expertise, a savvy for credit documents and sometimes entire takeovers, as eventually took place with the $16.5bn deal for software company Citrix in 2022.

Another former employee said working at Elliott was particularly good for learning “how to prosecute a wide variety of weird and hairy and messy situations”.

The classic pod-shop trades are, like, “buy the stocks that will go up next week and short the ones that will go down,” or “buy the stocks that will be added to the index,” or “buy Treasury bonds and sell futures.” Those trades tend to earn modest returns and benefit from a lot of leverage, and they naturally demand a large scale.

The classic Elliott trades are, like, “notice a mistake in a bond indenture,” [5] or “ruthlessly make fun of a public-company CEO’s drinking problem.” Those trades sort of have whatever scale they have: If there are $500 million of bonds with the mistaken indenture, or if the company with the drunk CEO has a $1 billion market capitalization, then that’s the cap on the opportunity. It seems plausible that drunk CEOs and mistaken bond documents are more common at smaller targets. If you run an $80 billion fund, those trades might not move the needle. It’s not worth it to apply a lot of manual effort to every drunk CEO or mistaken bond indenture.

On the other hand, if you’re an Elliott employee, you can go start your own hedge fund and do Elliott-style trades that are too small for Elliott. The FT reports:

Members of the “Elliott diaspora”, as some former staffers call it, have founded at least seven hedge funds since 2020, mirroring the “Tiger cubs” that came out of Julian Robertson’s Tiger Management around the turn of the millennium.

They include Adam Katz’s Irenic Capital, Dan Gropper’s Carronade Capital, Quentin Koffey’s Politan Capital and James Smith’s Palliser Capital. Unlike Robertson, Elliott has not invested in any of the new firms, according to two people familiar with the matter. But even so, they have begun to find success in the same field.

Scale is a necessity for a pod shop, but it is perhaps an impediment for a weird-and-hairy-situation shop.

My basic model of Strategy and its “Stretch” preferred stock is that it is soft fuzzy banking. That is:

1. Strategy issues Stretch to raise money to buy Bitcoin.
2. Stretch is a perpetual preferred stock whose dividend rate resets each month to make it trade at par. It’s effectively short-term financing that automatically rolls over each month at whatever Strategy’s market-clearing interest rate is.
3. Stretch is intended to work like a money-market instrument, like a bank deposit or money market fund. One dollar of Stretch is supposed to always be worth a dollar; it is supposed to have very little duration. Strategy is essentially funding Bitcoin purchases with bank deposits.
4. But not really. Unlike bank depositors, Stretch holders can’t take their money out whenever they want; Stretch is perpetual. The monthly interest-rate reset is almost economically equivalent to refinancing Stretch every month, but not quite; if financing markets are shut the Stretch stays outstanding. Also, Strategy doesn’t actually have to pay that interest: It’s a preferred stock, and Strategy can always decline to pay the dividend. And while Strategy announced an intent to reset the dividend every month to make Stretch trade at par, you can’t hold it to that intent: It could just reset the dividend to a lower rate and allow Stretch to trade below par.

Soft fuzzy banking is in some ways superior to regular banking: If actual banks could fund themselves like this, it would be nice. On the other hand:

• Actual banks can’t. This only works if you’re a goofy Bitcoin treasury company with an audience of true believers.
• Also the tradeoff is that the cost of this capital is quite high. As of last week, Stretch was (1) paying 11.5% and (2) trading around 75 cents on the dollar, implying that its yield (to trade at par) should be about 15%. So, more than a bank deposit.

As we discussed last week, Strategy is now facing something of a soft fuzzy bank run: As Stretch has traded down, Strategy’s cost of capital has gone up, and it has faced pressure to (1) pay much higher interest on the Stretch and/or (2) pay back the Stretch — just like a regular bank facing a regular bank run. Less so, because all of this is technically optional, but still sort of a bank run. And yesterday Strategy announced that:

1. It no longer intends to keep Stretch at par: It’s raising the interest rate to 12%, not 15%, and in the future “will not necessarily increase the STRC dividend rate solely because STRC trades below its stated amount.” (It closed yesterday at 83.67 cents on the dollar.)
2. It is buying back some Stretch: It’s spending up to $1 billion to buy back its preferred stock, and “currently expects STRC to be the initial priority under the program.”
3. It’s doing assorted other stuff — building its cash reserve, selling some Bitcoin — to improve its liquidity so that it can keep paying all those Stretch dividends.

Some of it — paying higher interest, redeeming deposits, selling assets, shoring up its balance sheet — is what banks would do in a bank run. Some of it — not paying much higher interest, only redeeming some of the deposits (and not at par) — is what banks would prefer to do in a bank run, if they could, but they can’t.

A few weeks ago, ahead of SpaceX’s initial public offering, we talked about the technical complexities of keeping track of orders in a giant IPO. Those complexities are … not very complex? It’s, like, you call your clients and ask them how many shares they want, and then you write their answers down on a list. But we discussed a Bloomberg News story about how various banks and brokers and depositories and service providers were doing various sorts of practice runs to make sure they were ready for the SpaceX IPO, because that IPO would be a severe test of their systems (for writing down lists). I wrote:

On the one hand all of this is true, but on the other hand if SpaceX’s banks come to it tomorrow and say “ahhh there were just too many orders for stock so we lost track of them” that will not be acceptable.

Well! Well. “SpaceX IPO Left Korea Broker With No Shares on Misunderstanding,” Bloomberg reports:

Mirae Asset Securities Co. … inadvertently treated an early request to indicate investor interest as the point at which it had submitted binding orders, the people said. As a result, more than $1.1 billion worth of Korean demand was never entered into the IPO order book, they said. …

In mid-May, weeks before bookbuilding began, the bookrunners circulated an email asking underwriters to indicate investor demand, which was aggregated in a virtual data room in line with standard practice for large deals.

Mirae responded to that request believing it had placed its clients’ orders, according to some of the people, who are familiar with the firm’s thinking. But from the perspective of the Wall Street banks running the deal, those responses were only indications of interest, not bids. The actual orders were entered in June after a separate email from the bookrunners, as is customary for such IPOs.

Right, again, the technical complexities here are just not that complex: They ask you before the IPO “hey how many shares do you think you can sell,” and then they ask you at the end of the IPO “please submit the orders for the shares you sold,” and then they use the second list to allocate you some shares. And if you forget, oops, no shares. “The New York-based banks viewed Mirae as having submitted zero retail orders, and ultimately allocated it zero retail shares.” And:

“We bow our heads in apology for delivering such unfortunate and heavy news to customers who participated in the SpaceX IPO subscription with great interest and anticipation,” Mirae Vice Chairmen Kim Mi-seop and Heo Seon-ho said in a text message to clients on June 15, the Seoul Economic Daily reported. They pledged a review of the process and measures to “restore consumer trust,” according to the newspaper.

Yeah there is no good way to tell clients “sorry, we lost track of your stock orders.”

How the great wealth transfer is rattling Wall Street. SpaceX Pushes US Share Sales to Record $251 Billion at Midyear. Private equity fund investors turn to debt-like deals in downturn. World Bank drops climate finance target under US pressure. Ethiopia Bondholders Criticise IMF for ‘Poorly’ Handled Debt Rework. EY employee charged with accessing Australian prime minister’s bank details. QSBS trust stacking. New York’s Pied-à-Terre Tax Stymies Owners Looking for Loopholes. People are betting on wildfires.

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!