Political betting lines are relatively unprofitable for both retail traders and companies in the space. 

Traders struggle with the relative infrequency and inconsistency of political events that lend themselves to betting with a sustained edge. 

Paul Krishnamurty, as an example, one of the most prominent political gambler in the world, recently estimated that only about five percent of his winnings as a professional gambler over the past 20 years have come from politics, the rest being from sports. 

While sportsbooks make money off political lines, in large part by banning accounts perceived to have an edge, companies offering traditional prediction markets often lose money on political markets and justify their costs as loss leaders, academic endeavors, or as a public service. 

Even state legislators who favor legalized election betting acknowledge that revenues from political lines would be relatively small. In a state like Nevada, with a mature gaming industry, political bets would amount to a “tiny” subset of sports betting revenues.

Conservative forecasts about the commercial viability and revenue generating potential of political prediction markets have been borne out to some extent by publicly available data on PredictIt’s users and finances. In their filing to the CFTC, plaintiffs associated with PredictIt tout the fact that, between 2014 and 2022, more than 120,000 participants have traded on the site. 

Activity on PredictIt peaked in the run-up to the 2020 U.S. presidential election, when close to $150 million was traded on the site. This is the only time in the site’s history when PredictIt revenues exceeded its expenses. 

In the run-up to the 2022 U.S. midterm elections, the site averaged around 80,000-100,000 users, 30,000 of which PredictIt described as active traders. While an increase from 22,000 users in 2016, the site has not grown its user base much since 2018. 

Additionally, only a small percentage of PredictIt users trade on a weekly or monthly basis, and most do not appear to trade large amounts of money. 

In 2020, the average first time deposit was $216. PredictIt’s no-action letter requires the site to limit each market to 5,000 participants, but that number is typically only reached in a handful of marquis election markets close to election day. 

As a consequence, PredictIt’s expenses have exceeded revenues over the history of the site’s operation.

Currently, PredictIt’s user base and revenues are almost certainly below its 2022 levels. 

Beyond the fact that engagement on the site drops in non-election years, PredictIt has seen an exodus due to its regulatory constraints and uncertainties. 

PredictIt’s counsel disclosed before the Fifth Circuit that 14,478 traders held positions in 75 contracts on the site that were expected to expire after February 15, the deadline the CFTC gave PredictIt in August 2022. In the months following the August 2022 notice, many of the site’s highest-volume, prominent traders withdrew funds, particularly after the November 2022 midterm elections and the December 7 Georgia run-offs. 

Between August and December 2022, traders withdrew $18 million from PredictIt; 7,500 traders withdrew all of their funds and shut down their accounts. More than 4,700 of these traders did so between November and December. 

By December, trading volume on PredictIt lost more than 75% of the volume it had before the CFTC’s August 2022 announcement. 

Exacerbating doubts about the site’s future has been PredictIt’s refusal to clarify how it would resolve existing markets if legal appeals fail. 

PredictIt’s decision not to create new markets or even, until late June 2022, add new candidates to its election market has further diminished interest in the site. Lingering traders have been left with low liquidity, highly volatile markets.

Evidence from the UK suggests that the limited interest and profitability of political betting in the United States is only partly related to regulation. 

On the one hand, in the absence of restrictive regulations, major political events—notably U.S. presidential elections—have been among the largest single and most profitable betting events for UK bookmakers. 

Compared to $261 million ($281.9 million adjusted for inflation) in the 2016 U.S. election, $744 million was bet on Betfair Exchange alone on the 2020 U.S. presidential election, making it the largest betting event in the site’s history. These sums far surpassed comparable sporting events. For comparison, the August 2017 fight between Conor McGregor and Floyd Mayweather, the most-bet-on sporting event Betfair had offered to date, brought in $71.5 million. 

On Betonline.ag, more money was bet on the 2020 election than that year’s Super Bowl.

The absence of betting limits appears to have drawn ‘smart whales’ to take advantage of mispriced odds driven by retail traders. Although 46.6% of the money wagered was on Trump compared to 50.6% for Joe Biden, the 10 largest bets placed on Betfair were for Biden. These include a $1.3 million bet on Biden, which was the third largest wager in Betfair history, and at least six other Biden bets of more than $400,000. 

Nevertheless, election betting revenues in the UK pale in comparison to lines such as horse racing and soccer, which present far more repeatable betting opportunities than elections. 

The limited profitability potential of current affairs markets is compounded by the fact that operators must hire experts who can write complex rules for infrequent, unrepeatable events that can be understood by traders well enough to avoid ambiguous settlement situations. 

The UK-based betting exchange Smarkets, as an example, was founded by an American CEO, Jason Trost, who became interested in the space after learning about political prediction markets. However, the economics of the industry have led Smarkets to treat current events markets as a niche “passion project” subsidized by sports and other lines more likely to see growth. 

Because Smarkets’s goal of producing a publicly-available, “very accurate set of probabilities” with “social utility” requires the site to attract “clever, clued-up, informed” bettors, Smarkets deprioritizes profitability considerations for its current events lines and welcomes long-term winning accounts even as their competitors restrict them. 

Nevertheless, Trost concedes that he “wouldn’t put too much stock” in Smarkets’s political markets as an “information source” because they are “thinly-traded” and lack “critical mass” and a sufficiently “diverse group of bettors.”

While the 2020 elections point to a favorable trend for political betting, the awareness they have generated for political betting may prove to be a fleeting phenomenon. 

Industry veteran Matthew Shaddick has cautioned that recent spikes in political betting involvement, which crossed $1 billion globally in 2020, may largely stem from the unique “worldwide fame and notoriety and interest” that Trump personally engenders. 

Shaddick questions whether more conventional politicians would produce comparable levels of “excitement” in the betting markets. 

These doubts make it difficult for companies to expand product development, hire talented specialists in the political prediction betting space, and invest in the lobbying, public relations, and research efforts needed to influence regulators and build the political forecasting community. Even when they do occur, such investments tend to be inconsistent, as this area is one of the most obvious and immediate targets when companies are forced to cut costs.

Pratik Chougule is the executive director of the Coalition for Political Forecasting. Solomon Sia is a board member of the Coalition for Political Forecasting. This post is an excerpt from the authors’ report “Political Betting Regulation in the United States: Pathways to Liberalization.”