Thursday links

The beautiful, distracting game

When England played Argentina in the 2002 World Cup, it was in the middle of the trading day, London time. But the market might as well have been closed because trading came to a complete stop.

The three trading screens I sat in front of in those years were jammed full of stock quotes, the prices constantly blinking — green when a price ticked up and red when it ticked down. As soon as the match kicked off, though, there was so little blinking I thought for a moment my screens had frozen.

No one was trading because everyone was watching the game.

A new study has quantified what I saw on my screens: the World Cup disrupts financial markets by robbing them of attention.

“The same cultural phenomenon that generates billions in revenue appears to simultaneously withdraw investor attention from financial markets,” Yosef Bonaparte finds, “compressing trading volume, depressing market sentiment, and producing measurable negative abnormal returns in the US stock market.”

Bonaparte (who I’m guessing is rooting for Les Blues) frames the World Cup “as a natural experiment in mass cognitive reallocation” — and finds that it causes asset prices to deviate from fundamental values in “predictable and measurable ways.”

Fascinating.

He unfortunately begins with an own goal: “For the first time in the tournament’s history,” he says of the current tournament, “the United States is a primary host nation.”

Interesting! Roberto Baggio would be relieved to learn that his memory of blasting a penalty kick over the crossbar in Pasadena, California was just a dream, and not the worst moment in Italian sporting history.

Unfortunately, I can prove it was real: the World Cup hosted by the US in 1994 is why we’ve had to suffer Alexi Lalas as a halftime commentator ever since. (He wasn’t even good at football! Why is he on TV??)

The rest of Bonaparte’s paper, however, is admirably researched, documenting the many ways the World Cup affects markets. 

It makes investors euphoric, for example — or despondent: “Losing a knockout match produces a statistically significant next-day market decline in the losing country,” Bonaparte explains, “while winning nations briefly outperform global benchmarks.”

It disrupts our sleep: “Research examining stock market returns… documents a -26 basis point daily return for a day of sleeplessness associated with late-night matches.” (I might have guessed more, after last night, actually.)

It distracts us: “Football matches are associated with large contemporaneous declines in trading volume in the home markets of competing nations… The result is reduced trading volume, slower incorporation of new information into prices, and attenuated market reactions to firm-specific news.” (I knew that one!)

It makes us make bad decisions: “Heightened gambling sentiment during the World Cup spills over into equity market behavior… elevating demand for lottery-type stocks with high idiosyncratic volatility and skewness.” (Yikes.)

The net impact is not small. The paper cites data showing that “the average return on the US stock market during World Cup effect periods is -2.58%, compared to a benchmark of +1.21% over equivalent non-World Cup periods.”

Primary markets are particularly affected. US IPOs listed during a World Cup are priced 9% lower than otherwise.

I’ve added the emphasis on “US” here because Americans have historically been indifferent to the World Cup. The effect on the stock market, the paper finds, comes entirely from the changed behavior of international investors.

So what happens when Americans start getting distracted, too?

To get an idea, look at the crowds celebrating a US goal in Atlanta last night.

I doubt any of them were buying stocks this morning.

AI world cup bracket

If, like me, you find it difficult to keep track of who’s in what part of this expanded World Cup bracket, try the circular one at WC26.chat. It comes with an LLM attached that’s much easier than Googling.

You can ask it anything: who plays England if England beats Mexico, for example, and where. And when. Plus the odds of every game.

You can even ask where the World Cup was played in 1994.

L’Affaire Donk

The New York Times investigates the controversy over a $5.7 million Polymarket bet on whether an e-sports commentator said “donk” during a livestreamed tournament. 

It’s a colorful story, featuring characters named Dink, Donk, and Lancelot Chardonnet — one of whom was even given that name at birth. (I’ll leave it as an exercise to guess which.)

Amusing details are central to the appeal of prediction markets (I’m not sure there’s much point without them). But the main takeaway here is that Polymarket’s decentralized dispute mechanism has devolved into a Keynesian beauty contest.

UMA tokenholders, who are supposed to use their tokens to vote on the truth, vote instead on how they think others will vote.

I can’t really blame them. As the mechanism is designed, voting with the majority earns fees. Voting against the majority incurs penalties.

Predictably, nearly everyone votes with the majority, whatever the merits of the dispute.

In addition to learning how all that works, we’re also introduced to the term “velar stop slip” — a literal slip of the tongue that can make “don’t” sound like “donk.”

In short, Dink, the commentator, meant to say “don’t” but a velar stop slip made it sound like “Donk” — the name of a famous e-sports player. This should have resolved the $5.7 million mention market to “yes.” But bettors on “no” challenged the outcome and Lancelot Chardonnet, the founder of a Polymarket voting cartel, persuaded everyone to say that Dink only said don’t. 

Got that?

If not, the Times has a perfect summary “The game theory behind the oracle had gone kerflooey.”

Ancient Assyrian capitalist

Ancient clay tablets from Assyria reveal that complex market systems existed nearly four thousand years before economists developed theories to explain them.

The tablets were preserved by a fire that effectively turned the building they were housed in into a pottery kiln, hardening the clay until it became indestructible.

The preserved tablets document financial agreements involving “defined partners, contributed capital, profit-sharing ratios, and a liquidity penalty designed to align the interests of investors with the long-term needs of the enterprise.” In other words, pretty much everything a private-equity attorney or investment banker would need to broker a deal. (Assuming, that is, they can read Old Assyrian.) 

“People bought other merchants’ loan documents and used them as collateral for new loans,” the author notes. “One of the traders got caught smuggling tin in his undergarments to evade a ten percent import tax.”

The tablets add to evidence that sophisticated finance arises organically, no Econ textbooks required. Just the evolutionary drive to profit: “They had prices and trust and the patience to walk a thousand kilometers for a net margin.”

“That was enough,” the author concludes. “It always has been.”

Bitcoin after block rewards

As bad as things are for Bitcoin at the moment, a new study warns that the long-term outlook is even worse: when block rewards fall toward zero, miners will be incentivized to abandon the rules and manipulate the protocol for profit.

The paper models this behavior using a "deviation threshold" to identify the point at which it makes financial sense for a rational miner to begin cheating — by withholding blocks, deliberately splitting the chain, or picking and choosing transactions for private gain.

The author helpfully specifies that the moment to worry is when Gt ≥ ϕ(w) · Xt.

Less helpfully, he does not say when that might be.

The absolute deadline, however, is the year 2140, because that’s when block rewards fall all the way to zero and the “deviation threshold” collapses.

In a fee-only system, a miner only needs to see a tiny extra private payoff — just 0.17% of the block's total value — to switch from honest mining to protocol deviation.

Between now and then, the author says, the Bitcoin protocol will have to evolve to protect itself. He recommends introducing a congestion-adjusting base fee, a strict fee floor, and adaptive block-size limits.

The last of these seems like bad news, considering how contentious the last block-size debate was.

But there’s good news, too, though: algorithms don’t watch football.

— Byron Gilliam