Money Stuff: Biotech Dividend Arrived Early

Sometimes stocks trade at the wrong price. Sometimes this is a debatable, subjective matter: You have done a careful analysis of the company’s future cash flows, and you have determined that the current market price is much lower (or higher) than the value of those cash flows, so you buy (or sell) the stock, chuckle to yourself “all these other people have no idea what the fundamental value of this company is,” and eventually turn out to be right or wrong. More rarely, though, it is a simple objective mistake: You can buy a dollar, a real dollar, for 25 cents. Or you can sell 25 cents for a dollar. If you spot opportunities like that, you should probably seize them. But you should also, you know, triple-check to make sure that you’re right. If a stock price looks like a mistake, you have to be sure exactly what the mistake is and who is making it.

ESSA Pharma Inc. is a small Canadian biotechnology company listed on the Nasdaq; its market capitalization last September was about $300 million. Its mission was to find a treatment for prostate cancer, but last October it announced a failed clinical trial and gave up. The stock price, naturally, collapsed, but Essa wasn’t worthless: It had raised money for its research, and when it terminated that research it had money left over. It decided to wind up the company and give that money back to shareholders.

ESSA consisted of (1) a pile of cash, (2) a public listing and (3) some pretty speculative intellectual property (its prostate cancer treatment that didn’t work). So the company looked for a merger to extract some value from the IP and/or listing. [1] This July, it signed a merger agreement with a company called XOMA Royalty Corp.; the deal was that (1) XOMA would acquire ESSA, (2) ESSA’s existing shareholders would get most of the cash and (3) they’d get a contingent value right entitling them to some value if the legacy assets turned out to be worth anything.

The way the existing ESSA shareholders would get the cash was that ESSA would pay out a dividend before the merger closed. On Aug. 14, ESSA announced that it would pay out $80 million to existing shareholders (about $1.69 per share) on Aug. 22 (last Friday). At the time, ESSA’s stock was trading at about $1.92 per share (a $90 million market capitalization), so almost 90% of the value of the stock represented the right to that cash payout. That makes sense: There was a little stub of speculative legacy asset value, but mostly there was cash.

That dividend announcement, however, contained an off-by-one error. Here it is (emphasis added):

The Nasdaq Stock Market LLC (“Nasdaq”) has determined that the ex-dividend date for the Distribution will be August 25, 2025 (the “Ex-Dividend Date”), the first business day after the payment date. The Distribution is scheduled to be paid to Shareholders on August 22, 2025 and Shareholders will receive approximately $1.69 per Common Share in the Distribution (subject to applicable withholding).

Because the Distribution represents more than 25% of the price of the Common Shares, Nasdaq has determined that the Common Shares will trade with “due bills” representing an assignment of the right to receive the Distribution during the period from August 19, 2025 through and including August 25, 2025 (the “Due Bill Period”). Thus, the Common Shares will trade with this “due bill” and the assignment of the right to receive the Distribution during the Due Bill Period until the Ex-Dividend Date.

Shareholders who sell their Common Shares during the Due Bill Period will be selling their right to the Distribution, and such Shareholder will not be entitled to receive the Distribution (even if the trade will settle after the Due Bill Period). Persons who purchase Common Shares during the Due Bill Period will be entitled to receive the Distribution (even if the trade will settle after the Due Bill Period).

That is: 

1. The dividend was “scheduled to be paid to Shareholders on August 22.”
2. If you bought shares any time “through and including August 25,” you were entitled to get the dividend. Because stock trades settle T+1 — if you buy a share on Monday you don’t get it until Tuesday — buying a share on Aug. 25 meant that you wouldn’t actually own the share until Aug. 26, but that’s okay, because the person who sold you the share owed you the dividend (a “due bill”). [2]  
3. “The ex-dividend date for the Distribution will be August 25, 2025.” The ex-dividend date is the first day on which the stock trades “ex dividend,” meaning without the dividend. If you buy stock before the ex-dividend date, you get the dividend. If you buy stock after the ex-dividend date, you do not get the dividend. [3] If you buy stock on the ex-dividend date, you do not get the dividend. 
4. Thing 2 conflicts with Thing 3. The announcement says that if you buy stock “through and including” Aug. 25, you get the dividend, but it also says that if you buy stock on Aug. 25, you don’t get the dividend.
5. Oops!

From the broader context it seems clear that they meant Thing 3 (no dividend if you buy on Aug. 25), not Thing 2, just because the dividend was paid on Aug. 22 and it would be weird to get the dividend if you bought the stock after it was paid. On the other hand, if you just read the press release, you might well come away with the impression that they meant Thing 2 (yes dividend if you buy on Aug. 25), because the press release says that in so many words — “through and including August 25” — while Thing 3 is expressed more technically (“ex-dividend date”). Does “ex-dividend date” mean “first day without the dividend” or “last day with the dividend”? Well, it means “first day without the dividend,” but maybe you didn’t know that. Apparently no one asked? 

The stock closed at $1.93 on Friday, representing (1) the right to receive $1.69 in cash that day plus (2) a little stub worth about $0.24. As of 10 a.m. this morning, the stock was trading at around $0.25, representing only the little stub (the cash got paid out on Friday).

But yesterday, whoops! If you bought the stock yesterday, you either were or were not entitled to the $1.69 dividend. The stock opened at $0.7333 per share, which was definitely the wrong price: You were either hugely overpaying for a $0.25-ish stub company, or hugely underpaying for a $1.69 cash dividend. And so there was a lot of trading: People who thought they would get the dividend were happy to pay $0.73 for $1.69 in immediate cash, while people who did not think they would get the dividend were happy to sell the stub for $0.73. About 178 million shares — almost $100 million worth — changed hands yesterday; average volume this year before last Friday was under 250,000 shares per day. There were two possible readings of the press release, and differences of opinion are what make a market, so this press release created a lot of trading between people who thought the stock did not come with a dividend (and sold) and people who thought it did (and bought).

It didn’t. The stock was halted at 9:45 a.m., and it didn’t reopen until 4:30. Eventually the company put out a clarification:

The press release issued by the Company on August 14, 2025, inadvertently stated that the Common Shares would trade with “due bills” representing an assignment of the right to receive the Distribution during the period from August 19, 2025 through and including August 25, 2025 (the “Due Bill Period”). The correct Due Bill Period was in fact August 19, 2025 through and including August 22, 2025. Shareholders should be advised that the Distribution was paid to Shareholders on August 22, 2025, and as a result Common Shares did not trade with due bills on August 25, 2025 and instead began trading on an “ex-dividend” basis as of such date.

Ah well. Sometimes you really can buy $1.69 of cash for $0.73. Sometimes, though, the press release is just wrong. I assume there will be lawsuits.

For much of US history, sports gambling was regulated by the states. It was illegal in most states, with Nevada being the main exception. But “by the 1990s, there were signs that the trend that had brought about the legalization of many other forms of gambling might extend to sports gambling,” and the US Congress passed a federal law to stop that. [4] The Professional and Amateur Sports Protection Act of 1991 essentially forbade the states from legalizing sports gambling, with a few exceptions (again mostly Nevada). So starting in 1991, sports gambling was not a matter of state regulation: The federal government mostly prohibited states from allowing it.

In 2018, though, the Supreme Court struck down that law, finding it unconstitutional. “PASPA,” wrote the Court, “‘regulate[s] state governments’ regulation’ of their citizens. ... The Constitution gives Congress no such power.” So starting in 2018, sports gambling was again a matter for state regulation, and the trend of legalization did continue: Most states now allow some legal sports gambling, though about a dozen do not. 

But this year the system changed again: The federal government has apparently again prohibited states from regulating sports gambling. But the direction has flipped: Unlike the PASPA in 1991, now the federal rule is that states are prohibited from stopping sports gambling. 

Also unlike the PASPA, this did not involve any new legislation. The new rule isn’t clearly stated anywhere. There’s no federal law saying “states can’t regulate sports gambling.” The new rule is an emergent result of regulatory silence. What happened is approximately that the US Commodity Futures Trading Commission was given the power to regulate “event contracts,” that is, prediction markets, and a prediction market named Kalshi became a CFTC-regulated futures exchange. As a futures exchange, Kalshi is allowed to “self-certify” new event contracts, which then become listed for trading unless the CFTC steps in to stop them. [5] Kalshi started self-certifying event contracts on sporting events, like who would win football games. The CFTC actually has a rule prohibiting event contracts that involve “gaming,” and when the rule was enacted everyone apparently agreed that “gaming” included sports betting, so it is not clear to me why Kalshi thinks that these contracts are legal. But the important thing is that, in 2025, the CFTC — which is quite Kalshi-friendly and might soon be headed by a Kalshi board member — has not objected to those contracts, so Kalshi can go ahead and trade sports bets.

And because Kalshi is a CFTC-regulated commodity futures exchange, that means that — however accidentally and unintuitively — its contracts are federally regulated commodities trades, and federal commodities regulation preempts state gaming regulation. (In the olden days, people thought that trading corn futures was gambling, and states tried to ban it, so federal commodities law preempted those efforts.) And so Kalshi started offering sports bets nationwide, and various state regulators sent it letters saying “hey you are doing sports gambling without complying with regulation,” and Kalshi told them “buzz off.” And Kalshi went to court, saying that it is illegal for the states to regulate its sports gambling offerings. And courts mostly agreed!

We have talked about this history a couple of times before. I find it extremely weird! Donald Trump did not run for the presidency on a platform of “states shouldn’t be allowed to regulate sports betting,” and nobody introduced a law saying that, and in fact as far as I can tell the CFTC’s rules still prohibit futures exchanges from offering sports bets, but here we are anyway. Sports betting is now a form of commodities futures trading, futures exchanges offer sports bets, and states can’t stop them.

Meanwhile the federal government now taxes 100% of gambling winnings, while allowing deductions for only 90% of gambling losses, making it extremely difficult to be a professional gambler. But you can still be a professional commodities trader, since all of your commodity trading losses are still deductible. So you pretty much have to do your sports gambling on a futures exchange, where the taxes are lower and the states can’t stop you. 

Maybe. Probably? Anyway there is some news.

First: Most courts have agreed with Kalshi that the states can’t stop them, but since we last discussed this, one court has disagreed. This month, a Maryland federal judge “observed that the states have a historically significant interest in regulating gambling activity, concluding that Kalshi had failed to show that it had a likelihood of success on the merits of its suit and denied its motion for a preliminary injunction.” So it seems like Maryland can still regulate sports gambling, for now.

Second: While Kalshi mostly seems to be allowed to offer sports gambling everywhere, the real prize is Robinhood. Robinhood is a giant retail stock brokerage firm that has partnered with Kalshi to offer sports bets to its stock-trading clients: The contracts are traded on Kalshi, but Robinhood provides the customers and the interface to let them make the bets. Apparently state regulators, even after losing to Kalshi in court, have told Robinhood not to offer sports bets. So last week Robinhood also went to court, suing the Nevada Gaming Control Board to let it offer sports bets. From its complaint:

Despite [the Kalshi] rulings, the Board continues to threaten to enforce preempted Nevada law against Robinhood, even though the Board is currently enjoined by this Court from doing so against Kalshi with respect to the same transactions. On May 6, 2025, after this Court’s decision in KalshiEx, … Robinhood met with the Board and explained that it believed it should be able to offer sports-related event contracts trading through Kalshi’s exchange for as long as this Court’s order in KalshiEx remains in effect. At the conclusion of that meeting, Board employees indicated they did not expect to be able to agree to refrain from enforcement action against Robinhood, even while the KalshiEx order is in place.

“Robinhood has activated its Nevada customers’ access to sports-related event contract trading,” it adds, in violation of Nevada law, and so “Robinhood now faces an immediate threat of civil penalties and criminal prosecution from the Board,” but it would prefer not to, so it sued. “The CFTC has already impliedly approved those same event contracts,” it says: By letting Kalshi trade them, the CFTC has silently prevented Nevada from stopping them.

Third: Sure yes this is fun for Robinhood and Kalshi, but it is a bit rough on traditional sports books, which still have to follow state law and whose customers now can deduct only 90% of their losses. The obvious solution for them is to become futures exchanges. Meanwhile, I suppose the traditional futures exchanges — where you can trade, like, corn futures or interest rates — presumably want to get in on the gambling business. The trade is obvious:

Flutter Entertainment’s FanDuel is dipping a toe into the popular but legally unsettled realm of prediction markets. The move might eventually pay out in a big way, according to analysts.

The sports-betting platform is teaming up with the exchange company CME Group to offer event-based contracts that would let FanDuel users place “yes” or “no” wagers on financial events and benchmarks, such as where the S&P 500 closes or where the price of gold lands on a given day.

The rollout, expected this year, is Flutter’s cautious early step into the wider world of prediction markets and might ultimately lead it to making markets for sports-related events, analysts said. In traditional sports wagering, odds are set by a bookmaker. Sports-related event contracts are priced based on how other traders are betting in the market.

A key difference: Sports-prediction markets aren’t subject to the same taxes and regulations as traditional sports wagering. If Flutter can get in on those markets, it could reel in new customers from states that haven’t legalized sports betting, analysts said.

Terrific. [6]  I keep saying that this is extremely weird, but in a sense it does feel right that, in 2025, gambling and financial markets are merging completely.

That is the folk wisdom, and if folk wisdom is around long enough eventually there will be an economics paper supporting it. Here is a new National Bureau of Economic Research working paper by Joshua Gans:

When should travellers leave for the airport? This paper develops a model for optimal airport arrival timing when travellers face uncertain travel times and can potentially board earlier flights. We show that access to earlier flights creates a “recourse option” that fundamentally changes optimal behaviour. While earlier flights always reduce the probability of missing one's scheduled departure, they may paradoxically increase expected waiting time when travellers adjust their arrival strategies.

I feel like “can potentially board earlier flights” is not in the traditional problem specification but this one is fun too.

Here’s an X post from Ruslan Kogan, chief executive officer of Australian online retailer Kogan.com Ltd., explaining that a robot did his earnings call:

In our $KGN earnings call today, it sounded like it’s @davidmshafer  and I presenting but it wasn’t. Our voices were AI generated and read the script. We uploaded several hours of our prior talks to train our voice model and the results were pretty incredible. This saved us a few hours preparation - valuable time we could spend working on other areas of the business.

Really? So: They wrote the 25-minute script for the earnings presentation, the robot read the script, and then they answered questions? I guess? It would have taken them several hours to prepare to read the script? I feel like it would take me approximately 25 minutes to read a 25-minute earnings-call script? I suppose they had that 25 minutes to check email while the robot was reading the script, but still. Man. Have the robot write the script, sure, that’s time-consuming. (Make sure it doesn’t get anything wrong!) Have the robot answer analysts’ follow-up questions, sure, that’s impressive. But having the robot read the script so you don’t have to seems like a pretty minimal time savings.

On the other hand, a lot of the action in modern AI-for-earnings-calls is along the lines of “tone and body language,” where the CEO reads the earnings script and the investors’ artificial intelligence tools analyze his voice to see if he’s optimistic or panicking or lying or whatever. I suppose having your AI read the script makes you a bit harder to analyze.

Trump Moves to Fire Fed’s Cook, Setting Up Historic Legal Fight. US long-term debt sells off after Trump’s attempted firing of Fed governor. UniCredit Raises Physical Stake in Commerzbank to 26%. Trump Jr. Joins Polymarket Advisory Board as 1789 Boosts Stake. Trump Channels Hatred for Wind Farms Into Strike Against Orsted. JPMorgan, Already No. 1 in Card Spending, Wants Even More. AT&T to Buy Spectrum Licenses from EchoStar for $23 Billion. Elon Musk’s xAI Sues Apple and OpenAI, Alleging They Are Monopolists. Perplexity Is Launching a New Revenue-Share Model for Publishers. Squirrel honored with Topps card after interrupting Yankees game.

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!