from ABC News:
A Google employee fraudulently made more than $1 million by using inside information to place Polymarket bets on what users were searching for on Google, according to a federal criminal complaint unsealed Wednesday in New York.
...
Spagnuolo, 36, is charged with commodities fraud, wire fraud and money laundering.
"Unlike the counterparties to his trades, Spagnuolo knew the outcome of these wagers before the trading public did because he had accessed Google's confidential, commercially valuable internal data," the complaint said.
He correctly bet -- using an account under the name AlphaRaccoon -- that Google's most-searched person in 2025 would be the singer known as D4vd, according to the complaint. At the time he placed that bet, the prediction market Polymarket "assigned a near-zero probability to d4vd being 'the #1 searched person on Google this year,'" the complaint said.
A few things surprise me about this story.
Two or three months ago, I was at a neighborhood get-together, and I found myself in a conversation with someone who works for TMZ. The discussion turned to d4vd's case. The tabloid news show had been running with it since early September, when the badly decomposed body of a 14-year-old girl was found in the trunk of his Tesla. This brings up my first question: how was this story assigned a near-zero probability on Polymarket?
Putting aside the insider-trading aspect, which we will get back to, this is a remarkable failure of prediction. Even a fairly cursory bit of research into the big stories of 2025 would have revealed the completely unsurprising fact that lots of people were interested in this incredibly lurid story involving a highly popular recording artist. This is a case of a market failing to price in publicly available, easily accessed data.
While Yogi Berra (or perhaps the even more widely quoted apocryphal Yogi Berra) was largely correct when he said, "It's tough to make predictions, especially about the future," some predictions are tougher than others. This is one of the persistent problems in discussing and evaluating prognostications, be they model-based or otherwise. Sometimes predictions are so easy as to be effectively worthless, and it's important to give credit accordingly. The person who predicts the sun will rise in the east may have 100% accuracy, while the one who predicts it will rise in the west may only be right one in 100 times, but it's the second guy we should be impressed with.
The fact that, if not for the guy with inside information, Polymarket would have apparently gotten an easy prediction this wrong raises questions in the "dog that didn't bark" sense.
The other thing that surprises me is that this particular case of insider trading was illegal. To be clear, I'm not saying that it wasn't wrong—it certainly was—or that the FBI should not have gone after this guy—they certainly should have—but given how incredibly forgiving courts and regulators have become toward insider trading and other financial crimes, I honestly don't see this rising to that level.
From Wired:
According to the complaint, Spagnuolo placed trades on Polymarket from around October 2025 to December 2025 using internal Google data. In one instance, he netted $1.2 million trading on who Google’s most-searched person of the year would be in 2025, correctly predicting that the winner would be D4vd, a once obscure singer who became the subject of intense public scrutiny after he was suspected of murder. (D4vd was ultimately charged in the case in April.) ["Once obscure" is an odd framing for a then rapidly rising young performer, but it is technically true in the vast majority of cases. -- MP]
“Unlike the counterparties to his trades, Spagnuolo knew the outcome of these wagers before the trading public did because he had accessed Google’s confidential, commercially valuable internal data,” FBI agent Brandon Racz wrote in the complaint.
Given that the courts currently hold that “insider trading is not about fairness, it’s about theft,” why is the FBI even bringing up the counterparties to the trades?
No comments:
Post a Comment