Retail bias or thin liquidity? Polymarket’s Brazil odds face scrutiny

Polymarket predicted Luiz Inácio Lula da Silva had about a 20% chance of winning the Brazilian election. However, Lula won on Sunday despite his low odds, a blow to the prediction market world that raises questions about the real determinants of prediction odds.

The left-wing Lula managed to defeat his right-wing opponent Flávio Bolsonaro in a second round of voting. The second round was needed as Bolsonaro was leading the first one in October but couldn’t gain a majority. This time, Lula managed to secure a majority of votes.

Prior to the voting, Polymarket predicted Lula’s chances of winning at 20% while Bolsonaro’s stood at around 80%. On such platforms, the contract settles at $1, which means that a 60-cent “yes” price meant that there was a 60% chance that Lula would win. Thus, pre-voting odds presented Lula’s victory as an unlikely scenario. However, the news of Lula’s win caused the Brazilian stock market to fall.

A Polymarket whale blames retail

One of the big traders says the issue was the crowd. FhantomBets, an anonymous pseudonym for Polymarket‘s whale, stated in a report that veteran bettors realized the odds weren’t right and Lula’s probability of winning was actually more like 30% to 40%.

The disparity was most apparent after counting the votes. “The retail traders were so much against Lula that you could get unlimited size at 60-80c with Lula having [essentially] already won,” FhantomBets said, noting that Lula was certain by 6:15 pm ET but his “yes” contract was only priced at 60 cents. He noted that the whale and other traders exited their long positions to enter the Brazilian market.

The ban that thinned the book

The professionals-vs.-amateurs explanation does not necessarily hold true for everybody. As reported, Luke Nolan, head of research at CoinShares, stated that he would exercise caution when interpreting the results as smart traders winning against retail. According to him, increased participation could mean increased liquidity and insights into the situation.

According to both Nolan and Einar Knobel, CEO and co-founder of prediction-market platform TxODDS, there is an obvious reason for that: The government of Brazil banned prediction markets in April. They cited investor protection and increased problems with gambling addiction. Finance Minister Dario Durigan claimed that the platforms violated betting regulations and did not receive any regulation from their side. Therefore, since local players were locked out, Knobel stated that the price was formed “by people outside the country” who were trading the poll headlines and first-round narrative. The volume corresponded to that of only 4.5% of the activity seen in the 2024 U.S. presidential market.

Was it really a failure?

Nolan disagreed with this label of blowout. A 20% probability, said Nolan, is still something that can happen. And an outcome expected at a 20% probability should come true once out of five times, Nolan argued. In his opinion, this could just have been one such event.

Moreover, Nolan thought that the markets performed better than the traditional betting companies, pointing to Betfair’s odds of 83% for Hillary Clinton on the eve of the U.S. elections of 2016. In Nolan’s opinion, it was not just one wrong prediction but a pattern of overconfidence which should be alarming.